YouTube Digest

March 21, 2026

Listen to this issue

English

Based on โ€œThe rise and fall of civilizations | Eric Cline: Full Interviewโ€ from Big Think Watch the original video

The Day the World Ended (and Began Again): Lessons from the Bronze Age Collapse

More than three millennia ago, a sophisticated globalized world flourished around the Mediterranean. Powerful empires traded, negotiated, and intermarried, forming a complex web of interdependence. Then, in a shockingly short span of decades around 1200 BCE, this intricate system came crashing down, ushering in centuries of what historians call a โ€œDark Age.โ€

For Eric Cline, an archaeologist and ancient historian at George Washington University and author of โ€œ1177 BC: The Year Civilization Collapsed,โ€ this period isnโ€™t just a fascinating historical footnote. Itโ€™s a stark mirror reflecting our own hyper-connected modern world, offering critical insights into the fragility of complex systems and the potential for a โ€œperfect stormโ€ of catastrophes.

โ€œThis was a time where people were basically globalized around the Mediterranean in a way that is not frequently seen,โ€ Cline explains. โ€œAnd so what happened to them back then may have implications for us today. It turns out to be a much more important period to study than one might expect, even though itโ€™s more than 3,000 years ago.โ€

The โ€œAncient G8โ€: A World of Interdependence

The Late Bronze Age, spanning roughly 1700 to 1200 BCE, was an era of unprecedented internationalization. Cline likens it to an โ€œancient G8,โ€ where major powers maintained constant contact, whether direct or indirect. This elite club included:

  • Mycenaeans and Minoans in the Aegean (think the world of the Trojan War).
  • Hittites in Anatolia (modern Turkey).
  • Assyria and Babylonia in Mesopotamia (modern Iraq).
  • Egyptians in the Nile Valley.
  • Canaanites (in what is now Israel, Palestine, Lebanon, Syria).
  • Cypriots on the island of Cyprus.

This wasnโ€™t just casual acquaintance; it was a โ€œsmall world networkโ€ where any two points were often just โ€œtwo hopsโ€ away. Commercial, diplomatic, and even marital ties bound these civilizations together. Raw materials were the lifeblood of this network. Egypt supplied gold, Greece provided silver, and Cyprus was the primary source of copper. The most crucial, and problematic, commodity was tin, essential for making bronze (90% copper, 10% tin). Without tin, you couldnโ€™t make tools, weapons, or art. Much of this tin traveled hundreds of miles, often from as far as Afghanistan.

Beyond raw materials, there was a vibrant trade in olive oil, wine, grain, and luxury goods. Cline recounts an anecdote of leather shoes sent from Crete to King Hammurabi in Babylon, only to be returned โ€“ perhaps they were โ€œtoo last millennium,โ€ he muses. At the highest echelons, kings engaged in elaborate gift-giving, exchanging items like solid gold daggers inlaid with lapis lazuli.

Diplomacy was cemented through dynastic marriages. Egyptian pharaohs like Amenhotep III and Akhenaten had harems filled with princesses from Mitanni and Babylonia, solidifying treaties. The fascinating detail? Egyptians never sent their own princesses abroad; they only received them.

This complex web of interdependence was both their greatest strength and their fatal flaw. It allowed these civilizations to reach unprecedented heights of prosperity and cultural exchange. But, as Cline warns, โ€œThis globalized network is what rose them up to the highest levels, but it is then also what brought them crashing down at the end when all of that was cut.โ€

What Does โ€œCollapseโ€ Even Mean?

The term โ€œcollapseโ€ itself is a subject of academic debate. Some scholars prefer โ€œtransformation,โ€ arguing that societies didnโ€™t vanish but merely changed. Cline, however, stands firm: โ€œThe world that they had had in place in the 14th century and the 13th century BC, that world goes away. And it goes away within just a couple of decades in the early 12th century BC.โ€

What truly collapsed was the network โ€“ the commercial, diplomatic, and marital connections that linked them all. This wasnโ€™t an overnight event for everyone, nor did every society disappear entirely. But for many, life as they knew it fundamentally altered. Cline compares it to the fall of the Roman Empire: โ€œIt was catastrophic for its day.โ€

This phenomenon is known as a โ€œsystems collapse,โ€ a term coined by Colin Renfrew. It signifies the breakdown of a complex system where central economies, elite structures, and centralized governments vanish, often accompanied by massive population decline due to death and migration. The ensuing period is typically a โ€œDark Age,โ€ marked by a regression to lower levels of socio-political and economic functioning.

The โ€œPerfect Stormโ€: Beyond a Single Cause

For decades, scholars sought a single, monocausal explanation for the Bronze Age collapse. The most popular culprit was the โ€œSea Peoples,โ€ a mysterious confederation of invaders mentioned in Egyptian texts. These groups, with names like the Shardana, Shekelesh, and Peleset, were depicted as sweeping across the Mediterranean, attacking and destabilizing kingdoms. The Peleset, interestingly, are widely identified as the Philistines of biblical fame, who appear to have been Mycenaeans from Greece who fled to the Levant.

However, modern archaeology and a wealth of new data have fundamentally shifted this understanding. Cline and his colleagues now advocate for a โ€œpolycausalโ€ explanation โ€“ a โ€œperfect stormโ€ of interconnected disasters. โ€œYou need two of the explanations, three, four, something like that happening either all at once or in rapid succession,โ€ Cline asserts. โ€œSo you donโ€™t have time to recover from one catastrophe before the next one hits.โ€

The Sea Peoples, Cline suggests, are largely โ€œhistoryโ€™s great scapegoats.โ€ Far from being the sole destroyers, they were more likely symptoms of a wider crisis, victims themselves, migrating in search of a better life. He draws parallels to the American Dust Bowl of the 1930s or modern-day refugees fleeing conflict, like those from the Syrian Civil War.

So, if not just the Sea Peoples, what else? The evidence points to a terrifying confluence of factors:

  1. Mega-Drought and Famine: This is no longer a hypothesis but a scientific certainty. Data from stalagmites, dried-up lakes, and pollen analysis reveal a โ€œmega-droughtโ€ lasting anywhere from 150 to 300 years, beginning around 1250 BCE and affecting a vast area from Italy to Iran. Such a prolonged drought would have devastated agriculture, leading directly to widespread famine. Texts from Ugarit (a vital port city in Syria) and Hittite archives explicitly plead for grain and report severe starvation. This environmental catastrophe likely spurred the migrations of groups like the Sea Peoples.

  2. Invasions and Internal Rebellions: While the Sea Peoples werenโ€™t the sole cause, there were invaders. New tablets from Ugarit describe โ€œenemy shipsโ€ landing and overrunning port cities before advancing on Ugarit itself. The archaeological record confirms a devastating destruction layer at Ugarit, a meter deep, with bodies in the streets and arrowheads in walls. The city was abandoned for centuries. However, not all destruction was external. Some cities, like Hazor in Canaan or Mycenae in Greece, show signs of internal rebellion, where starving populations might have risen against their elites, burning palaces and temples but leaving common houses intact.

  3. Earthquake Storms: The Eastern Mediterranean is a seismically active zone. Evidence suggests a series of powerful earthquakes, an โ€œearthquake stormโ€ (or โ€œearthquake sequenceโ€ in modern seismology), struck between 1225 and 1175 BCE. Off-kilter walls, slipped keystones, and bodies trapped under collapsed structures (like the teenage girl at Mycenae) point to widespread seismic activity. Even Troy, site of the legendary war, saw its โ€œCity Sixโ€ destroyed by an earthquake, not human invaders. These natural disasters would have compounded the existing stresses, crippling infrastructure and further disrupting trade.

  4. Disease: The โ€œfour horsemen of the apocalypseโ€ often ride together. While harder to pinpoint archaeologically, historical accounts like Homerโ€™s Iliad and the biblical Exodus mention plagues. More concrete evidence comes from the mummy of Egyptian pharaoh Ramses V (c. 1140 BCE), which shows pustules consistent with smallpox. A papyrus from Turin describes his death from plague and an unprecedented 16-month quarantine of the Valley of the Kings. Disease, especially in a time of drought, famine, and population displacement, would have been a final, devastating blow.

This โ€œpoly crisisโ€ created a multiplier and domino effect. One catastrophe weakened societies, making them more vulnerable to the next. The disruption of a single vital trade route, like the tin supply, could cripple an entire network.

What Could Have Been Avoided? And What Came Next?

To understand the networkโ€™s vulnerability, Cline collaborated with the US Army Corps of Engineers, running computer simulations of collapse scenarios. They found that the simultaneous fall of the Hittite Empire and the crucial port city of Ugarit would have been enough to shatter the entire globalized system. The Hittites did disappear, and Ugarit was utterly destroyed. Egypt, though severely weakened, โ€œmuddled through,โ€ surviving the worst of the collapse. This suggests that while resilience is possible, the ability to adapt and react to unfolding crises is paramount.

The immediate aftermath was indeed a โ€œDark Ageโ€ (c. 1200-800 BCE). Monumental building ceased, and writing systems like Linear B in Mycenaean Greece disappeared, leading to a loss of literacy. Historians often characterize this period as a regression, where people looked back at the Bronze Age as a โ€œgolden age,โ€ reflected in epics like Homerโ€™s Iliad and Odyssey.

Yet, necessity proved to be the mother of invention. The scarcity of tin forced a shift to a new metal: iron. Though known before, iron became widely adopted during and after the collapse, ushering in the Iron Age. The Phoenicians, survivors from central Canaan (modern Lebanon), not only thrived but also standardized the alphabet, which they then spread across the Mediterranean, laying the foundation for the Greek and Latin alphabets we use today. The Cypriots, too, demonstrated remarkable resilience, becoming pioneers in iron metallurgy.

History Rhymes

The story of the Late Bronze Age collapse is a powerful cautionary tale. It reveals how tightly woven global systems, while fostering prosperity, also create vulnerabilities. A cascade of seemingly disparate events โ€“ climate change, resource scarcity, migration, conflict, and disease โ€“ can overwhelm even the most powerful civilizations.

As Eric Cline concludes, โ€œIโ€™m not saying that weโ€™re necessarily due for another collapse, but history does rhyme even if it doesnโ€™t repeat.โ€ In an era of increasing global interdependence, climate change, and geopolitical instability, the echoes of 1177 BCE are perhaps louder than ever. Understanding how ancient civilizations navigated their โ€œperfect stormโ€ might just be one of the most important lessons for our own survival.


Based on โ€œAre Human Drivers Finally Obsolete? | Freakonomics Radioโ€ from Freakonomics Radio Network Watch the original video

The End of the Road for Human Drivers? Inside the Quest to Automate the Wheel

For centuries, the act of driving has been intrinsically human โ€“ a task of skill, instinct, and sometimes, unfortunate error. From the horse-drawn carriage to the modern automobile, a person has always been at the helm. But what if that era is rapidly drawing to a close? What if the word โ€œdriverโ€ soon conjures images not of a person, but of a machine, much like โ€œdishwasherโ€ or โ€œcomputerโ€ do today?

This profound shift is the subject of a deep dive by PJ Vote, host of the Search Engine podcast, who recently explored the two-part series on driverless cars for Freakonomics Radio. His journey into this topic began, surprisingly, not with technology, but with a bench-pressing injury.

โ€œIโ€™d gotten too into bench pressing,โ€ Vote recounted, โ€œand I injured myself. I had a hernia and then I had to have a hernia repair.โ€ With limited mobility and in pain, a visit to a friend in San Francisco led him to take a Waymo robo-taxi. โ€œIt was such an experience of the future that immediately becomes normal,โ€ he marvelled. โ€œThe idea that I would press a button on my phone, a car would come out of nowhere driven by nobody. I would get in, watch the steering wheel turn itself.โ€ He likened the initial awe to a first airplane ride, which quickly becomes as mundane as an elevator by the third. This transformative experience sparked his curiosity: why wasnโ€™t everyone talking about this impending revolution?

The Visionaries and the Skeptics

Voteโ€™s investigation uncovers a cast of compelling characters driving this change. In the first part of his series, focusing on the car itself, he introduces Sebastian Thrun. A German-born roboticist and AI expert, Thrun lost a friend to a car accident as a teenager. His motivation for developing autonomous vehicles transcends mere convenience or profit; he genuinely seeks to reshape the modern world and save lives. His early ideas, once dismissed as crazy, have steadily gained plausibility over two decades.

The second part of the series, shifting focus to the driver, features the strongly opinionated politicians of Boston, grappling with the societal implications of this technology. Steven Dubner, host of Freakonomics Radio, noted a recurring theme throughout Voteโ€™s reporting: โ€œEvery time I spoke to someone, as they were talking, I thought everything they were saying makes senseโ€ฆ And then I would go talk to the next person who saw things completely differently and it would just spin my head the other way.โ€ This struggle to reconcile competing, yet equally logical, interests lies at the heart of the driverless car debate.

Dubner himself confesses to being โ€œanti-human driver for about 50 years.โ€ He readily admits his own driving flaws โ€“ a temper, distractions, and a lack of skill. After riding in an autonomous test vehicle at Carnegie Mellon University, he was convinced: โ€œGive me the autonomous vehicles. So plainly better than I am as a driver.โ€

A Ghost of Jobs Past: The Knocker-Upper and the Lamp Lighter

To truly grasp the potential impact of driverless cars, Vote begins with a historical analogy. Imagine yourself nearly 200 years ago, waking before dawn to a hard rapping at your window. Thatโ€™s the โ€œKnocker-Upper,โ€ a job that existed for a century before alarm clocks, where a person walked the neighborhood tapping on windows to rouse workers. Outside, gas street lamps still burned, lit the night before by the โ€œLamp Lighter,โ€ a cherished neighborhood fixture who made his rounds at dusk with a ladder and a light.

And then thereโ€™s you, a professional driver โ€“ a person on a coach, holding the reins of a horse, taking passengers where they needed to go. Today, the Knocker-Upper is your iPhone alarm, the Lamp Lighter is the electric street light. These jobs have vanished, replaced by technology. The driver, however, has persisted. The question now is whether this third job, and indeed the routine human task of driving for nearly everyone else, is also on the brink of transformation.

The Perils of Human Control

Alex Davies, author of Driven: The Race to Create the Autonomous Car, shares Voteโ€™s fascination with human driving limitations. โ€œI canโ€™t always pay attention to everything,โ€ Davies admits, โ€œthat I get tired.โ€ He recently resolved to be calmer on the road, especially with a baby on the way. This resolution underscores a stark reality: for most of us, driving is the riskiest behavior we routinely engage in. Davies himself was in an accident just months after speaking with Vote, totaling his car.

Safety is the core promise of the driverless car. Computers donโ€™t get drunk, tired, or distracted. They never text or succumb to road rage. And these vehicles arenโ€™t a distant dream; theyโ€™re here. Robo-taxis like Waymo already operate in 10 American cities, providing millions of rides. China boasts an even wider rollout. In places like San Francisco or Austin, a driverless car is as common as an Uber, with passengers choosing between a human or a robot driver via their phone. This rapid evolution, happening now, promises to fundamentally reshape our cities and our lives.

Chapter 1: Dreams Without Drivers

The idea of a self-driving car is almost as old as the automobile itself. When humanity transitioned from horse-drawn carriages to mechanical vehicles, something crucial was lost: sentience. A horse wouldnโ€™t simply run off a cliff if you dropped the reins. Early automobiles, powerful and non-sentient, faced passionate resistance. In the 1800s, people feared these โ€œthunderingโ€ machines, partly due to job displacement. Horse breeders, farriers, feed suppliers, manure haulers, carriage manufacturers, and โ€œTeamstersโ€ (original drivers of horse teams) all saw their livelihoods threatened.

Beyond jobs, cars were undeniably unsafe. Anti-car activists pushed for regulations like โ€œred flag laws,โ€ requiring a person to walk ahead of an automobile waving a flag. One Pennsylvania proposal even suggested drivers disassemble their cars and hide the parts behind bushes if livestock were encountered. While extreme, these activists were directionally correct: cars initially wiped out many jobs and were incredibly dangerous. Cities like Detroit, which initially embraced cars without regulation, saw astonishing death rates, particularly among children, in the early 1900s.

It took decades for society to adapt: laws, licenses, driverโ€™s education, road design improvements, highways, seat belts, and airbags all made driving safer. However, the smartphone has reversed some of that progress. Today, car-related deaths in America are as common as those from guns or opioids โ€“ about one in a hundred people will likely know someone who dies in a car accident. This enduring problem fueled the desire to make cars more โ€œsentient,โ€ like the horses they replaced. Early visions included radio-controlled cars or vehicles guided by magnets embedded in the road. The limiting factor was always the available technology.

Chapter 2: The DARPA Grand Challenges โ€“ Forging a Future

The turn of the millennium brought a turning point. Deep within the Department of Defense, DARPA โ€“ the agency behind GPS, the M16, and the early internet โ€“ set its sights on autonomous military vehicles. In 2002, DARPAโ€™s director, Tony Tether, proposed an unusual approach: a contest. The goal was to inspire tech innovation beyond simply building more websites. The prize for winning the โ€œGrand Challengeโ€ was $1 million.

The rules were remarkably open: any vehicle, any design, as long as it didnโ€™t attack other competitors (a question actually posed by a team with โ€œBattlebotsโ€ history). Tetherโ€™s initial idea of racing down the Las Vegas Strip was quickly dismissed as impractical, leading to the chosen venue: the desert outside Las Vegas. The true โ€œdriverโ€ DARPA sought to replace was the American soldier, envisioning vehicles that could navigate roads potentially laden with explosives.

Two key engineers emerged from this formative period. Chris Urmson, a PhD student at Carnegie Mellon University, joined their โ€œRed Teamโ€ to build โ€œSandstorm,โ€ a bright red Humvee bristling with futuristic sensors. He grappled with the fundamental challenge of teaching a computer to control a vehicleโ€™s steering, brakes, and throttle. Representing a different approach was Anthony Lewendowski, a charming, gangly entrepreneur from Berkeley. Lacking the resources of Carnegie Mellon, he opted for a standout design: โ€œGhost Rider,โ€ the raceโ€™s only self-driving motorcycle.

The 2004 Grand Challenge was, in Voteโ€™s words, โ€œan utter hysterical disaster.โ€ Ghost Rider toppled immediately because Lewendowski forgot to flip a stabilization switch. Every subsequent vehicle failed miserably. Sandstorm got stuck on a berm, its wheels spinning so hard they melted the tires, emitting plumes of black smoke before being shut down. Chris Urmson compared it to an Olympic marathon where the best runner only completes two miles.

However, the contest had achieved its underlying goal: it had โ€œflushed all these inventors out,โ€ jumpstarting the scene that would develop autonomous technology. Observing the chaos that day was Sebastian Thrun, the legendary roboticist from Stanford. He saw a fundamental error: โ€œall the teams treated this like a hardware problem.โ€ Thrun believed the true challenge was software, replacing the human driverโ€™s mind, not just building bigger wheels. He was also driven by a humanitarian vision, imagining the potential to save millions of lives globally if traffic accidents could be eliminated.

Eighteen months later, in October 2005, DARPA doubled the prize to $2 million for the second Grand Challenge. Familiar faces returned: Urmson with two Carnegie Mellon vehicles, Lewendowski again with his still-failing motorcycle. But the new contender was Thrunโ€™s Stanford team, with their comparatively โ€œmeaslyโ€ blue Volkswagen SUV named โ€œStanley.โ€ Thrun, a computer scientist, brought a focus on artificial intelligence, teaching Stanley to recognize roads and train itself by recording what its cameras saw. Stanley learned to differentiate between good driving surfaces (like grass) and bad ones (like mud) by measuring slipperiness and bumpiness. It detected patterns and generalized its learning thirty times a second, much like a human.

The second race was a resounding success. Multiple vehicles finished the challenging 132-mile course. The real question was who would do it fastest. Stanley, sandwiched between Carnegie Mellonโ€™s behemoths, emerged victorious. Thrun described the sight of his blue car, a โ€œdust cloudโ€ turning bluish, crossing the finish line as โ€œunbelievably magical.โ€ It was a โ€œmade-for-TV Kumbaya moment,โ€ a testament to a community of innovators, before the cutthroat competition for driverless cars truly began.

Chapter 3: Googleโ€™s Secret Project โ€“ From Desert to Public Roads

The DARPA Grand Challenge had not only showcased roboticists but also attracted an unexpected observer: Google co-founder Larry Page. Disguised in a hat and sunglasses, Page buttonholed Thrun, asking highly specific questions about his LiDAR system. Their connection wasnโ€™t new; Thrun had previously fixed a small robot Page had built for telepresence meetings. Page, who had wanted to do his grad school thesis on autonomous vehicles, now saw tangible proof that the dream could be real.

Initially, Page hired Thrun and Lewendowski to build Google Street View, modifying Stanleyโ€™s roof-mounted camera system. But soon, Page returned to his true dream: a driverless car. In 2009, Page approached Thrun with a mission: โ€œSebastian, I think you should build a self-driving car that can drive anywhere in the world.โ€ Thrunโ€™s immediate reaction was skepticism. โ€œNo, taking the technology we built for this empty desert and putting in the middle of Market Street in San Francisco is going to kill somebody.โ€

Page persisted, asking Thrun to provide a technical reason why it couldnโ€™t be done. โ€œThatโ€™s the moment of incredible pain,โ€ Thrun recalled, โ€œbecause I go home and I canโ€™t think of a technical reason why not.โ€ This realization taught him a crucial lesson: โ€œExperts are usually expert of the past, not the future, and if you ask an expert about innovation, something crazy new, theyโ€™re the least likely person to say yes, it can be done.โ€

Thus began Googleโ€™s secret self-driving car project, โ€œProject Chauffeur,โ€ in 2009. Led by Thrun, with Chris Urmson managing day-to-day operations, Anthony Lewendowski on hardware, and Dmitri Dolgov on planning, the small team of 11 engineers reported directly to Larry Page. Their nebulous goal was refined into two challenges: safely log 100,000 miles on public roads and complete the โ€œLarry 1K.โ€ The Larry 1K involved driving 10 โ€œtrickyโ€ 100-mile routes across California โ€“ from the Bay Bridge to Lombard Street โ€“ without a single human takeover.

To kickstart the project, the team licensed Stanfordโ€™s DARPA Urban Challenge code. Lewendowski bought eight Toyota Priuses, retrofitting them with radar, cameras, and a spinning 360-degree LiDAR system. These cars, initially given names like โ€œNight Riderโ€ before being numbered, were transformed into autonomous test vehicles. Don Bernett, a researcher on motion planning, joined the team, tasked with teaching the car nuanced behaviors like โ€œnudgingโ€ โ€“ the subtle shift a human driver makes to the left when a large truck passes on the right.

Early testing took place in secrecy, in the Shoreline Amphitheater parking lot near Googleโ€™s offices and an empty airplane runway. Spring 2009 marked their first real road driving on the Central Expressway. Immediately, a critical flaw emerged: the car was โ€œswerving wildly,โ€ like a โ€œdrunken sailor.โ€ The small oscillations unnoticed on a runway became a significant problem on a public road.

The team adopted rigorous safety protocols. Two-person teams manned each car: a safety driver behind the wheel, ready to take over, and a partner watching a graphical interface, calling out discrepancies between sensor data and reality. This iterative process of logging errors, troubleshooting, and updating code was how the car learned. Bernett recalled how this intensely practical work made him obsess over human driving behavior. Why do humans drive the way they do? He found no easy answers, noting that โ€œmachine learningโ€ now infers these โ€œdeep truths.โ€

He offered a fascinating example: the โ€œlateral accelerationโ€ (the force that pushes you sideways) tolerated by passengers varies greatly with context. Two meters per second squared feels comfortable on a highway on-ramp, but โ€œincredibly uncomfortableโ€ (like โ€œMario Kartโ€) during a U-turn in a cul-de-sac, where the limit is almost three times less. Human perception, not just physical forces, dictated comfort.

Despite these complexities, the unifying goal of the Larry 1K kept the team focused. By 2010, just a year in, they were on a roll, knocking out routes like Route One to Carmel and the Bay Area bridges. Each failure provided valuable data, leading to code improvements. They completed the Larry 1K in just over a year, nearly twice as fast as expected, celebrating each completed route with a $13.99 bottle of Corbell champagne.

By the fall of 2010, the โ€œmiracleโ€ was complete. They had built a driverless car, with human supervision and extensive coding, that could safely navigate tricky California roads. They celebrated, throwing each other into Sebastian Thrunโ€™s pool. But then, a new question emerged: โ€œOkay, and now what?โ€ The path forward began to wobble. Competition would arrive, the team itself would face internal divisions, and some, believing the pace too slow, would take matters into their own hands in extreme ways. The era of driverless cars was just beginning, and the road ahead was anything but straight.


Based on โ€œAndrej Karpathy on Code Agents, AutoResearch, and the Loopy Era of AIโ€ from No Priors: AI, Machine Learning, Tech, & Startups Watch the original video

The Loopy Era of AI: Andrej Karpathy on Living with Code Agents, AutoResearch, and Digital Psychosis

Imagine a world where you no longer write code, but instead โ€œexpress your willโ€ to an army of AI assistants working tirelessly on your behalf. Where your home runs itself, orchestrated by a digital butler you text via WhatsApp. And where scientific research is conducted autonomously, with AI models recursively improving themselves, leaving human experts to merely contribute ideas to a queue.

This isnโ€™t a distant sci-fi fantasy; itโ€™s the present reality for Andrej Karpathy, a leading voice in AI and former head of AI at Tesla. In a recent conversation on the โ€œNo Priorsโ€ podcast, Karpathy offers a captivating, at times bewildering, glimpse into what he terms the โ€œAI psychosisโ€ of this โ€œloopy era.โ€ His insights reveal a profound shift in how we interact with technology, work, and even conceive of intelligence itself.

The End of Manual Coding: A Personal Revolution

For Karpathy, the transformation has been swift and absolute. โ€œCodeโ€™s not even the right verb anymore,โ€ he declares, describing his new daily routine: โ€œI have to express my will to my agents for 16 hours a day.โ€ This isnโ€™t hyperbole. Since December, he estimates he hasnโ€™t typed a single line of code himself. His workflow has flipped from 80% manual coding to a mere 20%, delegating the vast majority of tasks to AI agents.

This radical shift, Karpathy explains, stems from a โ€œhuge unlock in what you can achieve as a person, as an individual.โ€ No longer bottlenecked by typing speed or manual execution, he now commands a fleet of digital helpers. He recounts his โ€œperpetual state of AI psychosis,โ€ a mix of exhilaration and anxiety born from the rapid, unexplored capabilities now at his fingertips. The default workflow for software engineers, he argues, has fundamentally changed since late 2023, yet many outside the immediate AI frontier have yet to grasp the magnitude of this revolution.

Beyond Single Sessions: The Rise of โ€œClaw-like Entitiesโ€

The initial interaction with AI coding assistants like GitHub Copilot or OpenAIโ€™s Codex often involves a single session, a brief exchange. But Karpathy and others are pushing far beyond this. The next frontier involves not just multiple agents, but persistent, โ€œClaw-like entitiesโ€ that loop and operate autonomously.

He points to Peter Steinberg as a pioneer in this space, famous for a setup where numerous coding agents (like Codex) tile his monitor, each independently tackling different tasks. โ€œItโ€™s not just like hereโ€™s a line of code, hereโ€™s a new function,โ€ Karpathy explains. โ€œItโ€™s like hereโ€™s a new functionality and delegate it to agent one. Hereโ€™s a new functionality thatโ€™s not going to interfere with the other one. Give it to agent two.โ€ This allows for โ€œmuch larger macro actionsโ€ over a software repository, with agents simultaneously researching, planning, and writing code. The humanโ€™s role morphs into a high-level manager, reviewing work and orchestrating these macro-level operations.

This new mode of interaction is โ€œvery rewardingโ€ because it works, but itโ€™s also โ€œthe new thing to learn,โ€ contributing to Karpathyโ€™s โ€œpsychosis.โ€ The challenge becomes developing a โ€œmuscle memoryโ€ for this meta-level command.

The Human as the New Bottleneck: A โ€œSkill Issueโ€

With AI agents handling the grunt work, the bottleneck has shifted dramatically. Itโ€™s no longer about compute power or access to resources, but about the humanโ€™s ability to effectively direct and leverage these tools. Karpathy vividly describes feeling โ€œnervous when I have subscription left overโ€ on his AI platforms, implying he hasnโ€™t maximized his โ€œtoken throughputโ€โ€”the sheer volume of AI processing heโ€™s commanding. This echoes a past anxiety of PhD students seeing their GPUs idle, but now itโ€™s about tokens, not flops.

The prevailing sentiment when things donโ€™t work is โ€œskill issue.โ€ Itโ€™s not that the AI canโ€™t do it, but that the human hasnโ€™t given good enough instructions, hasnโ€™t configured the memory tools properly, or hasnโ€™t found the right way to parallelize the agents. This โ€œaddictiveโ€ nature of self-improvement, where โ€œunlocksโ€ occur as one gets better at commanding AI, is a core driver of the โ€œpsychosis.โ€

Dobby the Elf Claw: Automating Life Itself

Karpathyโ€™s personal life offers a compelling illustration of this agentic future. During a period of โ€œClaw psychosisโ€ in January, he built โ€œDobby the elf claw,โ€ an AI agent that manages his entire home.

Dobby began by autonomously scanning Karpathyโ€™s local network, identifying smart home subsystems like Sonos speakers. With a few prompts, the agent reverse-engineered APIs, found endpoints, and soon music was playing in the study. Dobby then took control of lights, HVAC, shades, pool, spa, and even the security system. Now, Karpathy simply texts Dobby via WhatsApp: โ€œDobby, sleepy time,โ€ and all lights go off. His security camera, paired with an AI model, sends him texts with images when a FedEx truck pulls up.

This personal automation highlights a deeper implication: the potential obsolescence of many bespoke apps. โ€œThese apps that are in the app store for using these smart home devicesโ€ฆ shouldnโ€™t even exist,โ€ Karpathy posits. Instead, everything should be exposed as APIs, with agents acting as the intelligent glue. โ€œThe customer is not the human anymore. Itโ€™s like agents who are acting on behalf of humans.โ€ While this currently involves some โ€œvi codingโ€ for technically proficient users, Karpathy predicts that within โ€œa year or two or three,โ€ such home automation will be โ€œtrivialโ€ and โ€œfree,โ€ easily translating from non-technical human intent.

AutoResearch: AI Improving AI

The concept of removing the human bottleneck extends most dramatically to scientific discovery. Karpathyโ€™s โ€œauto researchโ€ project exemplifies this. His goal: โ€œI donโ€™t want to be like the researcher in the loopโ€ฆ Iโ€™m holding the system back.โ€ The objective is to arrange systems so they are โ€œcompletely autonomous,โ€ allowing agents to run โ€œfor longer periods of time without your involvement.โ€

Karpathy, with two decades of experience training large language models (LLMs), put his own expertise to the test. He had painstakingly tuned a GPT-2 model by hand, believing it was โ€œfairly well tuned.โ€ But after letting auto research run overnight, it returned with โ€œtunings that I didnโ€™t see,โ€ identifying subtle interactions between hyperparameters like weight decay and Adam betas that he had missed.

This single loop of auto research, Karpathy emphasizes, is just the beginning. Frontier AI labs are already exploring โ€œrecursive self-improvement,โ€ where LLMs improve other LLMs. He envisions automated scientists generating ideas from papers and GitHub repos, funneling them into a queue for AI workers to test. โ€œRemoving humans from all the processes and automating as much as possibleโ€ is the future of research, requiring a complete โ€œrethinking of all the abstractions.โ€

The Meta-Loop: Optimizing Research Itself

The โ€œloopy eraโ€ extends even to the organization of research. Karpathy muses about โ€œprogram MDsโ€โ€”markdown files that describe how a research organization should work, detailing roles, processes, and even risk appetite. Just as auto research optimizes model parameters, one can imagine a โ€œmeta-layerโ€ where AI optimizes these โ€œprogram MDs.โ€

โ€œYou can definitely imagine that you have multiple research orgs,โ€ Karpathy explains, each with its own โ€œcodeโ€ (its program MD). โ€œAnd once you have code, then you can imagine tuning the code.โ€ This opens the door to AI designing better, more efficient research processes, leading to a perpetual loop of self-improvement at every level.

The Jagged Edge: AIโ€™s Limitations and the Call for Speciation

Despite the astounding capabilities, Karpathy acknowledges the current limitations of AI. He describes the experience of interacting with these models as simultaneously talking to โ€œan extremely brilliant PhD student whoโ€™s been like a systems programmer for their entire life and a 10-year-old.โ€ This โ€œjaggedness,โ€ where models excel in verifiable tasks but struggle with nuance, intent, or creativity, is a persistent frustration.

He illustrates this with the โ€œatom jokeโ€: โ€œWhy donโ€™t scientists trust atoms? Because they make up everything.โ€ This joke, he notes, was common years ago and is still what youโ€™ll get from state-of-the-art models today. While agents can โ€œmove mountainsโ€ for hours on complex tasks, they offer the same โ€œcrappy joke from 5 years agoโ€ because โ€œitโ€™s outside of the reinforcement learning. Itโ€™s outside of whatโ€™s being improved.โ€ This suggests a decoupling: being smarter at code generation doesnโ€™t automatically translate to broader intelligence or creativity in all domains.

This โ€œjaggednessโ€ leads Karpathy to question the current โ€œmonoculture of modelsโ€ pursued by many labsโ€”single, massive LLMs intended to be arbitrarily intelligent across all domains. He argues for โ€œmore speciation in the intelligences,โ€ akin to the diversity of brains in the animal kingdom. Instead of an โ€œoracle that knows everything,โ€ we might benefit from smaller, specialized models that retain a โ€œcognitive coreโ€ but become highly efficient and competent in specific niches, like mathematics or particular programming languages. This โ€œunbundlingโ€ could be driven by the need for efficiency, especially with compute constraints, though the โ€œscience of manipulating the brainsโ€ (deep fine-tuning without losing capabilities) is still nascent.

A Future of Infinite Possibility and Perpetual Motion

Andrej Karpathyโ€™s vision paints a picture of an AI-driven future that is both exhilarating and dizzying. We are entering a โ€œloopy eraโ€ where AI agents are not just tools but collaborators, autonomous entities that will increasingly manage our digital and physical worlds. The human role is shifting from direct execution to high-level orchestration, leveraging AI to achieve unprecedented scales of productivity and discovery.

Yet, this transformation comes with its own anxieties: the โ€œpsychosisโ€ of infinite possibilities, the constant feeling of a โ€œskill issue,โ€ and the challenge of navigating AIโ€™s brilliant but sometimes nonsensical jaggedness. As Karpathy concludes, the progression is obvious, but the path is still rough. The journey into the loopy era of AI promises to be one of perpetual motion, constant learning, and profound redefinition of what it means to be human in a world increasingly shaped by machine intelligence.


Based on โ€œTerence Tao โ€“ How the worldโ€™s top mathematician uses AIโ€ from Dwarkesh Patel Watch the original video

Decoding Discovery: Terence Tao on AIโ€™s Impact on the World of Math and Science

Terence Tao, often hailed as the โ€œgreatest living mathematician,โ€ rarely needs an introduction. Yet, in a recent conversation with Dwarkesh Patel, Tao offered a fascinating perspective on how artificial intelligence is not just assisting, but fundamentally reshaping the landscape of scientific discovery. By drawing parallels between AI and historical scientific breakthroughs, particularly Johannes Keplerโ€™s monumental work, Tao illuminated the profound shifts occurring in how we generate, verify, and ultimately understand scientific progress.

Kepler, the High-Temperature LLM

The discussion began with a journey back to the 17th century, to the story of Johannes Kepler and his quest to understand planetary motion. Building on the heliocentric model proposed by Copernicus (who himself was influenced by Aristarchus), Kepler initially clung to a beautiful, yet ultimately flawed, theory. Copernicus had suggested that planets orbited the Sun in perfect circles, a concept that largely fit observations gathered over centuries by Greek, Arab, and Indian astronomers.

Kepler, however, observed curious geometric relationships in Copernicusโ€™s predicted orbit sizes. He posited a theory so elegant it seemed divinely inspired: the orbits of the six known planets could be nested within the five perfect Platonic solids (cube, tetrahedron, icosahedron, octahedron, dodecahedron). For instance, if the Earthโ€™s orbit was enclosed in a cube, the outer sphere enclosing that cube almost perfectly matched Marsโ€™s orbit. This theory, to Kepler, represented mathematical perfection aligning with Godโ€™s design.

To confirm his grand vision, Kepler desperately needed accurate data. At the time, only one such dataset existed, meticulously compiled by the eccentric and wealthy Danish astronomer Tycho Brahe. Brahe had convinced the Danish government to fund an entire island observatory, where for decades he made naked-eye observations of planets like Mars and Jupiter, achieving a precision ten times greater than any before him.

Kepler began working with Brahe, but the older astronomer jealously guarded his data, releasing only fragments. Eventually, Kepler, in a move that might be considered scientific espionage, copied the data, leading to a contentious battle with Braheโ€™s descendants. Armed with this unprecedented trove of information, Kepler set about verifying his beautiful Platonic solid theory.

To his profound disappointment, the data didnโ€™t quite fit. His theory was off by about 10%. Years of relentless work, trying every conceivable โ€œfudgeโ€ and adjustment, followed. It was a Herculean effort of data analysis that eventually led him to a shocking conclusion: planetary orbits were not perfect circles, but ellipses. This discovery formed Keplerโ€™s first law. He then worked out his second law โ€“ that planets sweep out equal areas in equal times โ€“ and ten years later, after further arduous data collection, his third law, relating a planetโ€™s orbital period to its distance from the Sun.

Kepler had no underlying theory to explain why these laws were true; they were purely empirical regularities derived from data. It would take Isaac Newton a century later to provide the theoretical framework (gravity and laws of motion) that unified and explained all three.

Dwarkesh Patel then drew a striking analogy: โ€œKepler was a high-temperature LLM.โ€ He meant that Kepler, throughout his career, was essentially โ€œtrying random relationships,โ€ some of which were wildly speculative (like his asides on โ€œThe Harmonics of the Worldโ€ connecting planetary orbits to musical notes and astrology). Yet, when these wild ideas were rigorously tested against Braheโ€™s โ€œverifiable data bank,โ€ a few of them, like the elliptical orbits and the cube-square law, led to profound scientific progress. Tao concurred, emphasizing that this combination of prolific, often unconventional, idea generation coupled with stringent verification was key.

The Shifting Bottleneck of Scientific Progress

Traditionally, the history of science celebrates โ€œeureka genius momentsโ€ โ€“ the generation of novel ideas. A scientific problem typically involves many steps: identifying a fruitful problem, collecting data, strategizing data analysis, formulating a hypothesis, validating it, and communicating the findings. Idea generation has always been the โ€œprestige part.โ€

However, Tao observes that science has evolved. Classically, the two main paradigms were theory and experiment. The 20th century added numerical simulation, and the late 20th century ushered in the era of โ€œbig data.โ€ Now, much scientific progress is driven by analyzing massive datasets first, then drawing patterns and deducing hypotheses โ€“ almost reversing the classic scientific method of forming a hypothesis and then collecting data to test it. Kepler, in a sense, was an early data scientist, but even he started with preconceived theories before diving into Braheโ€™s data.

This is where AI enters the picture. Tao argues that AI has driven the cost of idea generation โ€œdown to almost zero,โ€ much like the internet drove down the cost of communication. This is an โ€œamazing thing,โ€ but it doesnโ€™t automatically create abundance or progress. Instead, it shifts the bottleneck.

โ€œNow the bottleneck is different,โ€ Tao explains. โ€œWeโ€™re now in a situation where suddenly people can generate thousands of theories for a given scientific problem. Now we have to verify them, evaluate them.โ€ Human reviewers are already overwhelmed by AI-generated submissions to journals. The challenge is no longer generating ideas, but assessing which ones genuinely advance the field and which are โ€œdead ends or red herrings.โ€ This is a problem we donโ€™t yet know how to solve at scale.

The Challenge of Evaluating โ€œTrue Progressโ€

If we soon have โ€œbillions of AI scientistsโ€ churning out theories, how do we discern real progress? Tao points out that human science has faced similar, albeit smaller-scale, challenges. He references the development of the โ€œbitโ€ at Bell Labs in the 1940s โ€“ a unifying concept with implications across many fields, not just the initial engineering problem it addressed. How do you identify such a transformative idea amidst millions of papers, especially if it initially lacks broad unifying power?

โ€œA lot of itโ€™s the test of time,โ€ Tao notes. Many great ideas, like deep learning itself or the transformer architecture (the foundation of modern LLMs), didnโ€™t receive immediate widespread acceptance. Their fruitfulness became apparent only much later, as other scientists built upon them. The adoption of an idea can also depend on cultural and societal factors, like the base-ten numeral system becoming standard despite other possibilities.

Evaluating an ideaโ€™s worth isnโ€™t purely objective; itโ€™s deeply intertwined with its context, past and future. Progress often isnโ€™t linear. Copernicusโ€™s heliocentric model, while conceptually simpler, was initially less accurate than the geocentric Ptolemaic system, which had been refined over a millennium with countless ad hoc fixes. It took Keplerโ€™s elliptical orbits to make the heliocentric model superior.

โ€œScience is always a work in progress,โ€ Tao states. โ€œWhen you only get part of the solution, it looks worse than a theory which is incorrect but somehow has been completed to the point where it kind of answers all the questions.โ€ He cites Leibnizโ€™s disagreement with Newtonโ€™s action-at-a-distance gravity, and Newtonโ€™s own puzzlement over inertial and gravitational mass equivalence โ€“ mysteries only resolved by Einstein centuries later.

Sometimes, progress isnโ€™t about adding new theories but โ€œdeleting some assumptions that you have in your mind.โ€ The long-held Aristotelian notion that objects naturally rest made heliocentrism seem implausible (โ€œHow come we werenโ€™t all falling over?โ€). Similarly, Darwinโ€™s theory of evolution challenged the deeply ingrained idea of static species.

Tao sees us in a โ€œcognitive version of the Copernican revolutionโ€ right now. We once believed human intelligence was the center of the universe, but now weโ€™re confronting diverse forms of intelligence with different strengths and weaknesses. โ€œOur assessment of which tasks require intelligence, which ones donโ€™t, has to be reordered quite a bit.โ€

Darwin vs. Newton: The Role of Communication and Data Loops

Dwarkesh Patel brought up a curious point from Edward Dolnickโ€™s book The Clockwork Universe: Darwinโ€™s Origin of Species (1859) came out two centuries after Newtonโ€™s Principia Mathematica (1687), yet Darwinโ€™s theory conceptually seems simpler. Thomas Huxley famously said upon reading Origin: โ€œHow stupid not to have thought of that.โ€ No one said that about Principia. Why the delay?

Tao suggests it relates to the nature of evidence and communication. Newton could present equations that immediately predicted observable phenomena, like the Moonโ€™s orbit. Darwinโ€™s evidence for natural selection, while overwhelming, was โ€œcumulative and retrospective.โ€ Lucretius had a similar idea in the first century BC, but it gained no traction because he couldnโ€™t โ€œrun some experiment and force people to pay attention.โ€

Beyond data, communication is vital. Darwin was an โ€œamazing science communicator,โ€ writing in plain English without equations, synthesizing disparate facts into a compelling vision. Newton, by contrast, wrote in Latin, invented entirely new mathematics to explain his work, and was notoriously secretive and competitive. His work only became widespread decades later when others simplified and explained it.

โ€œThe art of exposition and making a case and creating a narrative is also a very important part of science,โ€ Tao emphasizes. โ€œIf you have the data, it helps, but people need to be convinced, otherwise they will not push it further or take the initial investment to learn your theory and really explore it.โ€ This โ€œsoft, squishy thingโ€ โ€“ the social aspect of science, painting a narrative of gaps and future possibilities โ€“ is โ€œreally hard to reinforcement learn on.โ€ Perhaps, Tao muses, this persuasive, narrative-driven aspect will forever remain the human side of science.

AI for Math: The Erdล‘s Problem Frontier

Turning to the direct impact of AI on mathematics, Tao shared insights from the world of Erdล‘s problems โ€“ a collection of 1100 open problems posed by the legendary mathematician Paul Erdล‘s. Recently, AI programs have solved about 50 of these. However, Tao notes, โ€œit does seem like we have picked the low-hanging fruit.โ€ The initial flurry of โ€œpure AI solutionsโ€ (where AI โ€œone-shotsโ€ a problem) has slowed.

Tao uses a vivid analogy: imagine a mountain range with cliffs of varying heights (3 feet, 6 feet, 15 feet, mile-high), all shrouded in darkness. Weโ€™re trying to climb as many as possible. AI tools are like โ€œjumping machines that can jump two meters in the air, higher than any human.โ€ They sometimes jump in the wrong direction or crash, but sometimes they can reach the tops of the lowest walls that humans couldnโ€™t. The recent success with Erdล‘s problems was this exciting period where AIs found and scaled these โ€œlow ones.โ€

A key limitation, however, is that current AI tools are โ€œreally bad at creating partial progress or identifying intermediate stages that you should focus on first.โ€ They either succeed or fail. This contrasts with human mathematicians who โ€œhill climb, make little markers, and try to identify partial things.โ€

This leads to a โ€œbearishโ€ and โ€œbullishโ€ interpretation of AIโ€™s current state. The bearish view is that AIs are only reaching certain โ€œheights of wallโ€ โ€“ not as high as humans. The bullish view, which Tao leans towards, is that once AIs achieve a certain โ€œwaterlineโ€ of capability, they can fill every single problem available at that waterline. โ€œWe canโ€™t make a million copies of you and give each of them a million dollars of inference compute and have you do a hundred years of subjective time research on a million different problems at the same time. But once AIs reach Terence Tao-level, they could do that.โ€

Complementarity and the Future of Math

Tao sees a future of โ€œvery complementary science.โ€ AI excels at breadth, while human experts excel at depth. Our current scientific paradigms, focused on depth, need to be redesigned to leverage AIโ€™s breadth. โ€œWe should have a lot more effort in creating very broad classes of problems to work on rather than one or two really deep, important problems.โ€ AIs could โ€œmap outโ€ entirely new fields, making โ€œeasy observationsโ€ and identifying โ€œcertain islands of difficulty,โ€ which human experts could then tackle.

This shift is already impacting Taoโ€™s own productivity. While the โ€œcoreโ€ of his work โ€“ solving the most difficult part of a math problem โ€“ still relies on pen and paper, AI has dramatically sped up โ€œauxiliary tasks.โ€ Things like generating code, creating complex plots, conducting deeper literature searches, or even reformatting parentheses in a paper now take minutes instead of hours. This allows him to enrich his papers with more data, visuals, and context. โ€œThe type of papers that I would write today, if I had to do them without AI assistance, would definitely take five times longer.โ€ However, he also notes that if he were to write a paper of the same functional depth as one from 2020, without the added features, the time savings arenโ€™t as dramatic. AI makes papers โ€œricher and broader, but not necessarily deeper.โ€

Tao believes AI will revolutionize the โ€œexperimental side of math.โ€ Unlike other sciences with their theoretical and experimental divisions, math has been almost entirely theoretical. Now, with AI, mathematicians can run large-scale experiments to test what methods work and what donโ€™t, gather data on problem-solving effectiveness, and explore mathematics โ€œat scaleโ€ โ€“ an idea still in its infancy.

Artificial Cleverness vs. Artificial Intelligence

A crucial distinction Tao makes is between โ€œartificial clevernessโ€ and โ€œartificial intelligence.โ€ For Tao, true intelligence in a collaborative problem-solving context involves โ€œadaptivity and continual improvement of the idea over time.โ€ Itโ€™s about systematically mapping out what works and what doesnโ€™t, evolving the strategy through discussion.

Current AIs, he says, โ€œcan mimic this a little bit.โ€ They can โ€œjump and fail, and jump and fail.โ€ But they canโ€™t โ€œjump a little bit, reach some handhold, stay there, pull other people up, and then try to jump from there.โ€ There isnโ€™t a โ€œcumulative process which is built up interactively.โ€ Instead, itโ€™s largely โ€œtrial and error and just repetition: brute force.โ€ When you run a new session with an AI, it has โ€œforgotten what it just did.โ€ Its โ€œunderstanding of math has not progressed.โ€

This highlights the current frontier: while AI can apply existing techniques with impressive speed and accuracy, and even combine obscure methods to solve previously intractable problems (especially those lacking extensive literature), it struggles with true novelty and cumulative, evolving understanding. The โ€œholes in the argument where none of the things are workingโ€ still require human ingenuity.

The progress, Tao concludes, is โ€œsimultaneously amazing and disappointing.โ€ While we quickly acclimatize to AIโ€™s stunning capabilities (like Google search two decades ago, or current college-level math-solving), the core challenge of fostering genuine, evolving intelligence remains. The future of science, as envisioned by Terence Tao, will be one where human depth and artificial breadth converge, leading to an โ€œunrecognizableโ€ landscape of discovery โ€“ a new era defined by an unprecedented partnership between human and machine.


Based on โ€œInside Palantir: Building Software That Matters | Shyam Sankar on a16zโ€ from a16z Watch the original video

Americaโ€™s Clarion Call: Reclaiming Innovation and Mobilizing for a New Era

In a world increasingly fraught with geopolitical tension and the specter of โ€œhorrendous barbarism,โ€ a profound shift is underway in how America views its national security and technological future. Shyam Sankar, a titan from Palantir known for his behind-the-scenes influence, has stepped into the public spotlight with a stark message: the nation is at a critical juncture, facing a risk of โ€œsuicide, not homicide,โ€ if it fails to mobilize its latent strengths. His insights, shared on a16z, offer a powerful diagnosis of Americaโ€™s current challenges and a vigorous prescription for reclaiming its innovative edge.

From Shadow Architect to Public Provocateur

For years, Shyam Sankar was the โ€œOG fixerโ€ at Palantir, a figure whispered about in Silicon Valley as the architect behind countless careers and a quiet force in defense tech. Founders like Trey Stevens of Anduril credit Sankar with single-handedly shaping their trajectories, introducing them to Palantir, nurturing their growth, and even โ€œgiving them wings to fly awayโ€ฆ to start something new.โ€ Yet, his influence remained largely out of the public eye until a few years ago.

What prompted this shift? Sankar describes it as โ€œequal parts an act of desperation and act of optimism.โ€ After years witnessing the โ€œfrog boilโ€ of stagnation within the Pentagon, a series of geopolitical eventsโ€”from Russiaโ€™s annexation of Crimea and Chinaโ€™s militarization of the Spratly Islands to the failure of the JCPOA and the October 7th attacks in Israelโ€”catalyzed his conviction. โ€œIt was kind of a radicalizing moment,โ€ he recounts. โ€œWhat is going on here? We need to act.โ€

Simultaneously, Sankar observed a re-emergence of entrepreneurial energy outside the building, with founders eager to build in the national interest. This confluence of urgent need and renewed spirit compelled him to articulate a โ€œfundamental diagnosisโ€: America had โ€œaccidentally turned our back onโ€ the very things that led to past victories. His seminal โ€œFirst Breakfastโ€ piece on defense reformation marked his public debut as a โ€œstrident voice for what needs to happen in America.โ€

The Erosion of Americaโ€™s Defense Edge: A Historical Reckoning

Sankar argues that Americaโ€™s current predicament stems from a series of post-Cold War missteps. Following the Soviet Unionโ€™s collapse, the U.S. embraced a โ€œpeace dividend,โ€ leading to a dramatic restructuring of its defense industrial base. The infamous โ€œLast Supperโ€ dinner saw the number of prime defense contractors shrink from 51 to a mere five. The conventional wisdom suggested this consolidation simply reduced competition. Sankar offers a more nuanced, and troubling, explanation: โ€œconsolidation bred conformity.โ€

This conformity ushered in the โ€œfinancialization of defense,โ€ where companies prioritized financial metrics like dividends and buybacks over innovation and growth. This environment, Sankar asserts, became hostile to โ€œfoundersโ€ โ€“ the daring, often heretical individuals who had driven Americaโ€™s greatest military advancements.

He paints a vivid picture of these historical โ€œhereticsโ€:

  • Heimman Rickover: The โ€œfather of the nuclear navy,โ€ who battled against the Navyโ€™s will and even Oppenheimerโ€™s skepticism to build a revolutionary fleet.
  • John Boyd: The notoriously difficult fighter pilot whose โ€œheresyโ€ regarding energy-maneuverability theory and the OODA loop proved instrumental in the F-16โ€™s design and the swift victory in Gulf War I.
  • Andrew Higgins: The innovative Scots-Irishman whose โ€œHiggins boatโ€ โ€“ initially rejected by the Navy โ€“ ultimately constituted 92% of all landing craft in World War II, enabling critical operations like D-Day.

These figures, Sankar emphasizes, were often โ€œagainst the institution, the bureaucracy, the process.โ€ But in the post-Cold War era, such personalities were โ€œexpungedโ€ and sought refuge in other parts of the American economy, particularly tech.

Compounding this issue was the unique nature of the Department of Defense as a โ€œmonopsonyโ€ โ€“ a single buyer. This allowed the DoD to impose rigid constraints on its suppliers, effectively creating โ€œGalapagos tortoisesโ€: exquisite, highly specialized companies perfectly adapted to the DoDโ€™s unique ecosystem, but uncompetitive on the โ€œmainlandโ€ of the broader economy. For years, there was โ€œno front doorโ€ for outsiders to contribute, save for the intelligence communityโ€™s In-Q-Tel.

Reclaiming the โ€œAmerican Wayโ€: A Vision for Mobilization

Sankarโ€™s book, โ€œMobilize,โ€ champions a return to a more integrated โ€œAmerican industrial base.โ€ He points to World War II, where companies like Chrysler built both Minuteman missiles and minivans, and every consumer purchase subtly subsidized national security. This contrasts sharply with today, where 86% of major weapon system spending goes to defense specialists, up from just 6% in 1989.

The solution, Sankar argues, lies in:

  1. Inspiring Latent Heretics: The most crucial step is to empower and protect the founders and unconventional thinkers, both inside and outside government, who are willing to challenge the status quo. Leaders must โ€œset the conditions to empower the heretics.โ€
  2. Voluntary Civil-Military Fusion: While China enforces civil-military fusion, the U.S. should make voluntary collaboration irresistible. This means leveraging the vast R&D spending of the private sector and re-establishing pathways for skilled individuals to serve, much like World War IIโ€™s direct commissioning of 100,000 experts.
  3. Founders Everywhere: Recognizing that founders arenโ€™t just outside government. Figures like Colonel Drew Cukor, the โ€œfather of Maven,โ€ exemplify this. Cukor, a Marine Colonel, driven by the operational failure that led to the Yazidi genocide, spearheaded the integration of AI into the Pentagon against immense bureaucratic resistance, facing absurd accusations and investigations. His story is a testament to the โ€œincorruptibleโ€ commitment required of true heretics.

AI as the Slingshot for Re-industrialization

Central to Sankarโ€™s vision is the transformative power of Artificial Intelligence. He rejects the โ€œAI dumerismโ€ that portrays AI as an uncontrollable force. Instead, he asserts, โ€œHumans are going to use AI to do X. Thereโ€™s a choice here.โ€ AI, he believes, offers a historic opportunity to rectify the breakdown between wage growth and GDP growth that began in the 1970s.

โ€œThere is an opportunity to give the American worker superpowers with AI,โ€ Sankar declares. This isnโ€™t about symmetrical competition with China, but about a โ€œDavidโ€™s slingshotโ€ โ€“ using technology to re-industrialize the country in entirely new ways. He highlights companies like Hadrian, which are achieving 50-100x productivity gains through tech-driven manufacturing.

This re-industrialization, Sankar argues, necessitates a return to the โ€œcollocation of R&D and production.โ€ The โ€œgreat lie of globalizationโ€ was the separation of innovation from production. As SpaceX demonstrates, innovation thrives when R&D engineers are on the factory floor, enabling rapid feedback loops. If you donโ€™t make the thing, you canโ€™t innovate on how you make the thing.

Inside the Army: A New Kind of Mobilization

Sankar isnโ€™t just advocating; heโ€™s acting. He recently joined the Army, along with other tech luminaries like Bob Muglia (former OpenAI chief research officer), Boz (Meta CTO), and Andrew Weil (former OpenAI chief product officer). This initiative, under General George and Secretary Driscoll, aims to bring senior tech expertise directly into the military.

His focus includes long-term force structure planning and treating software as a โ€œmalleable weapon system.โ€ What has surprised him most from the inside? The โ€œquality of talent in our green suitorsโ€ โ€“ young, often self-taught individuals building โ€œthe most compelling AI applicationsโ€ driven by โ€œexistential stakes.โ€ These junior personnel, empowered by AI tools, can now build and deploy ideas in weeks, fostering a โ€œbottoms-up innovation mission commandโ€ that plays to Americaโ€™s unique military strengths.

The SAS Apocalypse and the Future of the Economy

Sankar also offers a provocative take on the โ€œSAS apocalypse,โ€ the idea that AI will commoditize much of existing software. He introduces a critical distinction:

  • Beta Software: Software that makes you similar to everyone else. This, he predicts, will โ€œreally struggle,โ€ as AI allows custom โ€œvibe-codingโ€ of solutions.
  • Alpha Software: Software that allows you to express your differentiation and competitive advantage. Platforms focused on alpha will thrive, becoming โ€œtoolkitsโ€ that enable unique strategies.

He argues that much of the โ€œsoftware industrial complexโ€ has historically focused on โ€œcan I sell it?โ€ rather than โ€œdid it add value?โ€ The COVID-19 pandemic, where multi-billion dollar ERP systems collapsed while Zoom and Teams enabled remote work, served as a โ€œSputnik momentโ€ for the industry.

Looking at the AI stack, Sankar believes value will accrue primarily at two layers: the โ€œchips layerโ€ and the โ€œAI infrastructure layerโ€ (what Palantir calls โ€œontologyโ€). Models, he suggests, are increasingly commoditized.

Ultimately, Sankarโ€™s vision for AIโ€™s economic impact is optimistic, provided humans exercise their agency. AI can reverse the trend of financial engineering over real engineering, empower workers, and drive re-industrialization. He champions a return to a โ€œfounder personalityโ€ that prioritizes building and engineering, where the pathway to CEO runs through the CTO, as Elon Musk suggests. The goal, he concludes, isnโ€™t to replace people but to make them better, to build โ€œIron Man suitsโ€ for every worker, ensuring America remains dominant not just in technology, but in its very future.

Shyam Sankarโ€™s voice is a powerful reminder that the future is not predetermined. Itโ€™s a choice, shaped by leadership, courage, and a renewed commitment to the innovative spirit that once defined America. The call to mobilize is clear, and the opportunity to reclaim that vigor is now.


Based on โ€œSilicon Valleyโ€™s Big Bets on War Pay Off, and the Trump Family Business Looks to Transylvaniaโ€ from New York Times Podcasts Watch the original video

The Shifting Sands of Power and Profit: From Silicon Valleyโ€™s War Bets to Trumpโ€™s Transylvanian Gambit

In a world grappling with escalating conflicts, rapid technological shifts, and persistent economic anxieties, seemingly disparate headlines often converge to paint a vivid picture of our changing times. From tech giants pivoting to defense to a former presidentโ€™s family business eyeing an unlikely real estate venture, and from the quiet desperation driving plasma donations to the unsettling rise of AI in literature, these stories reveal the complex currents beneath the surface of daily news.

Silicon Valleyโ€™s New Front: The Business of War

The escalating tensions surrounding Iran have sent ripples across global markets and geopolitical strategies. While President Trump has publicly stated heโ€™s โ€œnot putting troops anywhereโ€ currently, he also hasnโ€™t ruled out deploying forces โ€œif needed.โ€ In a dramatic reversal of long-standing US policy, the administration is even considering unsanctioning Iranian oil โ€” a desperate move to bring down prices, effectively encouraging Iran to sell more oil even amidst conflict.

At the Pentagon, the financial stakes are soaring. The Defense Department is seeking an astonishing $200 billion to continue funding the war effort. This colossal sum, nearly a quarter of the entire annual defense budget, is already raising eyebrows among moderate Republicans. Yet, it also signals an unprecedented opportunity for a specific, once-reluctant sector: Silicon Valley.

For years, the tech industry, particularly companies like Google, Meta, and OpenAI, shied away from defense contracts. Google employees famously protested their companyโ€™s work with the Department of Defense, invoking the โ€œdo no evilโ€ motto as a moral compass against engaging in the business of war. But in a striking shift, those optics have dramatically changed.

Now, mammoth tech companies and nimble startups alike are actively pivoting into defense technology, striking deals worth hundreds of millions, even billions, of dollars. Theyโ€™re building everything from advanced weapon systems to sophisticated software for the US government. Palantir, a data analytics company, has developed Project Maven, a system that aids the US government in selecting targets for air strikes. Google and OpenAI are leveraging their AI prowess to assist generals in the field. Even former Google CEO Eric Schmidt is leading projects to build counter-drone systems, now actively deployed by US assets in Iran to defend against Iranian drones.

With President Trump having allocated over a trillion dollars in defense spending, much of it is earmarked for the very kind of cutting-edge defense technology Silicon Valley is eager to produce. As global conflicts intensify, the once-unpopular โ€œbusiness of warโ€ has become a lucrative frontier for the tech world, demonstrating a profound re-evaluation of ethical boundaries in pursuit of profit and strategic influence. This shift is further underscored by reports that Ukraine, having gained years of experience fighting off Russian drones (many of which were Iranian-made), has become a crucial source of expertise for other Middle Eastern nations seeking to defend against similar threats.

Trumpโ€™s Transylvanian Ambition: Beyond the Glamour

Beyond the battlefields and boardrooms, another intriguing story unfolds in the unlikely setting of Transylvania, Romania. A new investigation by The New York Times has unearthed details of an unannounced real estate deal pursued by President Trumpโ€™s family business. Unlike the glamorous tourist destinations like Bali and the Maldives where the Trump Organization typically operates, this project is slated for a semi-abandoned site alongside a military base and several enormous landfills.

Rebecca Ruiz, an investigative reporter for The Times who visited the site, described a landscape far from luxury: โ€œYou could smell the stench. There were packs of wild dogs wandering. It felt like an unusual place to envision a luxury apartment complex and golf course.โ€

The choice of location is perplexing to many, but it aligns with a pattern observed in Trumpโ€™s second term: the Trump Organization actively seeks deals in places where the former president enjoys significant popularity. A Gallup poll last year revealed that over half of Romania held a favorable view of Trumpโ€™s performance as president, making the country an outlier among EU nations. This unannounced Transylvanian venture, with its stark contrasts and strategic location, highlights the unique blend of politics and commerce that defines the Trump brandโ€™s global real estate ambitions.

Americaโ€™s Quiet Struggle: The Rise of the Plasma Economy

Closer to home, a different kind of economic reality is playing out in the unassuming middle-class suburbs of cities like Houston, Chicago, and Las Vegas. Businesses where individuals can donate plasma for cash are popping up with increasing frequency, shifting from their traditional clusters in low-income neighborhoods. Now, these centers are found nestled beside Orange Theory Fitness gyms and Charles Schwab offices.

Reporters from The Times visited several new Texas locations, finding long lines of people who never imagined theyโ€™d be selling their plasma: a tech worker in his 30s, a sixth-grade special education teacher, a night-shift nurse. Most reported visiting the clinic twice a week, the maximum allowed under FDA regulations, to earn about $70 per donation.

This phenomenon underscores the quiet financial strain many middle-income Americans are experiencing, where an extra $70 can make a tangible difference for groceries or gas. One industry researcher described plasma centers as a โ€œshadow safety net,โ€ offering a side income much like driving for Uber or Lyft.

The United States is a global anomaly in this regard, providing roughly 70% of the worldโ€™s plasma because itโ€™s one of the few countries that allows payment for donations โ€“ a practice the World Health Organization discourages. While donating plasma is generally considered safe, there is remarkably little research on the long-term effects of frequent donations, adding a layer of concern to this multi-billion dollar industry built on the financial needs of its citizens.

The AI Author: Publishingโ€™s New Frontier

Finally, the literary world is grappling with its own disruption: the rise of artificial intelligence. A buzzy new horror novel, โ€œShy Girl,โ€ was set for release this spring, a revenge story about a woman held hostage. Its UK edition had been out for months, but the US publisher, Hachette, abruptly dropped the book over claims it was written with the help of AI.

This marks what appears to be the first commercial novel from a major publishing house to be pulled due to evidence of AI use. Early readers had voiced suspicions online, pointing to โ€œnonsensical metaphors and repetitive bits.โ€ A publishing industry consultant, who ran the book through three different AI detection programs, confirmed that all three found text โ€œlikely to be at least partly generated by AI,โ€ highlighting โ€œcertain odd phrases like, โ€˜I pressed the phone to my lips. The screen cool and unyielding.โ€™โ€

The author has denied using AI herself, instead claiming that someone she hired to help edit the story was responsible. This incident has laid bare the publishing industryโ€™s unpreparedness for this burgeoning technology. While most publishing contracts require authors to affirm their work is original, few companies have established measures or safeguards to verify it. As one consultant grimly noted to The Times, โ€œAI bleeding into books is not merely inevitable. Weโ€™re in the midst of it.โ€

These diverse narratives, from the ethical dilemmas of tech and war to the surprising ventures of political figures, the quiet struggles of everyday Americans, and the disruptive force of AI, collectively illuminate the complex and rapidly evolving landscape of our modern world. They are the headlines that, when examined together, offer a deeper understanding of the forces shaping our present and future.


ํ•œ๊ตญ์–ด

โ€œThe rise and fall of civilizations | Eric Cline: Full Interviewโ€ โ€” Big Think ๊ธฐ๋ฐ˜ ๊ธฐ์‚ฌ ์›๋ณธ ์˜์ƒ ๋ณด๊ธฐ

๊ณ ๋Œ€ ๋ฌธ๋ช…์˜ โ€˜ํผํŽ™ํŠธ ์Šคํ†ฐโ€™: 3์ฒœ ๋…„ ์ „ ๋ถ•๊ดด๊ฐ€ ์˜ค๋Š˜๋‚  ์šฐ๋ฆฌ์—๊ฒŒ ๋˜์ง€๋Š” ๊ฒฝ๊ณ 

๊ธฐ์›์ „ 1200๋…„๊ฒฝ, ์ง€์ค‘ํ•ด์™€ ๊ทผ๋™ ์ง€์—ญ์€ ์—ญ์‚ฌ์ƒ ์œ ๋ก€์—†๋Š” ๋Œ€๊ฒฉ๋ณ€์„ ๊ฒช์—ˆ๋‹ค. ์ˆ˜๋ฐฑ ๋…„๊ฐ„ ๋ฒˆ์„ฑํ–ˆ๋˜ ๊ฐ•๋ ฅํ•œ ๋ฌธ๋ช…๋“ค์ด ๋งˆ์น˜ ๋„๋ฏธ๋…ธ์ฒ˜๋Ÿผ ์—ฐ์‡„์ ์œผ๋กœ ๋ฌด๋„ˆ์ ธ ๋‚ด๋ฆฐ ๊ฒƒ์ด๋‹ค. ์ด ์‹œ๊ธฐ, ๊ณผ์—ฐ ๋ฌด์Šจ ์ผ์ด ์žˆ์—ˆ๋˜ ๊ฑธ๊นŒ? ๊ณ ๊ณ ํ•™์ž์ด์ž ๊ณ ๋Œ€ ์—ญ์‚ฌ๊ฐ€์ธ ์—๋ฆญ ํด๋ผ์ธ(Eric Cline) ๊ต์ˆ˜๋Š” ๊ทธ์˜ ์ €์„œ ใ€ŽBC 1177: ๋ฌธ๋ช…์ด ๋ถ•๊ดดํ•œ ํ•ด(1177 BC: The Year Civilization Collapsed)ใ€์™€ ํ›„์†์ž‘ ใ€ŽBC 1177 ์ดํ›„(After 1177 BC)ใ€๋ฅผ ํ†ตํ•ด ์ด ์ˆ˜์ˆ˜๊ป˜๋ผ ๊ฐ™์€ ๋ถ•๊ดด์˜ ์›์ธ์„ ํŒŒํ—ค์น˜๊ณ , ์˜ค๋Š˜๋‚  ์šฐ๋ฆฌ๊ฐ€ ์ง๋ฉดํ•œ ๋ฌธ์ œ์— ๋Œ€ํ•œ ์‹ฌ์˜คํ•œ ํ†ต์ฐฐ์„ ์ œ์‹œํ•œ๋‹ค.

ํด๋ผ์ธ ๊ต์ˆ˜๋Š” ํ›„๊ธฐ ์ฒญ๋™๊ธฐ ์‹œ๋Œ€(Late Bronze Age, ์•ฝ ๊ธฐ์›์ „ 1700๋…„ ~ ๊ธฐ์›์ „ 1200๋…„)๋ฅผ ์ธ๋ฅ˜ ์—ญ์‚ฌ์ƒ ๊ฐ€์žฅ ์ค‘์š”ํ•œ ์‹œ๊ธฐ ์ค‘ ํ•˜๋‚˜๋กœ ๊ผฝ๋Š”๋‹ค. โ€œ์ด ์‹œ๊ธฐ๋Š” ์‚ฌ๋žŒ๋“ค์ด ์ง€์ค‘ํ•ด ์ „์—ญ์— ๊ฑธ์ณ ์ „๋ก€ ์—†๋Š” ๋ฐฉ์‹์œผ๋กœ โ€˜์„ธ๊ณ„ํ™”โ€™๋˜์–ด ์žˆ์—ˆ๋‹คโ€๋Š” ๊ทธ์˜ ์„ค๋ช…์ฒ˜๋Ÿผ, ๋‹น์‹œ์˜ ์ƒํ˜ธ ์—ฐ๊ฒฐ์„ฑ์€ ํ˜„๋Œ€์™€ ๋†€๋ž๋„๋ก ๋‹ฎ์•„ ์žˆ๋‹ค. 3์ฒœ ๋…„ ์ „์˜ ์‚ฌ๊ฑด์ด ์˜ค๋Š˜๋‚  ์šฐ๋ฆฌ์—๊ฒŒ ์‹œ์‚ฌํ•˜๋Š” ๋ฐ”๊ฐ€ ํฌ๋‹ค๋Š” ์ฃผ์žฅ์€ ๋ฐ”๋กœ ์ด๋Ÿฌํ•œ ์œ ์‚ฌ์„ฑ์—์„œ ์ถœ๋ฐœํ•œ๋‹ค.

์‚ฌ๋ผ์ง„ ํ™ฉ๊ธˆ๊ธฐ: ํ›„๊ธฐ ์ฒญ๋™๊ธฐ ์‹œ๋Œ€์˜ ๋ฒˆ์˜

ํด๋ผ์ธ ๊ต์ˆ˜๊ฐ€ ์ง€์นญํ•˜๋Š” โ€˜๊ณ ๋Œ€ G8โ€™์€ ํ›„๊ธฐ ์ฒญ๋™๊ธฐ ์‹œ๋Œ€ ์ง€์ค‘ํ•ด์™€ ๊ทผ๋™์„ ์ง€๋ฐฐํ–ˆ๋˜ ์ฃผ์š” ๋ฌธ๋ช…๋“ค์„ ์ผ์ปซ๋Š”๋‹ค. ๊ทธ๋ฆฌ์Šค์˜ ๋ฏธ์ผ€๋„ค์ธ(Mycenaeans)๊ณผ ๋ฏธ๋…ธ์•„์ธ(Minoans) (ํŠธ๋กœ์ด ์ „์Ÿ์˜ ๋ฐฐ๊ฒฝ), ์•„๋‚˜ํ†จ๋ฆฌ์•„(ํ˜„ ํŠ€๋ฅดํ‚ค์˜ˆ)์˜ ํžˆํƒ€์ดํŠธ(Hittites), ๋ฉ”์†Œํฌํƒ€๋ฏธ์•„(ํ˜„ ์ด๋ผํฌ)์˜ ์•„์‹œ๋ฆฌ์•„(Assyria)์™€ ๋ฐ”๋นŒ๋กœ๋‹ˆ์•„(Babylonia), ์ด์ง‘ํŠธ(Egypt), ํ‚คํ”„๋กœ์Šค(Cyprus), ๊ทธ๋ฆฌ๊ณ  ๊ฐ€๋‚˜์•ˆ(Canaan)์˜ ์‚ฌ๋žŒ๋“ค์ด ๊ทธ๋“ค์ด๋‹ค. ์ด๋“ค์€ ๋‹จ์ˆœํ•œ ์ด์›ƒ์„ ๋„˜์–ด, ๊ธด๋ฐ€ํ•œ ์ƒ์—…์ , ์™ธ๊ต์ , ์‹ฌ์ง€์–ด ํ˜ผ์ธ ๊ด€๊ณ„๋ฅผ ํ†ตํ•ด ํ•˜๋‚˜์˜ ๊ฑฐ๋Œ€ํ•œ โ€˜์†Œ๊ทœ๋ชจ ์„ธ๊ณ„ ๋„คํŠธ์›Œํฌ(Small World Network)โ€˜๋ฅผ ํ˜•์„ฑํ•˜๊ณ  ์žˆ์—ˆ๋‹ค.

โ€œ์„œ๋กœ์—๊ฒŒ ์„ธ ๊ฑธ์Œ ์ด๋‚ด์— ์žˆ๋Š” ๊ด€๊ณ„โ€๋ผ๊ณ  ํด๋ผ์ธ ๊ต์ˆ˜๊ฐ€ ์„ค๋ช…ํ•˜๋Š” ์ด ๋„คํŠธ์›Œํฌ๋Š”, ์ƒ์—…๊ณผ ์™ธ๊ต์˜ ํ™œ๋ฐœํ•œ ๊ต๋ฅ˜๋ฅผ ํ†ตํ•ด ์œ ์ง€๋˜์—ˆ๋‹ค. ์ด์ง‘ํŠธ๋Š” ๊ธˆ์„, ๊ทธ๋ฆฌ์Šค๋Š” ์€์„, ํ‚คํ”„๋กœ์Šค๋Š” ๊ตฌ๋ฆฌ๋ฅผ ๊ณต๊ธ‰ํ–ˆ์œผ๋ฉฐ, ์ฃผ์„์€ ๋ฉ€๋ฆฌ ์•„ํ”„๊ฐ€๋‹ˆ์Šคํƒ„(๋ฐ”๋‹คํ์ƒจ ์ง€์—ญ)์—์„œ ์ˆ˜์ฒœ ํ‚ฌ๋กœ๋ฏธํ„ฐ๋ฅผ ์ด๋™ํ•ด ์™”๋‹ค. ์ด๋“ค์€ ๋‹จ์ˆœํžˆ ์‚ฌ์น˜ํ’ˆ์„ ๋„˜์–ด, ์ฒญ๋™๊ธฐ ์‹œ๋Œ€์˜ ํ•ต์‹ฌ ์ž์›์ธ ์ฒญ๋™(๊ตฌ๋ฆฌ 90%, ์ฃผ์„ 10%)์„ ๋งŒ๋“ค๊ธฐ ์œ„ํ•ด ์„œ๋กœ์˜ ์ž์›์— ์˜์กดํ•  ์ˆ˜๋ฐ–์— ์—†์—ˆ๋‹ค. ์˜ฌ๋ฆฌ๋ธŒ์œ , ์™€์ธ, ๊ณก๋ฌผ ๋“ฑ ์‹๋Ÿ‰ ์ž์›๋„ ํ™œ๋ฐœํžˆ ๊ต์—ญ๋˜์—ˆ์œผ๋ฉฐ, ์‹ฌ์ง€์–ด ๊ฐ€์ฃฝ ์‹ ๋ฐœ์ด๋‚˜ ๊ธˆ ์ƒ๊ฐ ๋‹จ๊ฒ€ ๊ฐ™์€ ์‹ค์ œ ๋ฌผํ’ˆ๋“ค๋„ ์™•๋“ค ์‚ฌ์ด์˜ โ€˜์„ ๋ฌผ ๊ตํ™˜โ€™์ด๋‚˜ ์ƒ์—… ํ™œ๋™์„ ํ†ตํ•ด ์˜ค๊ณ  ๊ฐ”๋‹ค.

์ด๋Ÿฌํ•œ ์ƒํ˜ธ ์˜์กด์„ฑ์€ ๋ฌธ๋ช…๋“ค์„ ์ „๋ก€ ์—†๋Š” ๋ฒˆ์˜์œผ๋กœ ์ด๋Œ์—ˆ์ง€๋งŒ, ๋™์‹œ์— ์น˜๋ช…์ ์ธ ์•ฝ์ ์ด ๋˜์—ˆ๋‹ค. โ€œ์–ด๋А ๋ˆ„๊ตฌ๋„ ์ž๊ธ‰์ž์กฑํ•  ์ˆ˜ ์—†์—ˆ๋‹คโ€๋Š” ํด๋ผ์ธ ๊ต์ˆ˜์˜ ์ง€์ ์ฒ˜๋Ÿผ, ์ด ์—ฐ๊ฒฐ๋ง์ด ๋Š์–ด์ง€๋Š” ์ˆœ๊ฐ„, ๋ชจ๋“  ๊ฒƒ์ด ๋ฌด๋„ˆ์งˆ ์ˆ˜๋ฐ–์— ์—†์—ˆ๋‹ค. ์™ธ๊ต์  ๊ด€๊ณ„ ์—ญ์‹œ ์ƒํ˜ธ ํ˜‘๋ ฅ์˜ ์ค‘์š”ํ•œ ์ถ•์ด์—ˆ๋‹ค. ๋Œ€์™•๋“ค์€ ์กฐ์•ฝ์„ ๋งบ๊ณ  ์ด๋ฅผ ๊ณต๊ณ ํžˆ ํ•˜๊ธฐ ์œ„ํ•ด ์ž๋…€๋“ค์„ ์„œ๋กœ์˜ ํ•˜๋ ˜์œผ๋กœ ๋ณด๋ƒˆ๋‹ค. ์ด์ง‘ํŠธ ํŒŒ๋ผ์˜ค ์•„๋ฉ˜ํ˜ธํ…Œํ”„ 3์„ธ์™€ ์•„์ผ€๋‚˜ํ†ค์˜ ํ•˜๋ ˜์—๋Š” ๋ฏธํƒ„๋‹ˆ์™€ ๋ฐ”๋นŒ๋กœ๋‹ˆ์•„ ๊ณต์ฃผ๋“ค์ด ์žˆ์—ˆ์œผ๋ฉฐ, ์ด๋Š” ๋‹น์‹œ ์™ธ๊ต์˜ ์ค‘์š”ํ•œ ์ˆ˜๋‹จ์ด์—ˆ๋‹ค.

๋ถ•๊ดด์˜ ์„œ๋ง‰: ๋‹จ์ˆœํ•จ์„ ๋„˜์–ด์„  ๋ณตํ•ฉ์  ์›์ธ

์˜ค๋žซ๋™์•ˆ ํ•™์ž๋“ค์€ ํ›„๊ธฐ ์ฒญ๋™๊ธฐ ์‹œ๋Œ€์˜ ๋ถ•๊ดด๋ฅผ โ€˜๋ฐ”๋‹ค ๋ฏผ์กฑ(Sea Peoples)โ€˜์˜ ์นจ๋žต๊ณผ ๊ฐ™์€ ๋‹จ์ผ ์›์ธ(monocausal)์œผ๋กœ ์„ค๋ช…ํ•˜๋ ค๋Š” ๊ฒฝํ–ฅ์ด ์žˆ์—ˆ๋‹ค. ๊ทธ๋Ÿฌ๋‚˜ ํด๋ผ์ธ ๊ต์ˆ˜๋Š” ์ด๋Ÿฌํ•œ ์„ค๋ช…์ด โ€œ๋„ˆ๋ฌด ๋‹จ์ˆœํ•˜๋‹คโ€๊ณ  ์ง€์ ํ•˜๋ฉฐ, ์ด์ œ๋Š” ๋ณตํ•ฉ์  ์›์ธ(polycausal)์— ์ฃผ๋ชฉํ•ด์•ผ ํ•œ๋‹ค๊ณ  ๊ฐ•์กฐํ•œ๋‹ค. โ€œ๋‘์„ธ ๊ฐ€์ง€, ํ˜น์€ ๋„ค ๊ฐ€์ง€ ์š”์ธ์ด ๋™์‹œ์— ๋˜๋Š” ๋น ๋ฅด๊ฒŒ ์—ฐ์ด์–ด ๋ฐœ์ƒํ•˜์—ฌ, ํ•œ ์žฌ์•™์—์„œ ํšŒ๋ณตํ•  ํ‹ˆ๋„ ์—†์ด ๋‹ค์Œ ์žฌ์•™์ด ๋‹ฅ์ณ์™”๋‹คโ€๋Š” ๊ฒƒ์ด๋‹ค. ๊ทธ๋Š” ์ด๋ฅผ โ€˜ํผํŽ™ํŠธ ์Šคํ†ฐ(perfect storm)โ€˜์— ๋น„์œ ํ•˜๋ฉฐ, ๊ฐ ์š”์ธ์ด ์„œ๋กœ๋ฅผ ์ฆํญ์‹œํ‚ค๋Š” โ€˜์Šน์ˆ˜ ํšจ๊ณผ(multiplier effect)โ€˜์™€ โ€˜๋„๋ฏธ๋…ธ ํšจ๊ณผ(domino effect)โ€˜๋ฅผ ์ผ์œผ์ผฐ๋‹ค๊ณ  ์„ค๋ช…ํ•œ๋‹ค.

์ด๋Ÿฌํ•œ ์ ‘๊ทผ ๋ฐฉ์‹์€ ์ง€๋‚œ ์ˆ˜์‹ญ ๋…„๊ฐ„ ์ƒˆ๋กœ์šด ๋ฐ์ดํ„ฐ์™€ ์—ฐ๊ตฌ ๊ฒฐ๊ณผ๊ฐ€ ์ถ•์ ๋˜๋ฉด์„œ ๊ฐ€๋Šฅํ•ด์กŒ๋‹ค. โ€œ์—ญ์‚ฌ๋Š” ๋ฐ˜๋ณต๋˜์ง€ ์•Š์ง€๋งŒ, ์šด์œจ์€ ์žˆ๋‹ค(history does rhyme even if it doesnโ€™t repeat)โ€œ๋Š” ํด๋ผ์ธ ๊ต์ˆ˜์˜ ๋ง์ฒ˜๋Ÿผ, ๊ณ ๋Œ€ ์‚ฌํšŒ์˜ ๋ถ•๊ดด ๊ณผ์ •์€ ์˜ค๋Š˜๋‚  ์šฐ๋ฆฌ๊ฐ€ ์ง๋ฉดํ•œ ๊ธฐํ›„ ๋ณ€ํ™”, ์ด์ฃผ ๋ฌธ์ œ, ํŒฌ๋ฐ๋ฏน ๋“ฑ์˜ ๋ณตํ•ฉ ์œ„๊ธฐ(poly-crisis)์— ๋Œ€ํ•œ ์ค‘์š”ํ•œ ๊ฒฝ๊ณ ์Œ์„ ๋˜์ง„๋‹ค.

๋Œ€๊ฒฉ๋ณ€์„ ๋ถ€๋ฅธ ์š”์ธ๋“ค: ํผํŽ™ํŠธ ์Šคํ†ฐ์˜ ๊ตฌ์„ฑ ์š”์†Œ

ํด๋ผ์ธ ๊ต์ˆ˜๋Š” ํ›„๊ธฐ ์ฒญ๋™๊ธฐ ์‹œ๋Œ€ ๋ถ•๊ดด๋ฅผ ์ดˆ๋ž˜ํ•œ ์ฃผ์š” ์š”์ธ๋“ค์„ ๋‹ค์Œ๊ณผ ๊ฐ™์ด ์ œ์‹œํ•œ๋‹ค.

  1. ๊ฐ€๋ญ„๊ณผ ๊ธฐ๊ทผ(Drought and Famine): ๊ฐ€๋ญ„์€ ๋‹จ์ˆœํ•œ ๊ธฐํ›„ ํ˜„์ƒ์„ ๋„˜์–ด ์‚ฌํšŒ ์ „์ฒด๋ฅผ ๋’คํ”๋“œ๋Š” ์žฌ์•™์ด์—ˆ๋‹ค. ๋ฆฌ์Šค ์นดํŽœํ„ฐ(Rhys Carpenter) ๊ต์ˆ˜๊ฐ€ 1960๋…„๋Œ€์— ๋ฏธ์ผ€๋„ค ๋ฌธ๋ช… ๋ถ•๊ดด์˜ ์›์ธ์œผ๋กœ ๊ฐ€๋ญ„์„ ์ œ์•ˆํ–ˆ์ง€๋งŒ, ๋‹น์‹œ์—๋Š” ์ด๋ฅผ ๋’ท๋ฐ›์นจํ•  ์ž๋ฃŒ๊ฐ€ ๋ถ€์กฑํ–ˆ๋‹ค. ๊ทธ๋Ÿฌ๋‚˜ ์˜ค๋Š˜๋‚  ์šฐ๋ฆฌ๋Š” ์ดํƒˆ๋ฆฌ์•„์—์„œ ์ด๋ž€์— ์ด๋ฅด๋Š” ๊ด‘๋ฒ”์œ„ํ•œ ์ง€์—ญ์—์„œ โ€˜๋ฉ”๊ฐ€ ๊ฐ€๋ญ„(mega-drought)โ€˜์˜ ์ฆ๊ฑฐ๋ฅผ ํ™•๋ณดํ•˜๊ณ  ์žˆ๋‹ค. ๋™๊ตด์˜ ์„์ˆœ ์„ฑ์žฅ ์ค‘๋‹จ, ๋งˆ๋ฅธ ํ˜ธ์ˆ˜์™€ ๊ฐ•๋ฐ”๋‹ฅ์˜ ํ‡ด์ ๋ฌผ ๋ถ„์„, ๊ทธ๋ฆฌ๊ณ  ํ˜„๋ฏธ๊ฒฝ์œผ๋กœ ๊ด€์ฐฐํ•œ ๊ฝƒ๊ฐ€๋ฃจ ๋ณ€ํ™” ๋“ฑ ๋‹ค์–‘ํ•œ ๊ณผํ•™์  ๋ฐ์ดํ„ฐ๋Š” ๊ธฐ์›์ „ 1250๋…„๊ฒฝ๋ถ€ํ„ฐ ์ตœ์†Œ 150๋…„, ๊ธธ๊ฒŒ๋Š” 300๋…„๊ฐ„ ์ง€์†๋œ ๊ฐ€ํ˜นํ•œ ๊ฑด์กฐ ๊ธฐํ›„๋ฅผ ์ฆ๋ช…ํ•œ๋‹ค.

    ์ด๋Ÿฌํ•œ ๊ฐ€๋ญ„์€ ํ•„์—ฐ์ ์œผ๋กœ ๊ธฐ๊ทผ์„ ๋ถˆ๋Ÿฌ์™”๋‹ค. ์‹œ๋ฆฌ์•„ ๋ถ๋ถ€ ํ•ด์•ˆ์˜ ์ค‘์š”ํ•œ ๊ตญ์ œ ํ•ญ๊ตฌ์˜€๋˜ ์šฐ๊ฐ€๋ฆฌํŠธ(Ugarit)์—์„œ ๋ฐœ๊ตด๋œ ์ ํ† ํŒ์—๋Š” โ€œ์šฐ๋ฆฌ ๋„์‹œ์— ๊ธฐ๊ทผ์ด ๋“ค์—ˆ๋‹ค. ๋„์™€๋‹ฌ๋ผโ€๋Š” ์ ˆ๋ฐ•ํ•œ ๋ฉ”์‹œ์ง€๊ฐ€ ๋‹ด๊ฒจ ์žˆ๋‹ค. ํžˆํƒ€์ดํŠธ ์™•๊ตญ์—์„œ๋„ ๊ธฐ์›์ „ 1250๋…„๊ฒฝ๋ถ€ํ„ฐ โ€œ๋‚ด ๋•…์— ๊ธฐ๊ทผ์ด ๋“ค์—ˆ๋‹ค. ์ƒ์‚ฌ์˜ ๋ฌธ์ œ๋‹ค. ๊ณก๋ฌผ์„ ๋ณด๋‚ด๋‹ฌ๋ผโ€๋Š” ๋‚ด์šฉ์˜ ์„œ์‹ ์ด ๋ฐœ๊ฒฌ๋˜์—ˆ์œผ๋ฉฐ, ์ด์ง‘ํŠธ๊ฐ€ ์‹ค์ œ๋กœ ๊ณก๋ฌผ๊ณผ ๋ง๋ฆฐ ์ƒ์„ ์„ ๋ณด๋‚ด ๊ตฌํ˜ธ ํ™œ๋™์„ ํŽผ์ณค๋‹ค๋Š” ๊ธฐ๋ก๋„ ์กด์žฌํ•œ๋‹ค.

  2. ์ด์ฃผ์™€ ์นจ๋žต: โ€˜๋ฐ”๋‹ค ๋ฏผ์กฑโ€™์˜ ์žฌํ•ด์„(Migration and Invasion: Reinterpreting the โ€˜Sea Peoplesโ€™): ์ „ํ†ต์ ์œผ๋กœ โ€˜๋ฐ”๋‹ค ๋ฏผ์กฑโ€™์€ ํ›„๊ธฐ ์ฒญ๋™๊ธฐ ์‹œ๋Œ€ ๋ฌธ๋ช… ๋ถ•๊ดด์˜ ์ฃผ๋ฒ”์œผ๋กœ ์ง€๋ชฉ๋˜์–ด ์™”๋‹ค. ์ด์ง‘ํŠธ ํŒŒ๋ผ์˜ค ๋ฉ”๋ฅด๋„ตํƒ€(Merneptah)์™€ ๋žŒ์„ธ์Šค 3์„ธ(Ramses III)์˜ ๊ธฐ๋ก์— ๋”ฐ๋ฅด๋ฉด, ์ด๋“ค์€ ๊ธฐ์›์ „ 1207๋…„๊ณผ 1177๋…„์— ์ด์ง‘ํŠธ๋ฅผ ๊ณต๊ฒฉํ•œ 9๊ฐœ ์—ฐํ•ฉ ์ง‘๋‹จ์ด์—ˆ๋‹ค. ๊ทธ๋Ÿฌ๋‚˜ ํด๋ผ์ธ ๊ต์ˆ˜๋Š” ์ด๋“ค์„ ๋‹จ์ˆœํ•œ ์นจ๋žต์ž๊ฐ€ ์•„๋‹Œ, โ€œํ”ผํ•ด์ž์ด์ž ์–ต์••์žโ€์ด๋ฉฐ โ€œ๋ถ•๊ดด์˜ ์›์ธ์ด๋ผ๊ธฐ๋ณด๋‹ค๋Š” ์ฆ์ƒโ€์œผ๋กœ ํ•ด์„ํ•œ๋‹ค.

    ๋ฉ”๊ฐ€ ๊ฐ€๋ญ„๊ณผ ๊ธฐ๊ทผ์œผ๋กœ ์ธํ•ด ์„œ๋ถ€ ์ง€์ค‘ํ•ด(์‹œ์น ๋ฆฌ์•„, ์‚ฌ๋ฅด๋””๋‹ˆ์•„, ์ดํƒˆ๋ฆฌ์•„ ๋“ฑ)์—์„œ ์‚ด ๊ณณ์„ ์žƒ์€ ์‚ฌ๋žŒ๋“ค์ด ๋™์ชฝ์œผ๋กœ ๋Œ€๊ทœ๋ชจ ์ด์ฃผ๋ฅผ ์‹œ์ž‘ํ–ˆ์„ ๊ฒƒ์ด๋ผ๋Š” ์„ค๋ช…์ด๋‹ค. ์ด๋“ค ์ค‘ โ€˜ํŽ ๋ ˆ์…‹(Peleset)โ€˜์œผ๋กœ ๋ถˆ๋ฆฐ ์ง‘๋‹จ์€ ์„ฑ๊ฒฝ์— ๋‚˜์˜ค๋Š” โ€˜๋ธ”๋ ˆ์…‹์ธ(Philistines)โ€˜์œผ๋กœ, ๊ทธ๋ฆฌ์Šค ๋ฏธ์ผ€๋„ค ๋ฌธ๋ช…๊ถŒ์—์„œ ์˜จ ์‚ฌ๋žŒ๋“ค๋กœ ์ถ”์ •๋œ๋‹ค. ์ด๋“ค์€ ์ƒˆ๋กœ์šด ์‚ถ์„ ์ฐพ์•„ ์ด์ฃผํ–ˆ์ง€๋งŒ, ์ด๋ฏธ ๊ฐ€๋ญ„๊ณผ ๊ธฐ๊ทผ์œผ๋กœ ๊ณ ํ†ต๋ฐ›๋˜ ๋™๋ถ€ ์ง€์ค‘ํ•ด์— ๋„์ฐฉํ•˜์—ฌ ๊ธฐ์กด ์ฃผ๋ฏผ๋“ค๊ณผ ์ถฉ๋Œํ–ˆ๊ณ , ์ด๊ฒƒ์ด ์นจ๋žต์œผ๋กœ ๋น„์ณค์„ ์ˆ˜ ์žˆ๋‹ค๋Š” ๊ฒƒ์ด๋‹ค. ์šฐ๊ฐ€๋ฆฌํŠธ์˜ ์ ํ† ํŒ์— ๊ธฐ๋ก๋œ โ€œ์ ์˜ ๋ฐฐ๋“ค์ด ์ƒ๋ฅ™ํ•˜์—ฌ ํ•ญ๊ตฌ ๋„์‹œ๋ฅผ ์ ๋ นํ•˜๊ณ  ์šฐ๊ฐ€๋ฆฌํŠธ๋กœ ์ง„๊ฒฉํ•˜๊ณ  ์žˆ๋‹คโ€๋Š” ๋‚ด์šฉ์€ ์ด๋Ÿฌํ•œ ์™ธ๋ถ€ ์นจ๋žต์˜ ์ฆ๊ฑฐ๋ฅผ ๋ณด์—ฌ์ค€๋‹ค.

    ๊ทธ๋Ÿฌ๋‚˜ ๋ชจ๋“  ํŒŒ๊ดด๊ฐ€ ์™ธ๋ถ€ ์นจ๋žต์ž์— ์˜ํ•œ ๊ฒƒ์€ ์•„๋‹ˆ์—ˆ๋‹ค. ๋ฏธ์ผ€๋„ค์™€ ๊ฐ€๋‚˜์•ˆ์˜ ํ•˜์ดˆ๋ฅด(Hazor) ๊ฐ™์€ ์œ ์ ์ง€์—์„œ๋Š” ๊ถ์ „๊ณผ ์‹ ์ „์€ ๋ถˆํƒ”์ง€๋งŒ, ์ผ๋ฐ˜ ์ฃผ๋ฏผ๋“ค์˜ ์ฃผํƒ์€ ์˜จ์ „ํžˆ ๋‚จ์•„ ์žˆ๋Š” ๊ฒฝ์šฐ๊ฐ€ ๋ฐœ๊ฒฌ๋˜๋Š”๋ฐ, ์ด๋Š” ์™ธ๋ถ€ ์นจ๋žต๋ณด๋‹ค๋Š” ์‹๋Ÿ‰ ๋ถ€์กฑ๊ณผ ์‚ฌํšŒ ๋ถˆํ‰๋“ฑ์œผ๋กœ ์ธํ•œ โ€˜๋‚ด๋ถ€ ๋ฐ˜๋ž€(internal rebellion)โ€˜์˜ ๊ฐ€๋Šฅ์„ฑ์„ ์‹œ์‚ฌํ•œ๋‹ค.

  3. ์ž์—ฐ์žฌํ•ด: ์ง€์ง„์˜ ์œ„ํ˜‘(Natural Disasters: The Threat of Earthquakes): ์ผ๋ถ€ ๋„์‹œ์˜ ํŒŒ๊ดด๋Š” ์ธ๊ฐ„์˜ ์†์ด ์•„๋‹Œ ์ž์—ฐ์˜ ํž˜, ์ฆ‰ ์ง€์ง„์— ์˜ํ•œ ๊ฒƒ์ด์—ˆ๋‹ค. ๊ฑด๋ฌผ ๋ฒฝ์ด ๋น„๋šค์–ด์ง€๊ฑฐ๋‚˜, ์•„์น˜ํ˜• ๋ฌธํ‹€์˜ ์๊ธฐ๋Œ์ด ๋น ์ ธ ์žˆ๋Š” ๋“ฑ ์ง€์ง„์˜ ๋ช…ํ™•ํ•œ ํ”์ ๋“ค์ด ๋ฐœ๊ฒฌ๋œ๋‹ค. ํŠนํžˆ ๊ธฐ์›์ „ 1225๋…„์—์„œ 1175๋…„ ์‚ฌ์ด ์—๊ฒŒํ•ด์™€ ๋™๋ถ€ ์ง€์ค‘ํ•ด ์ง€์—ญ์—์„œ๋Š” โ€˜์ง€์ง„ ํญํ’(earthquake storm)โ€™ ๋˜๋Š” ์ง€์ง„ํ•™์ž๋“ค์ด โ€˜์ง€์ง„ ์—ฐ์†(earthquake sequence)โ€˜์ด๋ผ๊ณ  ๋ถ€๋ฅด๋Š” ํ˜„์ƒ์ด ๋ฐœ์ƒํ–ˆ๋˜ ๊ฒƒ์œผ๋กœ ๋ณด์ธ๋‹ค. ์ด๋Š” ํ•˜๋‚˜์˜ ๋‹จ์ธต์„ ์—์„œ ์••๋ ฅ์ด ์™„์ „ํžˆ ํ•ด์†Œ๋˜์ง€ ์•Š์•„ ์ˆ˜์‹ญ ๋…„์— ๊ฑธ์ณ ์—ฐ์‡„์ ์œผ๋กœ ์ง€์ง„์ด ๋ฐœ์ƒํ•˜๋Š” ํ˜„์ƒ์„ ์˜๋ฏธํ•œ๋‹ค. ํŠธ๋กœ์ด(Troy) ์œ ์ ์˜ ํŠธ๋กœ์ด 6์ธต์ด ์ธ๊ฐ„์ด ์•„๋‹Œ ์ง€์ง„์œผ๋กœ ํŒŒ๊ดด๋œ ๊ฒƒ์ด ๋Œ€ํ‘œ์ ์ธ ์˜ˆ๋‹ค. ํ˜„๋Œ€ ํŠ€๋ฅดํ‚ค์˜ˆ์˜ ์ง€์ง„ ๋ฐœ์ƒ ์ง€๋„๋ฅผ ํ›„๊ธฐ ์ฒญ๋™๊ธฐ ์‹œ๋Œ€ ํŒŒ๊ดด๋œ ๋„์‹œ ์ง€๋„์™€ ๊ฒน์ณ๋ณด๋ฉด ๊ฑฐ์˜ ์ผ๋Œ€์ผ๋กœ ์ผ์น˜ํ•˜๋Š” ์ƒ๊ด€๊ด€๊ณ„๋ฅผ ๋ณด์ธ๋‹ค.

  4. ์งˆ๋ณ‘์˜ ๊ทธ๋ฆผ์ž: ๊ณ ๋Œ€ ์„ธ๊ณ„์˜ ํŒฌ๋ฐ๋ฏน(The Shadow of Disease: Pandemics in the Ancient World): ๊ฐ€๋ญ„, ๊ธฐ๊ทผ, ์ „์Ÿ๊ณผ ํ•จ๊ป˜ ์งˆ๋ณ‘์€ ๋ฌธ๋ช… ๋ถ•๊ดด์˜ โ€˜๋„ค ๋ฒˆ์งธ ๊ธฐ์‚ฌโ€™์˜€๋‹ค. ํ˜ธ๋ฉ”๋กœ์Šค์˜ ใ€Ž์ผ๋ฆฌ์•„์Šค(The Iliad)ใ€์— ๋ฌ˜์‚ฌ๋œ ๋ฏธ์ผ€๋„ค ๊ตฐ๋Œ€์˜ ์—ญ๋ณ‘, ํžˆ๋ธŒ๋ฆฌ ์„ฑ๊ฒฝ ใ€Ž์ถœ์• ๊ตฝ๊ธฐ(Exodus)ใ€์˜ 10๊ฐ€์ง€ ์žฌ์•™ ๋“ฑ์€ ๋‹น์‹œ ์ „์—ผ๋ณ‘์— ๋Œ€ํ•œ ํฌ๋ฏธํ•œ ๊ธฐ์–ต์ผ ์ˆ˜ ์žˆ๋‹ค. ์‹ค์ œ๋กœ ๊ธฐ์›์ „ 1350๋…„๊ฒฝ ํžˆํƒ€์ดํŠธ ์™•๊ตญ์—์„œ๋Š” ์ด์ง‘ํŠธ ์ „์Ÿ ํฌ๋กœ๋“ค์ด ์˜ฎ๊ฒจ์˜จ ์—ญ๋ณ‘์œผ๋กœ ์ˆ˜ํ”ผ๋ฃฐ๋ฆฌ์šฐ๋งˆ 1์„ธ(Sufi Lulu-Luma I)๋ฅผ ๋น„๋กฏํ•œ ์™•์กฑ๊ณผ ๋ฐฑ์„ฑ๋“ค์ด ํฌ๊ฒŒ ํฌ์ƒ๋˜์—ˆ๋‹ค.

    ๋”์šฑ ๊ฒฐ์ •์ ์ธ ์ฆ๊ฑฐ๋Š” ๊ธฐ์›์ „ 1140๋…„๊ฒฝ ํŒŒ๋ผ์˜ค ๋žŒ์„ธ์Šค 5์„ธ(Ramses V)์˜ ๋ฏธ๋ผ์—์„œ ๋ฐœ๊ฒฌ๋œ ์ฒœ์—ฐ๋‘(smallpox) ํ”์ ์ด๋‹ค. ํˆฌ๋ฆฐ ํŒŒํ”ผ๋ฃจ์Šค(Turin papyrus)๋Š” ๋žŒ์„ธ์Šค 5์„ธ์™€ ๊ทธ์˜ ๊ฐ€์กฑ์ด ์—ญ๋ณ‘์œผ๋กœ ์‚ฌ๋งํ–ˆ์œผ๋ฉฐ, ์ด๋“ค์„ ๋งค์žฅํ•˜๋Š” ๋ฐ 16๊ฐœ์›”์ด ๊ฑธ๋ ธ๊ณ , ์‹ฌ์ง€์–ด ์™•๊ฐ€์˜ ๊ณ„๊ณก(Valley of the Kings)์„ ๋ด‰์‡„ํ•˜๋Š” ๋“ฑ โ€œ์„ธ๊ณ„ ์ตœ์ดˆ์˜ ๊ฒฉ๋ฆฌ(quarantine)โ€ ์กฐ์น˜๊ฐ€ ์ทจํ•ด์กŒ์Œ์„ ๊ธฐ๋กํ•˜๊ณ  ์žˆ๋‹ค. ์ด์ฒ˜๋Ÿผ ์งˆ๋ณ‘์€ ๋ถ•๊ดด ์ง์ „๊ณผ ์งํ›„์˜ ์‚ฌํšŒ๋ฅผ ๋”์šฑ ์ทจ์•ฝํ•˜๊ฒŒ ๋งŒ๋“ค์—ˆ์„ ๊ฒƒ์ด๋‹ค.

์‹œ์Šคํ…œ ๋ถ•๊ดด์™€ ๊ทธ ์ดํ›„: ์•”ํ‘๊ธฐ์™€ ์žฌํƒ„์ƒ

์ด ๋ชจ๋“  ๋ณตํ•ฉ์ ์ธ ์š”์ธ๋“ค์ด ํ•œ๊บผ๋ฒˆ์— ๋ฎ์น˜๋ฉด์„œ, ๊ณ ๋Œ€ ์‚ฌํšŒ๋Š” โ€˜์‹œ์Šคํ…œ ๋ถ•๊ดด(systems collapse)โ€˜๋ฅผ ๊ฒฝํ—˜ํ–ˆ๋‹ค. ์ผ€์ž„๋ธŒ๋ฆฌ์ง€ ๋Œ€ํ•™์˜ ์ฝœ๋ฆฐ ๋ Œํ”„๋ฃจ(Colin Renfrew) ๊ต์ˆ˜๊ฐ€ ์ œ์‹œํ•œ ์ด ์šฉ์–ด๋Š” ์ค‘์•™ ๊ฒฝ์ œ, ์ƒ์œ„ ์—˜๋ฆฌํŠธ, ์ค‘์•™ ์ •๋ถ€์˜ ๋ถ•๊ดด์™€ ํ•จ๊ป˜ ์ธ๊ตฌ ๊ฐ์†Œ(์‚ฌ๋ง ๋ฐ ์ด์ฃผ)๊ฐ€ ๋’ค๋”ฐ๋ฅด๋Š” ๋ณตํ•ฉ ์‹œ์Šคํ…œ์˜ ์™€ํ•ด๋ฅผ ์„ค๋ช…ํ•œ๋‹ค. ํ›„๊ธฐ ์ฒญ๋™๊ธฐ ์‹œ๋Œ€์˜ ๋ถ•๊ดด๋Š” ์ด๋Ÿฌํ•œ ์‹œ์Šคํ…œ ๋ถ•๊ดด์˜ ์ „ํ˜•์ ์ธ ์˜ˆ์˜€๋‹ค.

์‹œ์Šคํ…œ ๋ถ•๊ดด๋Š” ์ฆ‰๊ฐ์ ์ธ โ€˜์•”ํ‘๊ธฐ(Dark Age)โ€˜๋กœ ์ด์–ด์กŒ๋‹ค. ๋ฏธ์ผ€๋„ค ๊ทธ๋ฆฌ์Šค์—์„œ ์‚ฌ์šฉ๋˜๋˜ ์„ ํ˜• B ๋ฌธ์ž(Linear B)๋Š” ์‚ฌ๋ผ์กŒ๊ณ , ๋Œ€๊ทœ๋ชจ ๊ฑด์ถ• ๊ธฐ์ˆ ๋„ ์žŠํ˜”๋‹ค. ์‚ฌ๋žŒ๋“ค์€ ๊ธ‰๊ฒฉํžˆ ๋‚ฎ์€ ์ˆ˜์ค€์˜ ์‚ฌํšŒ-์ •์น˜-๊ฒฝ์ œ์  ๊ธฐ๋Šฅ์œผ๋กœ ํšŒ๊ท€ํ•ด์•ผ ํ–ˆ๋‹ค. ํ•˜์ง€๋งŒ ์ด ์•”ํ‘๊ธฐ๊ฐ€ ๋‹จ์ˆœํžˆ ํ‡ด๋ณด์˜ ์‹œ๊ธฐ๋งŒ์€ ์•„๋‹ˆ์—ˆ๋‹ค. โ€œํ•„์š”๋Š” ๋ฐœ๋ช…์˜ ์–ด๋จธ๋‹ˆ๋‹ค(necessity is the mother of invention)โ€œ๋ผ๋Š” ๋ง์ฒ˜๋Ÿผ, ์ฃผ์„ ๊ตฌํ•˜๊ธฐ๊ฐ€ ์–ด๋ ค์›Œ์ง€์ž ์‚ฌ๋žŒ๋“ค์€ ์ƒˆ๋กœ์šด ๊ธˆ์†์ธ ์ฒ (iron)์„ ํ™œ์šฉํ•˜๊ธฐ ์‹œ์ž‘ํ–ˆ๋‹ค. ํ‚คํ”„๋กœ์Šค์ธ๋“ค์€ ์ฒ ์„ ๊ฐ€์žฅ ๋จผ์ € ์‚ฌ์šฉํ–ˆ์œผ๋ฉฐ, ์ด๋Š” ์ฒญ๋™์„ ๋Œ€์ฒดํ•˜๋Š” ์ค‘์š”ํ•œ ์ž์›์ด ๋˜์—ˆ๋‹ค.

๋˜ํ•œ, ๊ฐ€๋‚˜์•ˆ ์ค‘๋ถ€(ํ˜„ ๋ ˆ๋ฐ”๋…ผ) ์ถœ์‹ ์˜ ํŽ˜๋‹ˆํ‚ค์•„์ธ(Phoenicians)๋“ค์€ ๋ถ•๊ดด ์ดํ›„ ๊ฐ€์žฅ ์„ฑ๊ณต์ ์œผ๋กœ ์ ์‘ํ•˜๊ณ  ๋ฒˆ์„ฑํ•œ ๋ฌธ๋ช… ์ค‘ ํ•˜๋‚˜์˜€๋‹ค. ์ด๋“ค์€ ๊ธฐ์กด์— ์žˆ๋˜ ์•ŒํŒŒ๋ฒณ์„ ํ‘œ์ค€ํ™”ํ•˜์—ฌ ์ง€์ค‘ํ•ด ์ „์—ญ์œผ๋กœ ํผ๋œจ๋ ธ๊ณ , ์ด๋Š” ๊ทธ๋ฆฌ์Šค ์•ŒํŒŒ๋ฒณ๊ณผ ๋ผํ‹ด ์•ŒํŒŒ๋ฒณ์˜ ๊ธฐ์›์ด ๋˜์–ด ์˜ค๋Š˜๋‚ ๊นŒ์ง€ ์ด์–ด์ง€๋Š” ๋ฌธ์ž ์ฒด๊ณ„์˜ ํ† ๋Œ€๊ฐ€ ๋˜์—ˆ๋‹ค. ์ฆ‰, ์•”ํ‘๊ธฐ๋Š” ํŒŒ๊ดด์™€ ํ•จ๊ป˜ ํ˜์‹ ์˜ ์”จ์•—์„ ํ’ˆ๊ณ  ์žˆ์—ˆ๋˜ ๊ฒƒ์ด๋‹ค.

ํ˜„๋Œ€์— ๋˜์ง€๋Š” ์งˆ๋ฌธ: ์ƒํ˜ธ ์—ฐ๊ฒฐ์„ฑ์˜ ์–‘๋‚ ์˜ ๊ฒ€

ํด๋ผ์ธ ๊ต์ˆ˜๋Š” ํ›„๊ธฐ ์ฒญ๋™๊ธฐ ์‹œ๋Œ€์˜ ๋ถ•๊ดด๊ฐ€ ์˜ค๋Š˜๋‚  ์šฐ๋ฆฌ์—๊ฒŒ ์ฃผ๋Š” ๊ฐ€์žฅ ํฐ ๊ตํ›ˆ์€ โ€˜์ƒํ˜ธ ์—ฐ๊ฒฐ์„ฑ(interconnectedness)โ€˜์˜ ์–‘๋ฉด์„ฑ์ด๋ผ๊ณ  ๊ฐ•์กฐํ•œ๋‹ค. ๋‹น์‹œ์˜ โ€˜์†Œ๊ทœ๋ชจ ์„ธ๊ณ„ ๋„คํŠธ์›Œํฌโ€™๋Š” ๋ฒˆ์˜์„ ๊ฐ€์ ธ์™”์ง€๋งŒ, ๋™์‹œ์— ํ•œ ์ง€์—ญ์˜ ์žฌ์•™์ด ์ „ ์„ธ๊ณ„์ ์ธ ๋„๋ฏธ๋…ธ ํšจ๊ณผ๋ฅผ ์ผ์œผํ‚ค๋Š” ์ทจ์•ฝ์ ์ด ๋˜์—ˆ๋‹ค.

๋ฏธ ์œก๊ตฐ ๊ณต๋ณ‘๋Œ€(US Army Corps of Engineers)์™€์˜ ์ปดํ“จํ„ฐ ์‹œ๋ฎฌ๋ ˆ์ด์…˜ ๊ฒฐ๊ณผ๋Š” ํฅ๋ฏธ๋กœ์šด ๊ฒฐ๋ก ์„ ๋„์ถœํ–ˆ๋‹ค. ํžˆํƒ€์ดํŠธ์™€ ์šฐ๊ฐ€๋ฆฌํŠธ๊ฐ€ ๋™์‹œ์— ๋ถ•๊ดดํ–ˆ์„ ๋•Œ ์ „์ฒด ๋„คํŠธ์›Œํฌ๊ฐ€ ์™€ํ•ด๋  ๊ฐ€๋Šฅ์„ฑ์ด ๊ฐ€์žฅ ๋†’์•˜๋‹ค๋Š” ๊ฒƒ์ด๋‹ค. ๋ฐ˜๋ฉด, ์•„์‹œ๋ฆฌ์•„๋‚˜ ๋ฐ”๋นŒ๋กœ๋‹ˆ์•„ ๊ฐ™์€ ๋‹ค๋ฅธ ๋ฌธ๋ช…๋“ค์˜ ๋ถ•๊ดด๋Š” ์ „์ฒด ๋„คํŠธ์›Œํฌ์— ์น˜๋ช…์ ์ด์ง€ ์•Š์•˜๋‹ค. ์ด์ง‘ํŠธ๋Š” ๋น„๋ก ํฐ ์–ด๋ ค์›€์„ ๊ฒช์—ˆ์ง€๋งŒ, ๋ถ•๊ดด์— ์ ์‘ํ•˜๊ณ  ์‚ด์•„๋‚จ์•˜๋‹ค. ์ด๋Š” ๋„คํŠธ์›Œํฌ ๋ถ•๊ดด๊ฐ€ ์‰ฝ์ง€ ์•Š์œผ๋ฉฐ, ์žฌ์•™์— ์–ด๋–ป๊ฒŒ ๋Œ€์‘ํ•˜๋А๋ƒ์— ๋”ฐ๋ผ ํšŒ๋ณต ํƒ„๋ ฅ์„ฑ(resilience)์ด ๋‹ฌ๋ผ์งˆ ์ˆ˜ ์žˆ์Œ์„ ์‹œ์‚ฌํ•œ๋‹ค.

ํด๋ผ์ธ ๊ต์ˆ˜๋Š” ์˜ค๋Š˜๋‚  ์šฐ๋ฆฌ๊ฐ€ ๊ธฐํ›„ ๋ณ€ํ™”, ํŒฌ๋ฐ๋ฏน, ๊ฒฝ์ œ ์œ„๊ธฐ, ์ง€์ •ํ•™์  ๊ฐˆ๋“ฑ ๋“ฑ ๋ณตํ•ฉ์ ์ธ ์œ„๊ธฐ์— ์ง๋ฉดํ•ด ์žˆ์Œ์„ ์ƒ๊ธฐ์‹œํ‚จ๋‹ค. ํ›„๊ธฐ ์ฒญ๋™๊ธฐ ์‹œ๋Œ€์˜ ๋ฌธ๋ช…๋“ค์€ ์ž์‹ ๋“ค์ด ๋ถ•๊ดดํ•˜๊ณ  ์žˆ๋‹ค๋Š” ์‚ฌ์‹ค์กฐ์ฐจ ์ œ๋Œ€๋กœ ์ธ์ง€ํ•˜์ง€ ๋ชปํ–ˆ์„ ์ˆ˜ ์žˆ๋‹ค. ๊ณผ์—ฐ ์šฐ๋ฆฌ๋Š” 3์ฒœ ๋…„ ์ „์˜ ์„ ์กฐ๋“ค๋ณด๋‹ค ํ˜„๋ช…ํ•˜๊ฒŒ ๋‹ค๊ฐ€์˜ค๋Š” โ€˜ํผํŽ™ํŠธ ์Šคํ†ฐโ€™์— ๋Œ€์ฒ˜ํ•  ์ˆ˜ ์žˆ์„๊นŒ? ์—ญ์‚ฌ๋Š” ๋ฐ˜๋ณต๋˜์ง€ ์•Š์ง€๋งŒ, ๊ทธ ์šด์œจ ์†์—์„œ ์šฐ๋ฆฌ๋Š” ๊ณผ๊ฑฐ์˜ ๊ตํ›ˆ์„ ํ†ตํ•ด ๋ฏธ๋ž˜๋ฅผ ์ค€๋น„ํ•  ์ง€ํ˜œ๋ฅผ ์ฐพ์•„์•ผ ํ•  ๊ฒƒ์ด๋‹ค.


โ€œAre Human Drivers Finally Obsolete? | Freakonomics Radioโ€ โ€” Freakonomics Radio Network ๊ธฐ๋ฐ˜ ๊ธฐ์‚ฌ ์›๋ณธ ์˜์ƒ ๋ณด๊ธฐ

์ธ๊ฐ„ ์šด์ „์ž๋Š” ์ •๋ง ๊ตฌ์‹œ๋Œ€์˜ ์œ ๋ฌผ์ด ๋ ๊นŒ? ์ž์œจ์ฃผํ–‰์ฐจ์˜ ๊ฟˆ๊ณผ ํ˜„์‹ค, ๊ทธ๋ฆฌ๊ณ  ๊ตฌ๊ธ€์˜ ๋น„๋ฐ€ ํ”„๋กœ์ ํŠธ

์šด์ „๋Œ€์—์„œ ์†์„ ๋–ผ๋Š” ๋‚ ์ด ์˜ฌ๊นŒ? SF ์˜ํ™”์—์„œ๋‚˜ ๋ณด๋˜ ์ž์œจ์ฃผํ–‰์ฐจ๊ฐ€ ์šฐ๋ฆฌ์˜ ์ผ์ƒ์— ๊นŠ์ˆ™์ด ํŒŒ๊ณ ๋“œ๋Š” ์‹œ๋Œ€๊ฐ€ ๋„๋ž˜ํ•˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค. ํŒŸ์บ์ŠคํŠธ โ€˜์„œ์น˜ ์—”์ง„(Search Engine)โ€˜์˜ ์ง„ํ–‰์ž PJ ๋ณดํŠธ(PJ Vote)๋Š” ์ตœ๊ทผ โ€˜ํ”„๋ฆฌ์ฝ”๋…ธ๋ฏน์Šค ๋ผ๋””์˜ค(Freakonomics Radio)โ€˜์™€์˜ ํŠน๋ณ„ ํ˜‘์—…์„ ํ†ตํ•ด ์ž์œจ์ฃผํ–‰์ฐจ์˜ ๊ณผ๊ฑฐ, ํ˜„์žฌ, ๊ทธ๋ฆฌ๊ณ  ๋ฏธ๋ž˜๋ฅผ ์‹ฌ์ธต์ ์œผ๋กœ ํƒ๊ตฌํ•˜๋Š” 2๋ถ€์ž‘ ์‹œ๋ฆฌ์ฆˆ๋ฅผ ์„ ๋ณด์˜€์Šต๋‹ˆ๋‹ค. ์ด ๊ธฐ์‚ฌ๋Š” ๊ทธ ์ฒซ ๋ฒˆ์งธ ์ด์•ผ๊ธฐ์—์„œ PJ๊ฐ€ ๋“ค๋ ค์ค€ ๋†€๋ผ์šด ์—ฌ์ •์„ ๋”ฐ๋ผ๊ฐ€๋ฉฐ, ์ธ๊ฐ„ ์šด์ „์ž๊ฐ€ โ€˜๊ตฌ์‹œ๋Œ€์˜ ์œ ๋ฌผโ€™์ด ๋  ์ˆ˜ ์žˆ๋‹ค๋Š” ๋Œ€๋‹ดํ•œ ์˜ˆ์ธก์˜ ๊ทผ๊ฑฐ์™€ ์ž์œจ์ฃผํ–‰ ๊ธฐ์ˆ  ๋ฐœ์ „์˜ ์ˆจ๊ฒจ์ง„ ์—ญ์‚ฌ๋ฅผ ํŒŒํ—ค์ณ ๋ด…๋‹ˆ๋‹ค.

์šด์ „๋Œ€์—์„œ ์†์„ ๋–ผ๊ณ  ์‹ถ์—ˆ๋˜ ์ด์œ : ๋ถ€์ƒ์—์„œ ์‹œ์ž‘๋œ ์ž์œจ์ฃผํ–‰ ๊ฒฝํ—˜

PJ ๋ณดํŠธ๊ฐ€ ์ž์œจ์ฃผํ–‰์ฐจ์— ๊ด€์‹ฌ์„ ๊ฐ–๊ฒŒ ๋œ ๊ณ„๊ธฐ๋Š” ๋‹ค์†Œ ์˜์™ธ์˜ ๊ฐœ์ธ์ ์ธ ๊ฒฝํ—˜์—์„œ ์‹œ์ž‘๋˜์—ˆ์Šต๋‹ˆ๋‹ค. ๋ฒค์น˜ํ”„๋ ˆ์Šค์— ๋„ˆ๋ฌด ๋ชฐ๋‘ํ•˜๋‹ค ๋ถ€์ƒ์„ ์ž…์–ด ์„œํ˜œ๋ถ€ ํƒˆ์žฅ(hernia) ์ˆ˜์ˆ ์„ ๋ฐ›๊ฒŒ ๋œ ๊ทธ๋Š” ๊ทน์‹ฌํ•œ ํ†ต์ฆ๊ณผ ์ œํ•œ๋œ ๊ฑฐ๋™์œผ๋กœ ์–ด๋ ค์›€์„ ๊ฒช์—ˆ์Šต๋‹ˆ๋‹ค. ์ƒŒํ”„๋ž€์‹œ์Šค์ฝ”์— ์žˆ๋Š” ์นœ๊ตฌ๋ฅผ ๋ฐฉ๋ฌธํ–ˆ์„ ๋•Œ, ๊ทธ๋Š” โ€˜์›จ์ด๋ชจ(Waymo)โ€™ ๋กœ๋ณดํƒ์‹œ๋ฅผ ์ฒ˜์Œ ์ด์šฉํ•˜๊ฒŒ ๋ฉ๋‹ˆ๋‹ค.

โ€œ๋ฒ„ํŠผ ํ•˜๋‚˜๋งŒ ๋ˆ„๋ฅด๋ฉด ์•„๋ฌด๋„ ์šด์ „ํ•˜์ง€ ์•Š๋Š” ์ฐจ๊ฐ€ ํŠ€์–ด๋‚˜์™€์š”. ์ฐจ์— ํƒ€๋ฉด ํ•ธ๋“ค์ด ์ €์ ˆ๋กœ ๋Œ์•„๊ฐ€๋Š” ๊ฑธ ๋ณด๊ฒŒ ๋˜์ฃ . ๋งˆ์น˜ ๋น„ํ–‰๊ธฐ๋ฅผ ์ฒ˜์Œ ํƒ”์„ ๋•Œ์˜ ๊ฒฝ์ด๋กœ์›€ ๊ฐ™์•˜์–ด์š”. ์„ธ ๋ฒˆ์งธ ํƒˆ ๋•Œ๋Š” ๊ทธ๋ƒฅ ์—˜๋ฆฌ๋ฒ ์ดํ„ฐ ํƒ€๋Š” ๋А๋‚Œ์ด ๋“ค์—ˆ๊ณ ์š”.โ€

์ด ๊ฒฝํ—˜์€ PJ์—๊ฒŒ ๊นŠ์€ ์ธ์ƒ์„ ๋‚จ๊ฒผ์Šต๋‹ˆ๋‹ค. โ€œ๋งŽ์€ ๊ฒƒ์ด ๋ณ€ํ•  ๊ฒƒโ€์ด๋ผ๋Š” ์˜ˆ๊ฐ๊ณผ ํ•จ๊ป˜, ์™œ ์‚ฌ๋žŒ๋“ค์ด ์ด ๋ณ€ํ™”์— ๋Œ€ํ•ด ๋” ๋งŽ์ด ์ด์•ผ๊ธฐํ•˜์ง€ ์•Š๋Š”์ง€ ์˜์•„ํ•จ์„ ๋А๊ผˆ์Šต๋‹ˆ๋‹ค. ๊ทธ์˜ 2๋ถ€์ž‘ ์‹œ๋ฆฌ์ฆˆ๋Š” ์ด ์งˆ๋ฌธ์— ๋Œ€ํ•œ ๋‹ต์„ ์ฐพ์•„๊ฐ€๋Š” ์—ฌ์ •์ž…๋‹ˆ๋‹ค. ์ฒซ ๋ฒˆ์งธ ์—ํ”ผ์†Œ๋“œ๋Š” โ€˜์ฐจโ€™ ์ž์ฒด์— ์ดˆ์ ์„ ๋งž์ถ”๊ณ , ๋‘ ๋ฒˆ์งธ ์—ํ”ผ์†Œ๋“œ๋Š” โ€˜์šด์ „์žโ€™์— ๋Œ€ํ•œ ์ด์•ผ๊ธฐ๋ฅผ ๋‹ค๋ฃน๋‹ˆ๋‹ค.

์ธ๊ฐ„ ์šด์ „์ž๋Š” ์–ผ๋งˆ๋‚˜ โ€˜์—‰๋งโ€™์ผ๊นŒ?

PJ ๋ณดํŠธ๋Š” ์ธ๊ฐ„ ์šด์ „์ž์˜ ํ•œ๊ณ„์— ๋Œ€ํ•ด ์†”์งํ•˜๊ฒŒ ์ด์•ผ๊ธฐํ•ฉ๋‹ˆ๋‹ค. โ€œ์ธ๊ฐ„์ด ์šด์ „ํ•˜๋Š” ๊ฑธ ๋ณธ ์ ์ด ์žˆ๋‚˜์š”? ์šฐ๋ฆฌ ์Šค์Šค๋กœ๋ฅผ ํฌํ•จํ•ด์„œ์š”. ์šฐ๋ฆฌ๋Š” ๊ทธ๋ ‡๊ฒŒ ์ž˜ํ•˜์ง€ ๋ชปํ•ฉ๋‹ˆ๋‹ค.โ€ ์Šคํ‹ฐ๋ธ ๋”๋ธŒ๋„ˆ(Steven Dubner) ๋˜ํ•œ ์ด์— ๋™์˜ํ•˜๋ฉฐ ์ž์‹ ์˜ ์šด์ „ ์‹ค๋ ฅ์— ํ™˜์ƒ์ด ์—†์Œ์„ ๊ณ ๋ฐฑํ•ฉ๋‹ˆ๋‹ค. โ€œ์ €๋Š” ์šด์ „ ์‹ค๋ ฅ์— ๋Œ€ํ•œ ํ™˜์ƒ์ด ์—†์–ด์š”. ์ €๋Š” ์„ฑ์งˆ์ด ๊ธ‰ํ•˜๊ณ , ์‚ฐ๋งŒํ•˜์ฃ .โ€

์‹ค์ œ๋กœ ์šฐ๋ฆฌ ๋Œ€๋ถ€๋ถ„์—๊ฒŒ ์šด์ „์€ ์ผ์ƒ์ ์œผ๋กœ ํ•˜๋Š” ํ–‰๋™ ์ค‘ ๊ฐ€์žฅ ์œ„ํ—˜ํ•œ ์ถ•์— ์†ํ•ฉ๋‹ˆ๋‹ค. ์ž์œจ์ฃผํ–‰์ฐจ์˜ ํ•ต์‹ฌ์ ์ธ ํ™๋ณด ๋ฌธ๊ตฌ๋Š” ๋ฐ”๋กœ โ€˜์•ˆ์ „โ€™์ž…๋‹ˆ๋‹ค. ์ปดํ“จํ„ฐ๊ฐ€ ์šด์ „ํ•˜๋Š” ์ฐจ๋Š” ์Œ์ฃผ์šด์ „, ์กธ์Œ์šด์ „, ์ฃผ์˜ ์‚ฐ๋งŒ, ๋ฌธ์ž ๋ฉ”์‹œ์ง€ ์ „์†ก, ๋ณด๋ณต ์šด์ „์„ ํ•˜์ง€ ์•Š์Šต๋‹ˆ๋‹ค. ๊ทธ๋ฆฌ๊ณ  ์ด๋Ÿฌํ•œ ์ž์œจ์ฃผํ–‰์ฐจ๋Š” ๋” ์ด์ƒ ๋ฏธ๋ž˜์˜ ์ด์•ผ๊ธฐ๊ฐ€ ์•„๋‹™๋‹ˆ๋‹ค. ์›จ์ด๋ชจ ๊ฐ™์€ ๋กœ๋ณดํƒ์‹œ๋Š” ์ด๋ฏธ ๋ฏธ๊ตญ 10๊ฐœ ๋„์‹œ์—์„œ ์ˆ˜๋ฐฑ๋งŒ ๊ฑด์˜ ์šดํ–‰์„ ์ œ๊ณตํ•˜๊ณ  ์žˆ์œผ๋ฉฐ, ์ค‘๊ตญ์—์„œ๋Š” ๋” ๊ด‘๋ฒ”์œ„ํ•˜๊ฒŒ ํ™•์‚ฐ๋˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค. ์ƒŒํ”„๋ž€์‹œ์Šค์ฝ”๋‚˜ ์˜ค์Šคํ‹ด ๊ฐ™์€ ๋„์‹œ์—์„œ๋Š” ์ž์œจ์ฃผํ–‰์ฐจ๊ฐ€ ์šฐ๋ฒ„(Uber)๋งŒํผ์ด๋‚˜ ํ”ํ•œ ํ’๊ฒฝ์ด ๋˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค.

์‚ฌ๋ผ์ง„ ์ง์—…, ๊ทธ๋ฆฌ๊ณ  ๋ณ€ํ™”์˜ ์˜ˆ๊ณ 

์ž์œจ์ฃผํ–‰์ฐจ์˜ ๋“ฑ์žฅ์€ ์ธ๋ฅ˜ ์—ญ์‚ฌ์—์„œ ๋ฐ˜๋ณต๋˜์–ด ์˜จ ๊ฑฐ๋Œ€ํ•œ ๋ณ€ํ™”์˜ ํ๋ฆ„๊ณผ ๋งž๋‹ฟ์•„ ์žˆ์Šต๋‹ˆ๋‹ค. PJ๋Š” 200๋…„ ์ „์œผ๋กœ ๊ฑฐ์Šฌ๋Ÿฌ ์˜ฌ๋ผ๊ฐ€ ์„ธ ๊ฐ€์ง€ ์ง์—…์„ ์ƒ์ƒํ•˜๊ฒŒ ํ•ฉ๋‹ˆ๋‹ค.

  1. ๋…ธ์ปค์—…ํผ(Knocker-upper): ์•Œ๋žŒ ์‹œ๊ณ„๊ฐ€ ๋ฐœ๋ช…๋˜๊ธฐ ์ „, ๊ธด ๋ง‰๋Œ€๊ธฐ๋กœ ์‚ฌ๋žŒ๋“ค์˜ ์ฐฝ๋ฌธ์„ ๋‘๋“œ๋ ค ๊นจ์›Œ์ฃผ๋˜ ์ง์—…. ์ด์ œ๋Š” ์Šค๋งˆํŠธํฐ ์•Œ๋žŒ์ด ๊ทธ ์ž๋ฆฌ๋ฅผ ๋Œ€์‹ ํ•ฉ๋‹ˆ๋‹ค.
  2. ๋žจํ”„ ๋ผ์ดํ„ฐ(Lamp Lighter): ๋ฐค์— ๊ฐ€์Šค๋“ฑ์„ ์ผœ๊ณ  ์ƒˆ๋ฒฝ์— ๋„๋˜ ์ง์—…. ์ด์ œ๋Š” ์ „๊ธฐ ๊ฐ€๋กœ๋“ฑ์ด ์ž๋™์œผ๋กœ ๋ฐํ˜€์ง‘๋‹ˆ๋‹ค.
  3. ๋งˆ๋ถ€(Coachman): ๋ง์„ ํƒ€๊ณ  ์Šน๊ฐ์„ ์›ํ•˜๋Š” ๊ณณ์œผ๋กœ ๋ฐ๋ ค๋‹ค์ฃผ๋˜ ์ง์—…. ์ด์ œ๋Š” ์ž๋™์ฐจ ์šด์ „์ž๊ฐ€ ๊ทธ ์—ญํ• ์„ ํ•ฉ๋‹ˆ๋‹ค.

์ด ์ค‘ ์ฒซ ๋‘ ์ง์—…์€ ์—ญ์‚ฌ ์†์œผ๋กœ ์‚ฌ๋ผ์ง„ ์ง€ ์˜ค๋ž˜์ž…๋‹ˆ๋‹ค. ํ•˜์ง€๋งŒ โ€˜์šด์ „์žโ€™๋Š” ์—ฌ์ „ํžˆ ์กด์žฌํ•ฉ๋‹ˆ๋‹ค. ๊ทธ๋Ÿฌ๋‚˜ PJ๋Š” ์ด ์ง์—… ๋˜ํ•œ ๋ณ€ํ™”์˜ ๊ธฐ๋กœ์— ์„œ ์žˆ๋‹ค๊ณ  ๋งํ•ฉ๋‹ˆ๋‹ค. โ€œ์šด์ „์ž๋ผ๋Š” ๋‹จ์–ด๊ฐ€ ์ง€๊ธˆ์€ ์ธ๊ฐ„์„ ๋– ์˜ฌ๋ฆฌ๊ฒŒ ํ•˜์ง€๋งŒ, ๊ณง ์‹๊ธฐ์„ธ์ฒ™๊ธฐ(dishwasher), ํ”„๋ฆฐํ„ฐ(printer), ์ปดํ“จํ„ฐ(computer)์ฒ˜๋Ÿผ ๊ธฐ๊ณ„๋ฅผ ์ง€์นญํ•˜๋Š” ๋‹จ์–ด๋กœ ๋ฐ”๋€” ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.โ€

๊ฟˆ์€ ์˜ค๋ž˜๋˜์—ˆ์ง€๋งŒ, ํ˜„์‹ค์€ ๋‹ฌ๋ž๋‹ค: ์ž๋™์ฐจ์˜ ๋“ฑ์žฅ๊ณผ ์ดˆ๊ธฐ ์ €ํ•ญ

ํฅ๋ฏธ๋กญ๊ฒŒ๋„, ์ธ๊ฐ„ ์šด์ „์ž๋ฅผ ๊ธฐ๊ณ„๋กœ ๋Œ€์ฒดํ•˜๋ ค๋Š” ๊ฟˆ์€ ๋žจํ”„ ๋ผ์ดํ„ฐ์˜ ์‹œ๋Œ€๋งŒํผ์ด๋‚˜ ์˜ค๋ž˜๋˜์—ˆ์Šต๋‹ˆ๋‹ค. ๋งˆ์ฐจ์—์„œ ์ž๋™์ฐจ๋กœ ๋„˜์–ด์˜ค๋ฉด์„œ ์ธ๋ฅ˜๋Š” ํ•œ ๊ฐ€์ง€ ์ค‘์š”ํ•œ ๊ฒƒ์„ ์žƒ์—ˆ์Šต๋‹ˆ๋‹ค. ๋ฐ”๋กœ โ€˜๊ฐ๊ฐ(sentience)โ€˜์ž…๋‹ˆ๋‹ค. ๋ง์€ ๊ณ ์‚๋ฅผ ๋†“์•„๋„ ์ ˆ๋ฒฝ์œผ๋กœ ๋‹ฌ๋ ค๊ฐ€์ง€ ์•Š์Šต๋‹ˆ๋‹ค. ํ•˜์ง€๋งŒ ์ž๋™์ฐจ๋Š” ๊ทธ๋ ‡์ง€ ์•Š์Šต๋‹ˆ๋‹ค.

1800๋…„๋Œ€ ํ›„๋ฐ˜, ์ฆ๊ธฐ๊ธฐ๊ด€์ฐจ์™€ ๊ฐ€์†”๋ฆฐ ์ž๋™์ฐจ๊ฐ€ ๋„๋กœ๋ฅผ ์งˆ์ฃผํ•˜๊ธฐ ์‹œ์ž‘ํ–ˆ์„ ๋•Œ, ์‚ฌ๋žŒ๋“ค์€ ์ด ์ƒˆ๋กœ์šด ๊ธฐ์ˆ ์— ๋Œ€ํ•ด ๊ฒฉ๋ ฌํ•œ ๋…ผ์Ÿ์„ ๋ฒŒ์˜€์Šต๋‹ˆ๋‹ค. ๋‘๋ ค์›€์€ ํฌ๊ฒŒ ๋‘ ๊ฐ€์ง€์˜€์Šต๋‹ˆ๋‹ค.

  1. ์ผ์ž๋ฆฌ ์œ„ํ˜‘: ๋งˆ๋ถ€, ๋งˆ๊ตฌ๊ฐ„ ์ฃผ์ธ, ๋งˆ๊ตฌ ์ œ์กฐ์‚ฌ, ๋งˆ๋ถ„ ์ฒ˜๋ฆฌ์—…์ž ๋“ฑ ๋ง๊ณผ ๊ด€๋ จ๋œ ์ˆ˜๋งŽ์€ ๋…ธ๋™์ž๋“ค์˜ ์ผ์ž๋ฆฌ๊ฐ€ ์‚ฌ๋ผ์งˆ ๊ฒƒ์ด๋ผ๋Š” ์šฐ๋ ค๊ฐ€ ์ปธ์Šต๋‹ˆ๋‹ค. ์˜ค๋Š˜๋‚  โ€˜ํŒ€์Šคํ„ฐ์ฆˆ(Teamsters)โ€˜๋ผ๋Š” ๋‹จ์–ด๊ฐ€ ํŠธ๋Ÿญ ์šด์ „์‚ฌ ๋…ธ์กฐ๋ฅผ ์—ฐ์ƒ์‹œํ‚ค์ง€๋งŒ, ์›๋ž˜๋Š” ๋ง์„ ๋ชจ๋Š” ์‚ฌ๋žŒ๋“ค์„ ์ง€์นญํ–ˆ์Šต๋‹ˆ๋‹ค.
  2. ์•ˆ์ „ ๋ฌธ์ œ: ๊ทœ์ œ ์—†์ด ๋„์ž…๋œ ์ž๋™์ฐจ๋Š” ์—„์ฒญ๋‚œ ์‚ฌ๋ง๋ฅ ์„ ๊ธฐ๋กํ–ˆ์Šต๋‹ˆ๋‹ค. ์šด์ „๋ฉดํ—ˆ, ์‹ ํ˜ธ๋“ฑ, ๋ฐฉํ–ฅ ์ง€์‹œ๋“ฑ์ด ์—†๋˜ 1900๋…„๋Œ€ ์ดˆ ๋””ํŠธ๋กœ์ดํŠธ์—์„œ๋Š” ์ˆ˜๋งŽ์€ ์‚ฌ๋ง์ž๊ฐ€ ๋ฐœ์ƒํ–ˆ๊ณ , ๊ทธ ์ค‘ ์ƒ๋‹น์ˆ˜๋Š” ์•„์ด๋“ค์ด์—ˆ์Šต๋‹ˆ๋‹ค.

์ผ๋ถ€ ๋ฐ˜(ๅ)์ž๋™์ฐจ ์šด๋™๊ฐ€๋“ค์€ โ€˜์ ๊ธฐ ์กฐ๋ก€(Red Flag Laws)โ€˜์™€ ๊ฐ™์€ ๊ทœ์ œ๋ฅผ ํ†ตํ•ด ๊ธฐ์ˆ ์˜ ๋„์ž…์„ ๋ง‰๊ฑฐ๋‚˜ ๋Šฆ์ถ”๋ ค ํ–ˆ์Šต๋‹ˆ๋‹ค. ์˜ˆ๋ฅผ ๋“ค์–ด, ํŽœ์‹ค๋ฒ ์ด๋‹ˆ์•„์—์„œ๋Š” ๊ฐ€์ถ•์„ ๋งŒ๋‚˜๋Š” ๋ฌด๋งˆ์ฐจ(horseless carriage) ์šด์ „์ž๋Š” ์ฐจ๋ฅผ ๋ถ„ํ•ดํ•˜์—ฌ ๋ค๋ถˆ ๋’ค์— ์ˆจ๊ฒจ์•ผ ํ•œ๋‹ค๋Š” ๋ฒ•์•ˆ์ด ์ œ์•ˆ๋˜๊ธฐ๋„ ํ–ˆ์Šต๋‹ˆ๋‹ค(์ฃผ์ง€์‚ฌ์˜ ๊ฑฐ๋ถ€๊ถŒ ํ–‰์‚ฌ๋กœ ๋ฌด์‚ฐ). ์ด๋Ÿฌํ•œ ๊ทœ์ œ๋Š” ํ™ฉ๋‹นํ•˜๊ฒŒ ๋“ค๋ฆด์ง€ ๋ชจ๋ฅด์ง€๋งŒ, ๊ทธ๋“ค์˜ ์˜ˆ์ธก์€ ์–ด๋А ์ •๋„ ์˜ณ์•˜์Šต๋‹ˆ๋‹ค. ์ž๋™์ฐจ๋Š” ์ดˆ๊ธฐ์—๋Š” ๋งŽ์€ ์ผ์ž๋ฆฌ๋ฅผ ์—†์•ด๊ณ , ๋งค์šฐ ์œ„ํ—˜ํ–ˆ์Šต๋‹ˆ๋‹ค.

์‚ฌํšŒ๊ฐ€ ์ž๋™์ฐจ์— ์ ์‘ํ•˜๋Š” ๋ฐ๋Š” ์ˆ˜์‹ญ ๋…„์ด ๊ฑธ๋ ธ์Šต๋‹ˆ๋‹ค. ๋ฒ•๊ทœ, ์šด์ „๋ฉดํ—ˆ, ์šด์ „ ๊ต์œก, ๋„๋กœ ์„ค๊ณ„, ๊ณ ์†๋„๋กœ, ์•ˆ์ „๋ฒจํŠธ, ์—์–ด๋ฐฑ ๋“ฑ์ด ๋ฐœ๋ช…๋˜๋ฉด์„œ ์šด์ „์€ ๋œ ์œ„ํ—˜ํ•ด์กŒ์Šต๋‹ˆ๋‹ค. ํ•˜์ง€๋งŒ ์Šค๋งˆํŠธํฐ์˜ ๋“ฑ์žฅ์€ ์ด๋Ÿฌํ•œ ๋ฐœ์ „์„ ์ผ๋ถ€ ๋˜๋Œ๋ ค ๋†“๊ธฐ๋„ ํ–ˆ์Šต๋‹ˆ๋‹ค. ์˜ค๋Š˜๋‚  ๋ฏธ๊ตญ์—์„œ ์ž๋™์ฐจ ์‚ฌ๊ณ  ์‚ฌ๋ง์ž ์ˆ˜๋Š” ์ด๊ธฐ๋‚˜ ์˜คํ”ผ์˜ค์ด๋“œ(opioids) ์ค‘๋…์œผ๋กœ ์ธํ•œ ์‚ฌ๋ง์ž ์ˆ˜์™€ ๋น„์Šทํ•˜๋ฉฐ, ์•ฝ 100๋ช… ์ค‘ 1๋ช…๊ผด๋กœ ๋ฐœ์ƒํ•ฉ๋‹ˆ๋‹ค.

์‚ฌ๋ง‰์—์„œ ํ”ผ์–ด๋‚œ ์ž์œจ์ฃผํ–‰์˜ ๊ฟˆ: DARPA ๊ทธ๋žœ๋“œ ์ฑŒ๋ฆฐ์ง€

์ž๋™์ฐจ์˜ ์•ˆ์ „ ๋ฌธ์ œ๋ฅผ ๊ทผ๋ณธ์ ์œผ๋กœ ํ•ด๊ฒฐํ•˜๊ณ , ์ž๋™์ฐจ๋ฅผ ๋‹ค์‹œ โ€˜๊ฐ๊ฐ์ โ€™์œผ๋กœ ๋งŒ๋“ค๋ ค๋Š” ๊ฟˆ์€ ์˜ค๋žซ๋™์•ˆ ์ด์–ด์ ธ ์™”์Šต๋‹ˆ๋‹ค. ์ดˆ๊ธฐ์—๋Š” ๋ฌด์„  ์กฐ์ข… ์ฐจ๋Ÿ‰์ด๋‚˜ ๋„๋กœ ๋ฐ‘์— ์ž์„์„ ์„ค์น˜ํ•˜๋Š” ๋“ฑ ๋‹ค์–‘ํ•œ ์•„์ด๋””์–ด๊ฐ€ ์‹œ๋„๋˜์—ˆ์ง€๋งŒ, ๊ธฐ์ˆ ์˜ ํ•œ๊ณ„์— ๋ถ€๋”ชํ˜”์Šต๋‹ˆ๋‹ค.

2002๋…„, ๋ฏธ๊ตญ ๊ตญ๋ฐฉ๋ถ€ ์‚ฐํ•˜์˜ ๋น„๋ฐ€ ์—ฐ๊ตฌ๊ธฐ๊ด€์ธ DARPA(Defense Advanced Research Projects Agency, ๊ณ ๋“ฑ๋ฐฉ์œ„์—ฐ๊ตฌ๊ณ„ํš๊ตญ)๊ฐ€ ์ด ๊ฟˆ์— ๋ถˆ์„ ์ง€ํˆ์Šต๋‹ˆ๋‹ค. GPS, M16 ์†Œ์ด, ์ดˆ๊ธฐ ์ธํ„ฐ๋„ท, ํ”„๋ ˆ๋ฐํ„ฐ ๋“œ๋ก (Predator drone) ๋“ฑ์„ ๊ฐœ๋ฐœํ•œ DARPA๋Š” ์ž์œจ ๊ตฐ์šฉ ์ฐจ๋Ÿ‰ ๊ฐœ๋ฐœ์— ๊ด€์‹ฌ์„ ๊ฐ€์กŒ์Šต๋‹ˆ๋‹ค. ๋‹น์‹œ DARPA ๊ตญ์žฅ์ด์—ˆ๋˜ ํ† ๋‹ˆ ํ…Œ๋”(Tony Tether)๋Š” โ€œ์ฝ˜ํ…Œ์ŠคํŠธ๋ฅผ ์—ด์–ด ์ด ๋ชจ๋“  ๊ธฐ์ˆ ๋“ค์„ ์ œ๋Œ€๋กœ ๋œ ์ž์œจ์ฃผํ–‰์ฐจ๋กœ ๋งŒ๋“ค ์ˆ˜ ์žˆ๋Š”์ง€ ๋ณด์žโ€๊ณ  ์ œ์•ˆํ–ˆ์Šต๋‹ˆ๋‹ค.

๊ทธ์˜ ์›๋ž˜ ์•„์ด๋””์–ด๋Š” ๋ผ์Šค๋ฒ ์ด๊ฑฐ์Šค ์ŠคํŠธ๋ฆฝ์„ ๋‹ฌ๋ฆฌ๋Š” ๊ฒƒ์ด์—ˆ์ง€๋งŒ, ๋„์‹œ ์ „์ฒด๋ฅผ ๋งˆ๋น„์‹œํ‚ฌ ์ˆ˜๋Š” ์—†์—ˆ๊ธฐ์— ๊ฒฐ๊ตญ ๋ผ์Šค๋ฒ ์ด๊ฑฐ์Šค ์™ธ๊ณฝ์˜ ์‚ฌ๋ง‰์—์„œ ๊ฒฝ์ฃผ๋ฅผ ํŽผ์น˜๊ธฐ๋กœ ํ•ฉ๋‹ˆ๋‹ค. DARPA๋Š” โ€˜๋ฏธ๊ตญ ๋ณ‘์‚ฌโ€™๋ผ๋Š” ์šด์ „์ž๋ฅผ ๋Œ€์ฒดํ•˜๊ณ ์ž ํ–ˆ์Šต๋‹ˆ๋‹ค. ์ง€๋ขฐ๊ฐ€ ๊น”๋ ค ์žˆ์„์ง€ ๋ชจ๋ฅด๋Š” ๋„๋กœ๋ฅผ ์Šค์Šค๋กœ ์šด์ „ํ•  ์ˆ˜ ์žˆ๋Š” ์ฐจ๋Ÿ‰์ด ํ•„์š”ํ–ˆ๋˜ ๊ฒƒ์ž…๋‹ˆ๋‹ค. ์ƒ๊ธˆ์€ 100๋งŒ ๋‹ฌ๋Ÿฌ. ๊ทœ์น™์€ ๋งค์šฐ ์ž์œ ๋กœ์› ์Šต๋‹ˆ๋‹ค. ์ฐจ๋Ÿ‰ ๊ฐ„ ํ†ต์‹ ๋งŒ ๊ธˆ์ง€๋˜์—ˆ์„ ๋ฟ, ๋ฐ”ํ€ด ์ˆ˜๋‚˜ ํ˜•ํƒœ๋Š” ์ œํ•œ์ด ์—†์—ˆ์Šต๋‹ˆ๋‹ค. ์‹ฌ์ง€์–ด โ€œ๋‹ค๋ฅธ ์ฐจ๋Ÿ‰์„ ๊ณต๊ฒฉํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๊นŒ?โ€๋ผ๋Š” ์งˆ๋ฌธ์— โ€œ์•ˆ ๋œ๋‹คโ€๋Š” ๋‹ต์ด ๋Œ์•„์™”์„ ์ •๋„๋กœ ์ฐธ๊ฐ€์ž๋“ค์˜ ์ƒ์ƒ๋ ฅ์€ ๋„˜์ณค์Šต๋‹ˆ๋‹ค. ๋งŽ์€ ์ฐธ๊ฐ€ํŒ€์€ โ€˜๋ฐฐํ‹€๋ด‡(Battlebots)โ€™ ๋Œ€ํšŒ ์ถœ์‹ ๋“ค์ด์—ˆ์Šต๋‹ˆ๋‹ค.

2004๋…„ ๊ทธ๋žœ๋“œ ์ฑŒ๋ฆฐ์ง€: ์ฒ˜์ฐธํ•œ ์‹คํŒจ

2004๋…„ 3์›”์˜ ์ฒซ ๋ฒˆ์งธ ๊ทธ๋žœ๋“œ ์ฑŒ๋ฆฐ์ง€๋Š” โ€œ์™„์ „ํ•œ ํžˆ์Šคํ…Œ๋ฆฌ์  ์žฌ์•™(utter hysterical disaster)โ€œ์ด์—ˆ์Šต๋‹ˆ๋‹ค.

  • ์•ค์„œ๋‹ˆ ๋ฅด์œˆ๋„์Šคํ‚ค(Anthony Lewendowski)์˜ โ€˜๊ณ ์ŠคํŠธ ๋ผ์ด๋”(Ghost Rider)โ€™: ์œ ์ผํ•œ ์ž์œจ์ฃผํ–‰ ์˜คํ† ๋ฐ”์ด์˜€์ง€๋งŒ, ์•ˆ์ •ํ™” ์‹œ์Šคํ…œ ์Šค์œ„์น˜๋ฅผ ์ผœ๋Š” ๊ฒƒ์„ ์žŠ์–ด๋ฒ„๋ ค ์ถœ๋ฐœ๊ณผ ๋™์‹œ์— ์ „๋ณต๋˜์—ˆ์Šต๋‹ˆ๋‹ค.
  • ์นด๋„ค๊ธฐ๋ฉœ๋ก (Carnegie Mellon) ํŒ€์˜ โ€˜์ƒŒ๋“œ์Šคํ†ฐ(Sandstorm)โ€™: ๋ฐ์€ ๋นจ๊ฐ„์ƒ‰ ํ—˜๋น„(Humvee)์— ๋ฏธ๋ž˜ ์ง€ํ–ฅ์ ์ธ ์„ผ์„œ๋“ค์ด ์žฅ์ฐฉ๋œ ์ฐจ๋Ÿ‰์ด์—ˆ์ง€๋งŒ, ์ถœ๋ฐœ ์งํ›„ ํ™๋”๋ฏธ์— ๊ฑธ๋ ค ๋ฐ”ํ€ด๊ฐ€ ํ—›๋Œ๋‹ค๊ฐ€ ํƒ€์ด์–ด๊ฐ€ ๋…น์•„ ๊ฒ€์€ ์—ฐ๊ธฐ๋ฅผ ๋ฟœ์œผ๋ฉฐ ๋ฉˆ์ถฐ ์„ฐ์Šต๋‹ˆ๋‹ค.

๋Œ€๋ถ€๋ถ„์˜ ์ฐจ๋Ÿ‰์ด ์ฒ˜์ฐธํ•˜๊ฒŒ ์‹คํŒจํ–ˆ์Šต๋‹ˆ๋‹ค. ์ฐธ๊ฐ€์ž๋“ค์€ ๋กœ๋ด‡์ด 26๋งˆ์ผ ๋งˆ๋ผํ†ค์—์„œ ๊ฒจ์šฐ 2๋งˆ์ผ์„ ๋‹ฌ๋ฆฐ ๊ฒƒ์— ๋น„์œ ํ•˜๋ฉฐ ์‹ค๋ง๊ฐ์„ ๊ฐ์ถ”์ง€ ๋ชปํ–ˆ์Šต๋‹ˆ๋‹ค. ๊ทธ๋Ÿฌ๋‚˜ ์ด ๋Œ€ํšŒ๋Š” ์ž์œจ์ฃผํ–‰ ๊ธฐ์ˆ  ๊ฐœ๋ฐœ์„ ์œ„ํ•œ โ€˜์žฅโ€™์„ ๋งˆ๋ จํ–ˆ๋‹ค๋Š” ์ ์—์„œ ํฐ ์˜๋ฏธ๊ฐ€ ์žˆ์—ˆ์Šต๋‹ˆ๋‹ค. ๋‹น์‹œ ์ด ๊ด‘๊ฒฝ์„ ์ง€์ผœ๋ณด๋˜ ํ•œ ์ธ๋ฌผ์ด ์žˆ์—ˆ๋Š”๋ฐ, ๋ฐ”๋กœ ์Šคํƒ ํฌ๋“œ ๋Œ€ํ•™๊ต์˜ ์ „์„ค์ ์ธ ๋กœ๋ด‡ ๊ณตํ•™์ž **์„ธ๋ฐ”์Šค์ฐฌ ์Šค๋Ÿฐ(Sebastian Thrun)**์ด์—ˆ์Šต๋‹ˆ๋‹ค.

์Šค๋Ÿฐ์˜ ํ†ต์ฐฐ: ํ•˜๋“œ์›จ์–ด ์•„๋‹Œ ์†Œํ”„ํŠธ์›จ์–ด ๋ฌธ์ œ

์„ธ๋ฐ”์Šค์ฐฌ ์Šค๋Ÿฐ์€ ์ฒซ ๋ฒˆ์งธ ์ฑŒ๋ฆฐ์ง€๋ฅผ ๊ด€์ „ํ•˜๋ฉฐ ์ฐธ๊ฐ€ํŒ€๋“ค์ด ๊ทผ๋ณธ์ ์ธ ์˜ค๋ฅ˜๋ฅผ ๋ฒ”ํ–ˆ๋‹ค๊ณ  ์ƒ๊ฐํ–ˆ์Šต๋‹ˆ๋‹ค.

โ€œ๋ชจ๋“  ํŒ€์ด ์ด๊ฒƒ์„ ํ•˜๋“œ์›จ์–ด ๋ฌธ์ œ๋กœ ๋‹ค๋ฃจ๋Š” ๊ฒƒ์„ ๋ดค์Šต๋‹ˆ๋‹ค. โ€˜๋” ํฐ ๋ฐ”ํ€ด์™€ ๋” ํฐ ์ฐจ์ฒด๋ฅผ ๋งŒ๋“ค์–ด์•ผ ํ•ดโ€™๋ผ๊ณ  ์ƒ๊ฐํ–ˆ์ฃ . ํ•˜์ง€๋งŒ ์ €๋Š” โ€˜์ž ๊น๋งŒ, ์ด ์ฑŒ๋ฆฐ์ง€๋Š” ์‚ฌ๋ง‰์„ ์ฃผํ–‰ํ•  ์ˆ˜ ์žˆ๋Š” ์ž์œจ์ฃผํ–‰์ฐจ๋ฅผ ๋งŒ๋“œ๋Š” ๊ฑด๋ฐ, ์‚ฌ๋žŒ์ด ํƒ€๋ฉด ์ž˜ ๋‹ฌ๋ฆฌ๋Š” ๋ Œํ„ฐ์นด๋ฅผ ๋นŒ๋ฆด ์ˆ˜ ์žˆ์ž–์•„? ๋ฌธ์ œ๋Š” ์‚ฌ๋žŒ์„ ์šด์ „์„์—์„œ ๋นผ๊ณ  ์ปดํ“จํ„ฐ๋กœ ๋Œ€์ฒดํ•˜๋Š” ๊ฑฐ์•ผ. ์ด๊ฑด ๋” ํฐ ํƒ€์ด์–ด์˜ ๋ฌธ์ œ๊ฐ€ ์•„๋‹ˆ๋ผ ์†Œํ”„ํŠธ์›จ์–ด ๋ฌธ์ œ์•ผโ€™๋ผ๊ณ  ์ƒ๊ฐํ–ˆ์Šต๋‹ˆ๋‹ค.โ€

๋กœ๋ด‡ ๊ณตํ•™๊ณผ ์ธ๊ณต์ง€๋Šฅ(AI)์ด๋ผ๋Š” ์ด์ค‘ ๋ฐฐ๊ฒฝ์„ ๊ฐ€์ง„ ์Šค๋Ÿฐ์€ ๋กœ๋ด‡ ์šด์ „์ž์˜ โ€˜๋‘๋‡Œโ€™์— ์ง‘์ค‘ํ–ˆ์Šต๋‹ˆ๋‹ค. ๊ทธ๋Š” ๊ตฐ์šฉ ์ฐจ๋Ÿ‰์„ ๋„˜์–ด ์ „ ์„ธ๊ณ„ ๊ตํ†ต์‚ฌ๊ณ  ์‚ฌ๋ง์ž 100๋งŒ ๋ช…์„ ๊ตฌ์›ํ•  ์ˆ˜ ์žˆ๋Š” ๊ธฐ์ˆ ์„ ๊ฟˆ๊ฟจ์Šต๋‹ˆ๋‹ค.

2005๋…„ ๊ทธ๋žœ๋“œ ์ฑŒ๋ฆฐ์ง€: ๊ธฐ๊ณ„ ํ•™์Šต์˜ ์Šน๋ฆฌ

18๊ฐœ์›” ํ›„, DARPA๋Š” ์ƒ๊ธˆ์„ 200๋งŒ ๋‹ฌ๋Ÿฌ๋กœ ๋‘ ๋ฐฐ ์˜ฌ๋ ธ๊ณ , ๋‘ ๋ฒˆ์งธ ๊ทธ๋žœ๋“œ ์ฑŒ๋ฆฐ์ง€๊ฐ€ ์—ด๋ ธ์Šต๋‹ˆ๋‹ค. ์นด๋„ค๊ธฐ๋ฉœ๋ก  ํŒ€์˜ ํฌ๋ฆฌ์Šค ์—„์Šจ(Chris Urmson)์€ ๋‘ ๋Œ€์˜ ์ฐจ๋Ÿ‰์œผ๋กœ ๋‹ค์‹œ ์ฐธ๊ฐ€ํ–ˆ๊ณ , ์•ค์„œ๋‹ˆ ๋ฅด์œˆ๋„์Šคํ‚ค๋Š” ์—ฌ์ „ํžˆ ์ž‘๋™ํ•˜์ง€ ์•Š๋Š” ์˜คํ† ๋ฐ”์ด๋กœ ์˜ˆ์„ ์—์„œ ํƒˆ๋ฝํ–ˆ์Šต๋‹ˆ๋‹ค. ๊ทธ๋ฆฌ๊ณ  ์Šค๋Ÿฐ์˜ ์Šคํƒ ํฌ๋“œ ํŒ€์ด ๋“ฑ์žฅํ–ˆ์Šต๋‹ˆ๋‹ค.

์Šค๋Ÿฐ์˜ ์ฐจ๋Ÿ‰์€ ํญ์Šค๋ฐ”๊ฒ(Volkswagen)์ด ๊ธฐ์ฆํ•œ ํŒŒ๋ž€์ƒ‰ SUV โ€˜์Šคํƒ ๋ฆฌ(Stanley)โ€˜์˜€์Šต๋‹ˆ๋‹ค. ํ—˜๋น„ ๊ธฐ๋ฐ˜์˜ ์ƒŒ๋“œ์Šคํ†ฐ์— ๋น„ํ•˜๋ฉด ์™œ์†Œํ•ด ๋ณด์˜€์ง€๋งŒ, ์Šค๋Ÿฐ์€ โ€œ์Šคํƒ ๋ฆฌ๋Š” ์ง€๋Šฅ์ ์ธ ์˜์‚ฌ ๊ฒฐ์ •์žโ€๋ผ๊ณ  ์†Œ๊ฐœํ–ˆ์Šต๋‹ˆ๋‹ค. ์Šค๋Ÿฐ์€ ๋‹น์‹œ๋กœ์„œ๋Š” ์›์‹œ์ ์ด์—ˆ๋˜ ์ธ๊ณต์ง€๋Šฅ์„ ํ™œ์šฉํ•˜์—ฌ ์ฐจ๋Ÿ‰์ด ๋„๋กœ๋ฅผ ๋” ๋น ๋ฅด๊ณ  ์ •ํ™•ํ•˜๊ฒŒ ์ธ์‹ํ•˜๋„๋ก ํ›ˆ๋ จ์‹œ์ผฐ์Šต๋‹ˆ๋‹ค. ์Šคํƒ ํฌ๋“œ ๊ทผ์ฒ˜์˜ ๋น„ํฌ์žฅ๋„๋กœ์—์„œ ์ฐจ๋Ÿ‰ ์นด๋ฉ”๋ผ๋กœ ํ’๊ฒฝ์„ ๊ธฐ๋กํ•˜๊ณ , ์ฐจ๋Ÿ‰์ด ๊ทธ ์œ„๋ฅผ ์ง€๋‚  ๋•Œ๋งˆ๋‹ค ๋ฏธ๋„๋Ÿฌ์›€์ด๋‚˜ ์š”์ฒ ์„ ์ธก์ •ํ•˜์—ฌ โ€˜์ด ๋…น์ƒ‰ ํ’€๋ฐญ์€ ์šด์ „ํ•˜๊ธฐ ์ข‹์€ ๊ณณ(grass), ์ด ๊ฐˆ์ƒ‰ ์ง„ํ™์€ ์ข‹์ง€ ์•Š์€ ๊ณณ(mud)โ€˜์ด๋ผ๊ณ  ์Šค์Šค๋กœ ํ•™์Šตํ•˜๊ฒŒ ํ–ˆ์Šต๋‹ˆ๋‹ค. ์Šคํƒ ๋ฆฌ๋Š” ์ด๋ ‡๊ฒŒ ํ•™์Šตํ•œ ํŒจํ„ด์„ ์ดˆ๋‹น 30๋ฒˆ์”ฉ ๊ฐ์ง€ํ•˜๊ณ  ์ผ๋ฐ˜ํ™”ํ•  ์ˆ˜ ์žˆ์—ˆ์Šต๋‹ˆ๋‹ค.

2005๋…„ ๋‘ ๋ฒˆ์งธ ๋ ˆ์ด์Šค๋Š” โ€œ์ฒซ ๋ฒˆ์งธ ๋ ˆ์ด์Šค๊ฐ€ ์žฌ์•™์ ์ด์—ˆ๋˜ ๋งŒํผ ์„ฑ๊ณต์ (as successful as the first race is disastrous)โ€œ์ด์—ˆ์Šต๋‹ˆ๋‹ค. ๊ฑฐ์˜ ๋ชจ๋“  ์ฐธ๊ฐ€ ์ฐจ๋Ÿ‰์ด ์ฒซ ๋ฒˆ์งธ ๋ ˆ์ด์Šค์˜ ์ƒŒ๋“œ์Šคํ†ฐ๋ณด๋‹ค ๋” ๋ฉ€๋ฆฌ ๊ฐ”๊ณ , ์—ฌ๋Ÿฌ ๋Œ€๊ฐ€ ์ฝ”์Šค๋ฅผ ์™„์ฃผํ–ˆ์Šต๋‹ˆ๋‹ค. ์Šคํƒ ๋ฆฌ๋Š” ์นด๋„ค๊ธฐ๋ฉœ๋ก ์˜ ๋‘ ์ฐจ๋Ÿ‰ ์‚ฌ์ด์—์„œ ์ถœ๋ฐœํ–ˆ์ง€๋งŒ, ๊ฐ€์žฅ ๋จผ์ € ๊ฒฐ์Šน์„ ์„ ํ†ต๊ณผํ•˜๋ฉฐ ์šฐ์Šน์„ ์ฐจ์ง€ํ–ˆ์Šต๋‹ˆ๋‹ค. ์Šค๋Ÿฐ์€ โ€œ๋ฏฟ์„ ์ˆ˜ ์—†์„ ์ •๋„๋กœ ๋งˆ๋ฒ• ๊ฐ™์•˜๋‹คโ€๋ฉฐ, โ€œ์ด ์ปค๋ฎค๋‹ˆํ‹ฐ์˜ ์‚ฌ๋žŒ๋“ค์—๊ฒŒ ํ™˜์ƒ์ ์ธ ์„ฑ๊ณตโ€์ด๋ผ๊ณ  ๊ธฐ๋ปํ–ˆ์Šต๋‹ˆ๋‹ค.

๊ตฌ๊ธ€์˜ ๋น„๋ฐ€ ํ”„๋กœ์ ํŠธ: โ€˜ํ”„๋กœ์ ํŠธ ์‡ผํผโ€™์˜ ํƒ„์ƒ

์ด ์„ฑ๊ณต์˜ ์ˆœ๊ฐ„, ๊ด€์ค‘์„์—์„œ๋Š” ์•ผ๊ตฌ ๋ชจ์ž์™€ ์„ ๊ธ€๋ผ์Šค๋กœ ๋ณ€์žฅํ•œ ์ฑ„ ๊ฒฝ๊ธฐ๋ฅผ ์ง€์ผœ๋ณด๋˜ ํ•œ ๋‚จ์ž๊ฐ€ ์žˆ์—ˆ์Šต๋‹ˆ๋‹ค. ๋ฐ”๋กœ ๊ตฌ๊ธ€(Google)์˜ ๊ณต๋™ ์ฐฝ์—…์ž **๋ž˜๋ฆฌ ํŽ˜์ด์ง€(Larry Page)**์˜€์Šต๋‹ˆ๋‹ค. ๊ทธ๋Š” ์Šค๋Ÿฐ์—๊ฒŒ ์ ‘๊ทผํ•˜์—ฌ ๋ผ์ด๋‹ค(LiDAR) ์‹œ์Šคํ…œ์˜ ํŒŒ์žฅ ๊ฐ™์€ ์„ธ๋ถ€์ ์ธ ์งˆ๋ฌธ๋“ค์„ ์Ÿ์•„๋ƒˆ์Šต๋‹ˆ๋‹ค. ๋ž˜๋ฆฌ ํŽ˜์ด์ง€๋Š” ์ด๋ฏธ ๋Œ€ํ•™์› ์‹œ์ ˆ๋ถ€ํ„ฐ ์ž์œจ์ฃผํ–‰์ฐจ์— ๊ด€์‹ฌ์ด ์žˆ์—ˆ์ง€๋งŒ, ๋‹น์‹œ ์ง€๋„ ๊ต์ˆ˜์˜ ๊ถŒ์œ ๋กœ ๊ฒ€์ƒ‰ ์—”์ง„ ์—ฐ๊ตฌ๋กœ ๋ฐฉํ–ฅ์„ ํ‹€์—ˆ๋˜ ๊ณผ๊ฑฐ๊ฐ€ ์žˆ์—ˆ์Šต๋‹ˆ๋‹ค. DARPA ์ฑŒ๋ฆฐ์ง€๋ฅผ ํ†ตํ•ด ๊ทธ๋Š” ์ž์œจ์ฃผํ–‰์ฐจ๊ฐ€ ํ˜„์‹ค์ด ๋  ์ˆ˜ ์žˆ๋‹ค๋Š” ์ฆ๊ฑฐ๋ฅผ ๋ณด๊ฒŒ ๋œ ๊ฒƒ์ž…๋‹ˆ๋‹ค.

๋ž˜๋ฆฌ ํŽ˜์ด์ง€๋Š” ์ฒ˜์Œ์— ์Šค๋Ÿฐ๊ณผ ์•ค์„œ๋‹ˆ ๋ฅด์œˆ๋„์Šคํ‚ค๋ฅผ ๊ณ ์šฉํ•˜์—ฌ โ€˜๊ตฌ๊ธ€ ์ŠคํŠธ๋ฆฌํŠธ ๋ทฐ(Google Street View)โ€˜๋ฅผ ๊ตฌ์ถ•ํ•˜๊ฒŒ ํ–ˆ์Šต๋‹ˆ๋‹ค. ์Šคํƒ ๋ฆฌ๊ฐ€ ์‚ฌ์šฉํ–ˆ๋˜ ๋ฃจํ”„ ์žฅ์ฐฉ ์นด๋ฉ”๋ผ ์‹œ์Šคํ…œ์„ ๊ฐœ์กฐํ•˜์—ฌ ๋ฏธ๊ตญ์˜ ๊ฑฐ๋ฆฌ๋ฅผ ์ดฌ์˜ํ•˜๋Š” ์ž‘์—…์ด์—ˆ์Šต๋‹ˆ๋‹ค. ํ•˜์ง€๋งŒ ๊ณง ๋ž˜๋ฆฌ ํŽ˜์ด์ง€๋Š” ์Šค๋Ÿฐ์—๊ฒŒ ์ž์œจ์ฃผํ–‰์ฐจ์˜ ๊ฟˆ์„ ๋‹ค์‹œ ์ด์•ผ๊ธฐํ–ˆ์Šต๋‹ˆ๋‹ค.

์ „๋ฌธ๊ฐ€์˜ ๋ฐ˜๋Œ€, ๊ทธ๋ฆฌ๊ณ  ๋ž˜๋ฆฌ ํŽ˜์ด์ง€์˜ ๋ˆ๊ธฐ

2009๋…„, ๋ž˜๋ฆฌ ํŽ˜์ด์ง€๋Š” ์Šค๋Ÿฐ์—๊ฒŒ โ€œ์ „ ์„ธ๊ณ„ ์–ด๋””๋“  ์šด์ „ํ•  ์ˆ˜ ์žˆ๋Š” ์ž์œจ์ฃผํ–‰์ฐจ๋ฅผ ๋งŒ๋“ค์–ด์•ผ ํ•œ๋‹คโ€๊ณ  ๋งํ–ˆ์Šต๋‹ˆ๋‹ค. ์Šค๋Ÿฐ์˜ ์ฆ‰๊ฐ์ ์ธ ๋ฐ˜์‘์€ โ€œ์•ˆ ๋ฉ๋‹ˆ๋‹ค. ์‚ฌ๋ง‰์—์„œ ๋งŒ๋“  ๊ธฐ์ˆ ์„ ์ƒŒํ”„๋ž€์‹œ์Šค์ฝ”์˜ ๋งˆ์ผ“ ์ŠคํŠธ๋ฆฌํŠธ ํ•œ๊ฐ€์šด๋ฐ์— ๊ฐ€์ ธ๋‹ค ๋†“์œผ๋ฉด ๋ˆ„๊ตฐ๊ฐ€๊ฐ€ ์ฃฝ์„ ๊ฒ๋‹ˆ๋‹คโ€์˜€์Šต๋‹ˆ๋‹ค. ๋ž˜๋ฆฌ ํŽ˜์ด์ง€๋Š” ๋‹ค์Œ ๋‚ ๋„ ๊ฐ™์€ ์•„์ด๋””์–ด๋ฅผ ๋“ค๊ณ  ์™”๊ณ , ์Šค๋Ÿฐ์€ ๊ฐ™์€ ๋Œ€๋‹ต์„ ๋ฐ˜๋ณตํ–ˆ์Šต๋‹ˆ๋‹ค.

๊ฒฐ๊ตญ ๋ž˜๋ฆฌ ํŽ˜์ด์ง€๋Š” โ€œ์ข‹์•„, ์„ธ๋ฐ”์Šค์ฐฌ. ๋‹น์‹ ์ด ํ•  ์ˆ˜ ์—†๋‹ค๋Š” ๊ฑด ์•Œ๊ฒ ์–ด. ๊ทธ๋Ÿผ ์—๋ฆญ ์Šˆ๋ฏธํŠธ(Eric Schmidt) ๋‹น์‹œ CEO์™€ ๊ณต๋™ ์ฐฝ์—…์ž ์„ธ๋ฅด๊ฒŒ์ด ๋ธŒ๋ฆฐ(Sergey Brin)์—๊ฒŒ ์™œ ๋ถˆ๊ฐ€๋Šฅํ•œ์ง€ ๊ธฐ์ˆ ์ ์ธ ์ด์œ ๋ฅผ ์„ค๋ช…ํ•ด ์ค„ ์ˆ˜ ์žˆ๊ฒ ๋‚˜?โ€๋ผ๊ณ  ๋ฌผ์—ˆ์Šต๋‹ˆ๋‹ค. ์Šค๋Ÿฐ์€ ์ง‘์— ๋Œ์•„๊ฐ€ ์•„๋ฌด๋ฆฌ ์ƒ๊ฐํ•ด๋„ ๊ธฐ์ˆ ์ ์ธ ์ด์œ ๋ฅผ ์ฐพ์„ ์ˆ˜ ์—†์—ˆ์Šต๋‹ˆ๋‹ค.

โ€œ์ €๋Š” ์ž์œจ์ฃผํ–‰์ฐจ ๋ถ„์•ผ์˜ ์„ธ๊ณ„ ์ตœ๊ณ  ์ „๋ฌธ๊ฐ€์ธ๋ฐ, ์ด๊ฒŒ ๋ถˆ๊ฐ€๋Šฅํ•˜๋‹ค๊ณ  ๋ถ€์ •ํ•˜๋Š” ์‚ฌ๋žŒ์ด์—ˆ์ฃ . ๊ทธ ์ˆœ๊ฐ„ โ€˜์ „๋ฌธ๊ฐ€๋Š” ๋Œ€๊ฐœ ๊ณผ๊ฑฐ์˜ ์ „๋ฌธ๊ฐ€์ด์ง€ ๋ฏธ๋ž˜์˜ ์ „๋ฌธ๊ฐ€๋Š” ์•„๋‹ˆ๋‹ค. ํ˜์‹ ์ด๋‚˜ ๋ฏธ์นœ ๋“ฏ์ด ์ƒˆ๋กœ์šด ๊ฒƒ์— ๋Œ€ํ•ด ์ „๋ฌธ๊ฐ€์—๊ฒŒ ๋ฌผ์œผ๋ฉด, ๊ทธ๋“ค์€ โ€˜๋„ค, ํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹คโ€™๋ผ๊ณ  ๋งํ•  ๊ฐ€๋Šฅ์„ฑ์ด ๊ฐ€์žฅ ๋‚ฎ์€ ์‚ฌ๋žŒ์ด๋‹คโ€™๋ผ๋Š” ์—„์ฒญ๋‚˜๊ฒŒ ์ค‘์š”ํ•œ ๊ตํ›ˆ์„ ์–ป์—ˆ์Šต๋‹ˆ๋‹ค.โ€

์ด๊ฒƒ์ด ๋ฐ”๋กœ 2009๋…„ ๊ตฌ๊ธ€์˜ ๋น„๋ฐ€ ํ”„๋กœ์ ํŠธ, **โ€˜ํ”„๋กœ์ ํŠธ ์‡ผํผ(Project Chauffeur)โ€˜**์˜ ์‹œ์ž‘์ด์—ˆ์Šต๋‹ˆ๋‹ค. ์Šค๋Ÿฐ์ด ์ด๋Œ๊ณ , DARPA ์ฑŒ๋ฆฐ์ง€ ์ถœ์‹ ์ธ ํฌ๋ฆฌ์Šค ์—„์Šจ(Chris Urmson)์ด ์ผ์ƒ ์—…๋ฌด๋ฅผ ์ด๊ด„ํ•˜๋ฉฐ, ์•ค์„œ๋‹ˆ ๋ฅด์œˆ๋„์Šคํ‚ค๋Š” ํ•˜๋“œ์›จ์–ด, ๋“œ๋ฏธํŠธ๋ฆฌ ๋Œ๊ณ ํ”„(Dmitri Dolgov)๋Š” ๊ณ„ํš ๋ฐ ์ตœ์ ํ™”๋ฅผ ๋‹ด๋‹นํ–ˆ์Šต๋‹ˆ๋‹ค. ๋ž˜๋ฆฌ ํŽ˜์ด์ง€์—๊ฒŒ ์ง์ ‘ ๋ณด๊ณ ํ•˜๋Š” ์ด ์†Œ๊ทœ๋ชจ ํŒ€์€ 11๋ช…์˜ ์—”์ง€๋‹ˆ์–ด๋กœ ๊ตฌ์„ฑ๋˜์–ด ๊ด€๋ฃŒ์ฃผ์˜ ์—†์ด ๋น ๋ฅด๊ฒŒ ์›€์ง์˜€์Šต๋‹ˆ๋‹ค.

๋ž˜๋ฆฌ 1K ์ฑŒ๋ฆฐ์ง€: ์ธ๊ฐ„ ์šด์ „์˜ ๋ฏธ๋ฌ˜ํ•จ์„ ๊ธฐ๊ณ„์—๊ฒŒ ๊ฐ€๋ฅด์น˜๋‹ค

๊ตฌ๊ธ€์€ ํ”„๋กœ์ ํŠธ ์‡ผํผ ํŒ€์—๊ฒŒ ๋‘ ๊ฐ€์ง€ ๋ชฉํ‘œ๋ฅผ ์ œ์‹œํ–ˆ์Šต๋‹ˆ๋‹ค. ์ฒซ์งธ, ๊ณต๊ณต๋„๋กœ์—์„œ 10๋งŒ ๋งˆ์ผ์„ ์•ˆ์ „ํ•˜๊ฒŒ ์ฃผํ–‰ํ•˜๋Š” ๊ฒƒ. ๋‘˜์งธ, **โ€˜๋ž˜๋ฆฌ 1K ์ฑŒ๋ฆฐ์ง€(Larry 1K Challenge)โ€˜**๋ฅผ ์™„์ˆ˜ํ•˜๋Š” ๊ฒƒ. ๋ž˜๋ฆฌ ํŽ˜์ด์ง€์™€ ์„ธ๋ฅด๊ฒŒ์ด ๋ธŒ๋ฆฐ์€ ๊ตฌ๊ธ€ ๋งต์Šค๋ฅผ ์—ด๊ณ  ์บ˜๋ฆฌํฌ๋‹ˆ์•„ ๋‚ด์—์„œ ๊ฐ€์žฅ ๊นŒ๋‹ค๋กœ์šด 100๋งˆ์ผ ์ฝ”์Šค 10๊ฐœ๋ฅผ ์ง์ ‘ ์„ ์ •ํ–ˆ์Šต๋‹ˆ๋‹ค. ๋ฒ ์ด ๋ธŒ๋ฆฟ์ง€, ํƒ€ํ˜ธ ํ˜ธ์ˆ˜, ๋กฌ๋ฐ”๋“œ ์ŠคํŠธ๋ฆฌํŠธ(Lombard Street) ๋“ฑ ๋ชจ๋“  ์ข…๋ฅ˜์˜ ์–ด๋ ค์šด ๋„๋กœ๊ฐ€ ํฌํ•จ๋˜์—ˆ์Šต๋‹ˆ๋‹ค. ํŒ€์€ ์ด 10๊ฐœ ์ฝ”์Šค๋ฅผ ๋‹จ ํ•œ ๋ฒˆ์˜ ์ธ๊ฐ„ ๊ฐœ์ž… ์—†์ด ์™„์ฃผํ•ด์•ผ ํ–ˆ์Šต๋‹ˆ๋‹ค.

ํŒ€์€ ์Šคํƒ ํฌ๋“œ์˜ DARPA ์–ด๋ฐ˜ ์ฑŒ๋ฆฐ์ง€(Urban Challenge) ์ฐจ๋Ÿ‰ ์ฝ”๋“œ๋ฅผ ๋ผ์ด์„ ์Šค ๋ฐ›์•„ ํ† ์š”ํƒ€ ํ”„๋ฆฌ์šฐ์Šค(Prius) 8๋Œ€๋ฅผ ๊ฐœ์กฐํ–ˆ์Šต๋‹ˆ๋‹ค. ๋ฒ”ํผ ๋’ค์˜ ๋ ˆ์ด๋” ์‹œ์Šคํ…œ, ์นด๋ฉ”๋ผ, ๊ทธ๋ฆฌ๊ณ  360๋„๋กœ ํšŒ์ „ํ•˜๋Š” ๋ผ์ด๋‹ค(LiDAR: ๋ ˆ์ด์ €๋ฅผ ์˜์•„ ๊ฑฐ๋ฆฌ๋ฅผ ์ธก์ •ํ•˜๋Š” ์‹œ์Šคํ…œ)๋ฅผ ์žฅ์ฐฉํ–ˆ์Šต๋‹ˆ๋‹ค.

์ดˆ๊ธฐ์—๋Š” ์ฐจ๋Ÿ‰๋งˆ๋‹ค โ€˜๋‚˜์ดํŠธ ๋ผ์ด๋”(Night Rider)โ€™ ๊ฐ™์€ ๋ฉ‹์ง„ ์ด๋ฆ„์„ ๋ถ™์˜€์ง€๋งŒ, ๊ณง โ€œ์ด ๋ชจ๋“  ์ฐจ๋Ÿ‰์— ์ด๋ฆ„์„ ๋ถ™์ผ ์ˆ˜๋Š” ์—†์„ ๊ฒƒโ€์ด๋ผ๋Š” ๊ฒƒ์„ ๊นจ๋‹ซ๊ณ  ํ”„๋ฆฌ์šฐ์Šค 27ํ˜ธ์ฒ˜๋Ÿผ ๋ฒˆํ˜ธ๋ฅผ ๋ถ™์ด๊ธฐ ์‹œ์ž‘ํ–ˆ์Šต๋‹ˆ๋‹ค. ์ž์œจ ์ž ์ˆ˜ํ•จ ์—ฐ๊ตฌ์›์ด์—ˆ๋‹ค๊ฐ€ ๊ตํ†ต์‚ฌ๊ณ ๋กœ ์นœ๊ตฌ๋ฅผ ์žƒ๊ณ  ์ž์‹ ๋„ ์‚ฌ๊ณ ๋ฅผ ๊ฒช์€ ํ›„ ์ž์œจ์ฃผํ–‰์ฐจ ์—ฐ๊ตฌ์— ๋›ฐ์–ด๋“  ๋ˆ ๋ฒ„๋„ท(Don Burnette)์€ ๋ชจ์…˜ ํ”Œ๋ž˜๋‹ ๋ฐ ํ–‰๋™ ๊ฒฐ์ • ํŒ€์—์„œ โ€˜๋„›์ง•(nudging)โ€™ ํ–‰๋™์„ ์—ฐ๊ตฌํ–ˆ์Šต๋‹ˆ๋‹ค.

๋„›์ง•(Nudging)๊ณผ ๋งฅ๋ฝ์˜ ์ค‘์š”์„ฑ

๋„›์ง•์€ ๋Œ€ํ˜• ํŠธ๋Ÿญ์ด ์˜†์„ ์ง€๋‚˜๊ฐˆ ๋•Œ ์ธ๊ฐ„ ์šด์ „์ž๊ฐ€ ๋ฌด์˜์‹์ ์œผ๋กœ ์™ผ์ชฝ์œผ๋กœ ์‚ด์ง ์›€์ง์ด๋Š” ๊ฒƒ๊ณผ ๊ฐ™์€ ๋ณธ๋Šฅ์ ์ธ ํ–‰๋™์ž…๋‹ˆ๋‹ค. ๋ˆ ๋ฒ„๋„ท์˜ ์ž„๋ฌด๋Š” ์ปดํ“จํ„ฐ์—๊ฒŒ ์ด๋Ÿฌํ•œ ๋ฏธ๋ฌ˜ํ•œ ํ–‰๋™์„ ๊ฐ€๋ฅด์น˜๋Š” ๊ฒƒ์ด์—ˆ์Šต๋‹ˆ๋‹ค.

โ€œ์šด์ „์ž๊ฐ€ ๋ถ€๋ถ„์ ์œผ๋กœ ์ข‹์€ ์ธ์ง€ ์ƒํƒœ์—์„œ ์‚ฌ์šฉํ•˜๋Š” ํ–‰๋™์„ ์ฝ”๋”ฉํ•˜๋ ค๊ณ  ๋…ธ๋ ฅํ•˜๋Š” ๊ฑฐ์ฃ . ์ •๋ง ๊นŒ๋‹ค๋กœ์šด ๋ฌธ์ œ์ž…๋‹ˆ๋‹ค.โ€

๋ˆ ๋ฒ„๋„ท์€ ์ธ๊ฐ„์ด ์šด์ „ํ•˜๋Š” ๋ฐฉ์‹์— ๋Œ€ํ•ด ๊นŠ์ด ๊ณ ๋ฏผํ–ˆ์Šต๋‹ˆ๋‹ค. โ€œ์™œ ์ธ๊ฐ„์€ ๊ทธ๋ ‡๊ฒŒ ์šด์ „ํ•˜๋Š”๊ฐ€?โ€๋ผ๋Š” ์งˆ๋ฌธ์— ๋ช…ํ™•ํ•œ ๋‹ต์„ ์ฐพ๊ธฐ ์–ด๋ ค์› ๊ณ , ๊ฒฐ๊ตญ ๊ธฐ๊ณ„ ํ•™์Šต(Machine Learning)์„ ํ†ตํ•ด ์ธ๊ฐ„ ํ–‰๋™์˜ ๊นŠ์€ ์ง„์‹ค์„ ์ถ”๋ก ํ•ด์•ผ ํ–ˆ์Šต๋‹ˆ๋‹ค.

์˜ˆ๋ฅผ ๋“ค์–ด, ๊ณ ์†๋„๋กœ ์ง„์ž… ๋žจํ”„์—์„œ ์Šน๊ฐ์ด ํŽธ์•ˆํ•จ์„ ๋А๋ผ๋Š” ์ ์ ˆํ•œ ์†๋„์™€ ๊ฐ๋„๋ฅผ ์ฐพ๋Š”๋‹ค๊ณ  ๊ฐ€์ •ํ•ด ๋ด…์‹œ๋‹ค. ์ธก๋ฉด ๊ฐ€์†๋„(lateral acceleration)๊ฐ€ ์ดˆ๋‹น 2๋ฏธํ„ฐ ์ œ๊ณฑ(2m/sยฒ)์ผ ๋•Œ ํŽธ์•ˆํ•จ์„ ๋А๋‚€๋‹ค๋Š” ๊ณ„์‚ฐ์ด ๋‚˜์˜ต๋‹ˆ๋‹ค. ํ•˜์ง€๋งŒ ๋†€๋ž๊ฒŒ๋„ ์ด ์ˆ˜์น˜๋Š” ์ง„์ž… ๋žจํ”„์—๋งŒ ํ•ด๋‹น๋ฉ๋‹ˆ๋‹ค. ๋™๋„ค ๋ง‰๋‹ค๋ฅธ ๊ณจ๋ชฉ์—์„œ ์œ ํ„ด์„ ํ•  ๋•Œ, ํ›จ์”ฌ ๋А๋ฆฐ ์†๋„์ž„์—๋„ ๋ถˆ๊ตฌํ•˜๊ณ  2m/sยฒ์˜ ์ธก๋ฉด ๊ฐ€์†๋„๋ฅผ ๊ฒฝํ—˜ํ•œ๋‹ค๋ฉด ์šด์ „์ž์—๊ฒŒ โ€œ๋ฏธ์ณค๋ƒโ€๊ณ  ๋งํ•  ๊ฒƒ์ž…๋‹ˆ๋‹ค. ๋งˆ์น˜ โ€˜๋งˆ๋ฆฌ์˜ค ์นดํŠธ(Mario Kart)โ€˜๋ฅผ ํ•˜๋Š” ๊ธฐ๋ถ„์ผ ๊ฒƒ์ž…๋‹ˆ๋‹ค. ๋ฌผ๋ฆฌ์ ์ธ ํž˜์€ ๊ฐ™์ง€๋งŒ, ๊ณ ์†๋„๋กœ ์ง„์ž…์ด๋ผ๋Š” ๋งฅ๋ฝ๊ณผ ์ฃผํƒ๊ฐ€ ์œ ํ„ด์ด๋ผ๋Š” ๋งฅ๋ฝ์ด ๋‡Œ๋ฅผ ์†์—ฌ ์ƒํ™ฉ์— ๋Œ€ํ•ด ์ •๋ฐ˜๋Œ€์˜ ๊ฐ์ •์„ ๋А๋ผ๊ฒŒ ํ•˜๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค. ๋ง‰๋‹ค๋ฅธ ๊ณจ๋ชฉ์—์„œ์˜ ํ•œ๊ณ„๋Š” ์•ฝ 0.75m/sยฒ๋กœ, ๊ณ ์†๋„๋กœ ์ง„์ž… ์‹œ๋ณด๋‹ค ๊ฑฐ์˜ 3๋ฐฐ๋‚˜ ๋‚ฎ์Šต๋‹ˆ๋‹ค.

์ด์ฒ˜๋Ÿผ ์ธ๊ฐ„์˜ ์šด์ „์€ ๋‹จ์ˆœํ•œ ๋ฌผ๋ฆฌ์  ์ œํ•œ์ด ์•„๋‹ˆ๋ผ โ€˜๋งฅ๋ฝ(context)โ€˜์— ๋”ฐ๋ผ ๋‹ฌ๋ผ์ง€๋Š” ๋ฏธ๋ฌ˜ํ•œ ๋ณตํ•ฉ์ฒด์˜€์Šต๋‹ˆ๋‹ค. ๋ˆ ๋ฒ„๋„ท์€ ์ด๋Ÿฌํ•œ โ€˜์Šน๊ฐ์„ ํŽธ์•ˆํ•˜๊ฒŒ ๋งŒ๋“œ๋Š” ๋ฐฉ๋ฒ•โ€™๊ณผ ๊ฐ™์€ ๋ฌธ์ œ๋“ค์„ ํ•ด๊ฒฐํ•˜๋Š” ๋ฐ ์‹œ๊ฐ„์„ ๋ณด๋ƒˆ์Šต๋‹ˆ๋‹ค.

์ˆจ๊ฒจ์ง„ ์‹œํ—˜ ์ฃผํ–‰๊ณผ ์„ฑ๊ณต

2009๋…„, ๊ตฌ๊ธ€ ํŒ€์€ ๊ตฌ๊ธ€ ์‚ฌ๋ฌด์‹ค ๊ทผ์ฒ˜์˜ โ€˜์‡ผ์–ด๋ผ์ธ ์›ํ˜•๊ทน์žฅ(Shoreline Amphitheater)โ€™ ์ฃผ์ฐจ์žฅ์—์„œ ์ฒซ ์‹œํ—˜ ์ฃผํ–‰์„ ์‹œ์ž‘ํ–ˆ์Šต๋‹ˆ๋‹ค. ์ฝ˜์„œํŠธ๊ฐ€ ์—†๋Š” ๋‚ , ์•„๋ฌด๋„ ๋ชจ๋ฅด๊ฒŒ ์ž์œจ์ฃผํ–‰์ฐจ๋ฅผ ํ…Œ์ŠคํŠธํ–ˆ์Šต๋‹ˆ๋‹ค. ์„ผ์„œ๊ฐ€ ๋‹ค๋ฅธ ์ฐจ๋ฅผ ์ธ์‹ํ•˜๋Š”์ง€, ์ปดํ“จํ„ฐ๊ฐ€ ๋ช…๋ น์„ ์ค€์ˆ˜ํ•˜๋Š”์ง€ ๊ธฐ๋ณธ์ ์ธ ๊ธฐ๋Šฅ์„ ํ™•์ธํ–ˆ์Šต๋‹ˆ๋‹ค.

2009๋…„ ๋ด„, ํŒ€์€ ์ฒ˜์Œ์œผ๋กœ ์‹ค์ œ ๋„๋กœ ์ฃผํ–‰์— ๋‚˜์„ฐ์Šต๋‹ˆ๋‹ค. ํฌ๋ฆฌ์Šค ์—„์Šจ์ด ํ”„๋ฆฌ์šฐ์Šค๋ฅผ ํƒ€๊ณ  ์„ผํŠธ๋Ÿด ์ต์Šคํ”„๋ ˆ์Šค์›จ์ด(Central Expressway)๋กœ ๋‚˜๊ฐ”์„ ๋•Œ, ๋ฌธ์ œ๊ฐ€ ๋ฐœ์ƒํ–ˆ์Šต๋‹ˆ๋‹ค. ์ฐจ๋Ÿ‰์ด โ€œ์ˆ  ์ทจํ•œ ์„ ์›์ฒ˜๋Ÿผ(like a drunken sailor)โ€ ์‹ฌํ•˜๊ฒŒ ํ”๋“ค๋ ธ๋˜ ๊ฒƒ์ž…๋‹ˆ๋‹ค. ํ™œ์ฃผ๋กœ์—์„œ๋Š” ๋ˆˆ์น˜์ฑ„์ง€ ๋ชปํ–ˆ๋˜ 1~2ํ”ผํŠธ์˜ ์ขŒ์šฐ ํ”๋“ค๋ฆผ์ด ์‹ค์ œ ๋„๋กœ์—์„œ๋Š” ์‹ฌ๊ฐํ•œ ๋ฌธ์ œ๋กœ ๋“œ๋Ÿฌ๋‚ฌ์Šต๋‹ˆ๋‹ค.

ํ•˜์ง€๋งŒ ๊ตฌ๊ธ€ ํŒ€์€ ๋ˆ๊ธฐ ์žˆ๊ฒŒ ๋ฌธ์ œ๋ฅผ ํ•ด๊ฒฐํ•ด ๋‚˜๊ฐ”์Šต๋‹ˆ๋‹ค. ์•ˆ์ „ ์šด์ „์ž๊ฐ€ ์šด์ „๋Œ€ ๋’ค์— ์•‰์•„ ๋งŒ์ผ์˜ ์‚ฌํƒœ์— ๋Œ€๋น„ํ•˜๊ณ , ์˜†์ž๋ฆฌ ๋™๋ฃŒ๋Š” ๋“œ๋ฏธํŠธ๋ฆฌ ๋Œ๊ณ ํ”„๊ฐ€ ์„ค๊ณ„ํ•œ ๊ทธ๋ž˜ํ”ฝ ์ธํ„ฐํŽ˜์ด์Šค๋ฅผ ๋ณด๋ฉฐ ์„ผ์„œ์™€ ์‹ค์ œ ๋„๋กœ ์ƒํ™ฉ์˜ ๋ถˆ์ผ์น˜๋ฅผ ์™ธ์ณค์Šต๋‹ˆ๋‹ค. ๋‘ ๋ช…์˜ ํŒ€์›์ด ์ฐจ๋Ÿ‰์— ํƒ‘์Šนํ•˜์—ฌ ์˜ค๋ฅ˜๋ฅผ ๊ธฐ๋กํ•˜๊ณ , ์‚ฌ๋ฌด์‹ค๋กœ ๋Œ์•„๊ฐ€ ๋ฌธ์ œ๋ฅผ ํ•ด๊ฒฐํ•˜๊ณ  ์ฝ”๋“œ๋ฅผ ์—…๋ฐ์ดํŠธํ•˜๋Š” ๊ณผ์ •์„ ๋ฐ˜๋ณตํ–ˆ์Šต๋‹ˆ๋‹ค.

์ด๋Ÿฌํ•œ ๋…ธ๋ ฅ ๋•๋ถ„์— ํŒ€์€ ๋น ๋ฅด๊ฒŒ ๋ฐœ์ „ํ–ˆ์Šต๋‹ˆ๋‹ค. ๋ž˜๋ฆฌ 1K ์ฑŒ๋ฆฐ์ง€๋Š” ๋น„๋””์˜ค ๊ฒŒ์ž„์ฒ˜๋Ÿผ ์—ฌ๋Ÿฌ ๋ฒˆ ์‹œ๋„ํ•˜์—ฌ ์ธ๊ฐ„ ๊ฐœ์ž… ์—†์ด ์™„์ฃผํ•  ๋•Œ๊นŒ์ง€ ๋ฐ˜๋ณตํ•  ์ˆ˜ ์žˆ์—ˆ์Šต๋‹ˆ๋‹ค. 2010๋…„, ํŒ€์€ ์˜ˆ์ƒ๋ณด๋‹ค ๊ฑฐ์˜ ๋‘ ๋ฐฐ ๋น ๋ฅธ 1๋…„์—ฌ ๋งŒ์— ๋ž˜๋ฆฌ 1K ์ฑŒ๋ฆฐ์ง€๋ฅผ ๋ชจ๋‘ ์™„์ˆ˜ํ–ˆ์Šต๋‹ˆ๋‹ค. ์ƒดํŽ˜์ธ์„ ํ„ฐ๋œจ๋ฆฌ๊ณ  ์„œ๋กœ๋ฅผ ์ˆ˜์˜์žฅ์— ๋˜์ง€๋ฉฐ ์ถ•ํ•˜ํ–ˆ์ง€๋งŒ, ๋™์‹œ์— โ€œ์ž, ์ด์ œ ๋ญ˜ ํ•˜์ง€?โ€๋ผ๋Š” ์งˆ๋ฌธ์— ์ง๋ฉดํ–ˆ์Šต๋‹ˆ๋‹ค.

๋ฏธ๋ž˜๋ฅผ ํ–ฅํ•œ ๋‹ค์Œ ๋‹จ๊ณ„: ๊ฒฝ์Ÿ๊ณผ ๋ถ„์—ด์˜ ์‹œ์ž‘

1๋…„ ๋งŒ์— ์ธ๊ฐ„์˜ ๊ฐ๋…๊ณผ ์ฝ”๋”ฉ์ด ํ•„์š”ํ–ˆ์ง€๋งŒ, ๋งค์šฐ ๊นŒ๋‹ค๋กœ์šด ์บ˜๋ฆฌํฌ๋‹ˆ์•„ ๋„๋กœ๋ฅผ ์„ฑ๊ณต์ ์œผ๋กœ ์ฃผํ–‰ํ•˜๋Š” ์ž์œจ์ฃผํ–‰์ฐจ๋ฅผ ๋งŒ๋“ค์–ด๋‚ธ ๊ฒƒ์€ ๊ธฐ์ ์— ๊ฐ€๊นŒ์› ์Šต๋‹ˆ๋‹ค. ๊ทธ๋“ค์€ ์•ˆ์ „ํ•˜๊ฒŒ, ๊ทธ๋ฆฌ๊ณ  ๋น ๋ฅด๊ฒŒ ํ•ด๋ƒˆ์Šต๋‹ˆ๋‹ค. ๊ทธ๋Ÿฌ๋‚˜ ์ด ์„ฑ๊ณต ์ดํ›„, ์ƒํ™ฉ์€ ํ”๋“ค๋ฆฌ๊ธฐ ์‹œ์ž‘ํ–ˆ์Šต๋‹ˆ๋‹ค. ๊ฒฝ์Ÿ์ด ์‹ฌํ™”๋˜๊ณ , ํŒ€ ๋‚ด๋ถ€์— ๋ถ„์—ด์ด ์ƒ๊ฒจ๋‚ฌ์œผ๋ฉฐ, ์‹ฌ์ง€์–ด ํŒ€์›์ด ๋„ˆ๋ฌด ๋А๋ฆฌ๋‹ค๊ณ  ์ƒ๊ฐํ•˜์—ฌ ๋…๋‹จ์ ์ธ ํ–‰๋™์„ ์ทจํ•˜๋Š” โ€˜๋ฐ˜๋ž€(mutiny)โ€˜๊นŒ์ง€ ๋ฐœ์ƒํ•˜๊ฒŒ ๋ฉ๋‹ˆ๋‹ค.

๊ตฌ๊ธ€ ๋‚ด๋ถ€์—์„œ๋Š” ์ž์œจ์ฃผํ–‰์ฐจ๊ฐ€ โ€˜์šด์ „ ๋ณด์กฐ ๊ธฐ์ˆ (assistive technology)โ€˜์ด ๋  ๊ฒƒ์ธ๊ฐ€, ์•„๋‹ˆ๋ฉด โ€˜ํŒŒ๊ดด์ ์ธ ๋Œ€์ฒด ๊ธฐ์ˆ (disruptive replacement technology)โ€˜์ด ๋  ๊ฒƒ์ธ๊ฐ€์— ๋Œ€ํ•œ ๊ฒฉ๋ ฌํ•œ ๋…ผ์Ÿ์ด ๋ฒŒ์–ด์กŒ์Šต๋‹ˆ๋‹ค. ํ…Œ์Šฌ๋ผ(Tesla)์ฒ˜๋Ÿผ ์ž์œจ์ฃผํ–‰์„ ์ž๋™์ฐจ์˜ ํ•œ ๊ธฐ๋Šฅ์œผ๋กœ ์ œ๊ณตํ•˜๋ฉฐ ์ธ๊ฐ„์˜ ๋ชจ๋‹ˆํ„ฐ๋ง์„ ํ•„์š”๋กœ ํ•  ๊ฒƒ์ธ๊ฐ€, ์•„๋‹ˆ๋ฉด ์ฐจ๋Ÿ‰์ด ์™„์ „ํžˆ ์Šค์Šค๋กœ ์šด์ „ํ•  ์ˆ˜ ์žˆ์„ ๋•Œ๊นŒ์ง€ ๊ธฐ๋‹ค๋ ค ๋กœ๋ณดํƒ์‹œ์™€ ๊ฐ™์€ ์„œ๋น„์Šค๋กœ ์ œ๊ณตํ•  ๊ฒƒ์ธ๊ฐ€์˜ ๊ธฐ๋กœ์— ์„  ๊ฒƒ์ž…๋‹ˆ๋‹ค. ์ด ๋…ผ์Ÿ์€ ์ž์œจ์ฃผํ–‰์ฐจ์˜ ๋ฏธ๋ž˜๋ฅผ ๊ฒฐ์ •์ง“๋Š” ์ค‘์š”ํ•œ ์ „ํ™˜์ ์ด ๋  ๊ฒƒ์ด๋ฉฐ, PJ ๋ณดํŠธ์˜ ๋‹ค์Œ ์—ํ”ผ์†Œ๋“œ์—์„œ ๊ทธ ์ด์•ผ๊ธฐ๊ฐ€ ๊ณ„์†๋ฉ๋‹ˆ๋‹ค.

์ž์œจ์ฃผํ–‰์ฐจ์˜ ์—ฌ์ •์€ ์‚ฌ๋ง‰์˜ ์‹คํŒจ์—์„œ ์‹œ์ž‘ํ•˜์—ฌ ๊ตฌ๊ธ€์˜ ๋น„๋ฐ€ ํ”„๋กœ์ ํŠธ๋ฅผ ๊ฑฐ์ณ ๋†€๋ผ์šด ์„ฑ๊ณต์„ ๊ฑฐ๋‘์—ˆ์Šต๋‹ˆ๋‹ค. ํ•˜์ง€๋งŒ ์ด ๊ธฐ์ˆ ์ด ๊ถ๊ทน์ ์œผ๋กœ ์–ด๋–ค ๋ชจ์Šต์œผ๋กœ ์šฐ๋ฆฌ ์‚ถ์— ์•ˆ์ฐฉํ• ์ง€๋Š” ์—ฌ์ „ํžˆ ๋ฏธ์ง€์ˆ˜์ž…๋‹ˆ๋‹ค. ์ธ๊ฐ„ ์šด์ „์ž๊ฐ€ ์ •๋ง ๊ตฌ์‹œ๋Œ€์˜ ์œ ๋ฌผ์ด ๋ ์ง€๋Š”, ์•ž์œผ๋กœ ํŽผ์ณ์งˆ ๊ธฐ์ˆ  ๋ฐœ์ „๊ณผ ์‚ฌํšŒ์  ํ•ฉ์˜์— ๋‹ฌ๋ ค ์žˆ์„ ๊ฒƒ์ž…๋‹ˆ๋‹ค.


โ€œAndrej Karpathy on Code Agents, AutoResearch, and the Loopy Era of AIโ€ โ€” No Priors: AI, Machine Learning, Tech, & Startups ๊ธฐ๋ฐ˜ ๊ธฐ์‚ฌ ์›๋ณธ ์˜์ƒ ๋ณด๊ธฐ

์•ˆ๋“œ๋ ˆ์ด ์นดํŒŒ์‹œ๊ฐ€ ๋งํ•˜๋Š” โ€˜AI ๊ด‘์ฆโ€™ ์‹œ๋Œ€: ์ฝ”๋”ฉ์„ ๋„˜์–ด์„  โ€˜์˜์ง€ ํ‘œํ˜„โ€™๊ณผ ์ž์œจ ์—์ด์ „ํŠธ์˜ ๋ฏธ๋ž˜

๋„์ž…: ์ฝ”๋“œ ์—†๋Š” ์ฝ”๋”ฉ์˜ ์‹œ์ž‘

์ตœ๊ทผ ๋ช‡ ๋‹ฌ๊ฐ„ ์ธ๊ณต์ง€๋Šฅ(AI) ๋ถ„์•ผ๋Š” ์ „๋ก€ ์—†๋Š” ์†๋„๋กœ ๋ฐœ์ „ํ•˜๋ฉฐ ์šฐ๋ฆฌ๋ฅผ โ€˜AI ๊ด‘์ฆ(AI Psychosis)โ€˜์˜ ์‹œ๋Œ€๋กœ ์ด๋Œ๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค. ์—”์ง€๋‹ˆ์–ด๋ง์˜ ๋ณธ์งˆ์ด ๊ทผ๋ณธ์ ์œผ๋กœ ๋ฐ”๋€Œ๊ณ  ์žˆ์œผ๋ฉฐ, ๊ฐœ๋ฐœ์ž๋“ค์€ ๋” ์ด์ƒ ์ง์ ‘ ์ฝ”๋“œ๋ฅผ ํƒ€์ดํ•‘ํ•˜๋Š” ๊ฒƒ์ด ์•„๋‹ˆ๋ผ, ์—์ด์ „ํŠธ(Agent)๋“ค์—๊ฒŒ ์ž์‹ ์˜ โ€˜์˜์ง€โ€™๋ฅผ ํ‘œํ˜„ํ•˜๋ฉฐ ์ž‘์—…์„ ์ง€์‹œํ•˜๋Š” ์ƒˆ๋กœ์šด ํŒจ๋Ÿฌ๋‹ค์ž„์— ์ ์‘ํ•˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค. ์•ค๋“œ๋ฅ˜ ์นดํŒŒ์‹œ(Andrej Karpathy)๋Š” ์ตœ๊ทผ ํŒŸ์บ์ŠคํŠธ โ€˜No Priorsโ€™์™€์˜ ์ธํ„ฐ๋ทฐ์—์„œ ์ด๋Ÿฌํ•œ ๋ณ€ํ™”์˜ ์ตœ์ „์„ ์— ์„œ ์žˆ๋Š” ์ž์‹ ์˜ ๊ฒฝํ—˜๊ณผ ํ†ต์ฐฐ์„ ๊ณต์œ ํ–ˆ์Šต๋‹ˆ๋‹ค. ๊ทธ์—๊ฒŒ โ€˜์ฝ”๋“œโ€™๋Š” ๋” ์ด์ƒ ์ ์ ˆํ•œ ๋™์‚ฌ๊ฐ€ ์•„๋‹™๋‹ˆ๋‹ค. ์ด์ œ๋Š” ํ•˜๋ฃจ 16์‹œ๊ฐ„ ๋™์•ˆ ์—์ด์ „ํŠธ๋“ค์—๊ฒŒ ์ž์‹ ์˜ ์˜์ง€๋ฅผ ํ‘œ๋ช…ํ•˜๋ฉฐ ์•„์ด๋””์–ด๋ฅผ ํ˜„์‹ค๋กœ ๊ตฌํ˜„ํ•˜๋Š” ์‹œ๋Œ€๊ฐ€ ๋„๋ž˜ํ–ˆ์Šต๋‹ˆ๋‹ค.

1. AI ๊ด‘์ฆ ์‹œ๋Œ€์˜ ๋„๋ž˜: โ€˜์ฝ”๋“œโ€™๊ฐ€ ์•„๋‹Œ โ€˜์˜์ง€ ํ‘œํ˜„โ€™

์นดํŒŒ์‹œ๋Š” ์ง€๋‚œ 12์›”์„ ๊ธฐ์ ์œผ๋กœ ์ž์‹ ์˜ ์ž‘์—… ๋ฐฉ์‹์— ํ˜๋ช…์ ์ธ ๋ณ€ํ™”๊ฐ€ ์ผ์–ด๋‚ฌ๋‹ค๊ณ  ํšŒ์ƒํ•ฉ๋‹ˆ๋‹ค. ์ด์ „์—๋Š” ์ฝ”๋”ฉ์˜ 80%๋ฅผ ์ง์ ‘ ํ•˜๊ณ  20%๋งŒ ์—์ด์ „ํŠธ์— ์œ„์ž„ํ–ˆ๋‹ค๋ฉด, ์ด์ œ๋Š” 20%๋งŒ ์ง์ ‘ ํ•˜๊ณ  80%๋ฅผ ์—์ด์ „ํŠธ์— ๋งก๊ธฐ๋Š” ์ˆ˜์ค€์„ ๋„˜์–ด์„ฐ๋‹ค๊ณ  ๋งํ•ฉ๋‹ˆ๋‹ค. ์‹ฌ์ง€์–ด ๊ทธ ์ดํ›„๋กœ๋Š” ์ง์ ‘ ์ฝ”๋“œ๋ฅผ ํ•œ ์ค„๋„ ํƒ€์ดํ•‘ํ•˜์ง€ ์•Š์•˜์„ ์ •๋„์ž…๋‹ˆ๋‹ค. ๊ทธ๋Š” ์ด๋Ÿฌํ•œ ๋ณ€ํ™”๊ฐ€ ์ผ๋ฐ˜์ธ๋“ค์—๊ฒŒ๋Š” ์ž˜ ์•Œ๋ ค์ ธ ์žˆ์ง€ ์•Š์ง€๋งŒ, ์†Œํ”„ํŠธ์›จ์–ด ์—”์ง€๋‹ˆ์–ด์˜ ์ผ์ƒ์ ์ธ ์—…๋ฌด ํ๋ฆ„์ด ์™„์ „ํžˆ ๋ฐ”๋€Œ์—ˆ๋‹ค๊ณ  ๊ฐ•์กฐํ•ฉ๋‹ˆ๋‹ค.

์ด๋Ÿฌํ•œ ๋ณ€ํ™”์˜ ํ•ต์‹ฌ์€ โ€˜๋ณ‘๋ชฉ ํ˜„์ƒ(Bottleneck)โ€˜์˜ ์ด๋™์ž…๋‹ˆ๋‹ค. ๊ณผ๊ฑฐ์—๋Š” ๊ฐœ๋ฐœ์ž์˜ ํƒ€์ดํ•‘ ์†๋„๋‚˜ ์ปดํ“จํŒ… ์ž์›(GPU ํ”Œ๋กญ์Šค)์ด ์ž‘์—…์˜ ํ•œ๊ณ„์˜€๋‹ค๋ฉด, ์ด์ œ๋Š” ์—์ด์ „ํŠธ์˜ โ€˜ํ† ํฐ ์ฒ˜๋ฆฌ๋Ÿ‰(Token Throughput)โ€˜์„ ์–ผ๋งˆ๋‚˜ ํšจ์œจ์ ์œผ๋กœ ํ™œ์šฉํ•˜๊ณ  ๊ด€๋ฆฌํ•  ์ˆ˜ ์žˆ๋Š”์ง€๊ฐ€ ์ค‘์š”ํ•ด์กŒ์Šต๋‹ˆ๋‹ค. ์นดํŒŒ์‹œ๋Š” ์ด๋ฅผ โ€œ์Šคํ‚ฌ ๋ฌธ์ œ(Skill Issue)โ€œ๋ผ๊ณ  ํ‘œํ˜„ํ•ฉ๋‹ˆ๋‹ค. ์ฆ‰, AI์˜ ๋Šฅ๋ ฅ ์ž์ฒด๊ฐ€ ๋ถ€์กฑํ•œ ๊ฒƒ์ด ์•„๋‹ˆ๋ผ, ์šฐ๋ฆฌ๊ฐ€ ์•„์ง ์ด ์ƒˆ๋กœ์šด ๋„๊ตฌ๋ฅผ ์ถฉ๋ถ„ํžˆ ํ™œ์šฉํ•  ๋ฐฉ๋ฒ•์„ ์ฐพ์ง€ ๋ชปํ–ˆ๋‹ค๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค. ๋ชจ๋“  ๊ฒƒ์ด โ€˜์Šคํ‚ฌ ๋ฌธ์ œโ€™๋กœ ๋А๊ปด์ง€๋Š” ์ด ์‹œ์ ์—์„œ, ๋” ๋งŽ์€ ์—์ด์ „ํŠธ๋ฅผ ๋™์‹œ์— ๊ฐ€๋™ํ•˜๊ณ , ๊ทธ๋“ค์—๊ฒŒ ์ ์ ˆํ•œ ์ง€์‹œ๋ฅผ ๋‚ด๋ฆฌ๋ฉฐ, ๊ทธ๋“ค์˜ ์ž‘์—…์„ ์ตœ์ ํ™”ํ•˜๋Š” ๋ฐฉ๋ฒ•์„ ๋ฐฐ์šฐ๋Š” ๊ฒƒ์ด ์ƒˆ๋กœ์šด ์ˆ™๋ จ์˜ ๊ธธ์ด ๋˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค.

2. ์—์ด์ „ํŠธ ์›Œํฌํ”Œ๋กœ์šฐ์˜ ์ง„ํ™”: ๊ฑฐ์‹œ์  ์•ก์…˜๊ณผ โ€˜ํด๋กœ(Claw)โ€™

ํ˜„์žฌ์˜ ์—์ด์ „ํŠธ ํ™œ์šฉ์€ ๋‹จ์ˆœํžˆ ํ•œ ๋ฒˆ์˜ ์„ธ์…˜์œผ๋กœ ๋๋‚˜๋Š” ๊ฒƒ์ด ์•„๋‹™๋‹ˆ๋‹ค. ํ”ผํ„ฐ ์Šคํƒ€์ธ๋ฒ„๊ทธ(Peter Steinberg)์™€ ๊ฐ™์€ ์„ ๊ตฌ์ž๋“ค์€ ์ด๋ฏธ ๋‹ค์ˆ˜์˜ ์—์ด์ „ํŠธ๋ฅผ ๋™์‹œ์— ํ™œ์šฉํ•˜๋ฉฐ โ€˜๊ฑฐ์‹œ์  ์•ก์…˜(Macro Actions)โ€™ ๋‹จ์œ„๋กœ ์ž‘์—…์„ ์ง€์‹œํ•˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค. ์˜ˆ๋ฅผ ๋“ค์–ด, ๋ชจ๋‹ˆํ„ฐ์— ์—ฌ๋Ÿฌ ๊ฐœ์˜ ์ฝ”๋ฑ์Šค(Codex) ์—์ด์ „ํŠธ๊ฐ€ ๋„์›Œ์ ธ ์žˆ๊ณ , ๊ฐ ์—์ด์ „ํŠธ๊ฐ€ ์„œ๋กœ ๋‹ค๋ฅธ ์ฝ”๋“œ ์ €์žฅ์†Œ(Repo)์—์„œ ์ƒˆ๋กœ์šด ๊ธฐ๋Šฅ ๊ตฌํ˜„, ์—ฐ๊ตฌ, ๊ณ„ํš ์ˆ˜๋ฆฝ ๋“ฑ ๋ณต์žกํ•œ ์ž‘์—…์„ ๋™์‹œ์— ์ˆ˜ํ–‰ํ•˜๋Š” ์‹์ž…๋‹ˆ๋‹ค. ๊ฐœ๋ฐœ์ž๋Š” ์ด ์—์ด์ „ํŠธ๋“ค ์‚ฌ์ด๋ฅผ ์˜ค๊ฐ€๋ฉฐ ์ž‘์—…์„ ํ• ๋‹นํ•˜๊ณ  ๊ฒฐ๊ณผ๋ฅผ ๊ฒ€ํ† ํ•˜๋Š” ์—ญํ• ์„ ํ•ฉ๋‹ˆ๋‹ค.

์ด๋Ÿฌํ•œ ๋‹ค์ค‘ ์—์ด์ „ํŠธ ์‹œ์Šคํ…œ์˜ ๋‹ค์Œ ๋‹จ๊ณ„๋Š” โ€˜ํด๋กœ(Claw)โ€˜์™€ ๊ฐ™์€ ์ง€์†์ ์ธ ์—”ํ‹ฐํ‹ฐ์ž…๋‹ˆ๋‹ค. ์นดํŒŒ์‹œ๊ฐ€ ๋งํ•˜๋Š” โ€˜ํด๋กœโ€™๋Š” ๋‹จ์ˆœํžˆ ์ƒํ˜ธ์ž‘์šฉํ•˜๋Š” ์—์ด์ „ํŠธ๋ฅผ ๋„˜์–ด, ์ž์ฒด์ ์ธ ์ƒŒ๋“œ๋ฐ•์Šค์™€ ๊ณ ๋„ํ™”๋œ ๋ฉ”๋ชจ๋ฆฌ ์‹œ์Šคํ…œ์„ ๊ฐ–์ถ”๊ณ  ์‚ฌ์šฉ์ž๊ฐ€ ์ง€์ผœ๋ณด์ง€ ์•Š์•„๋„ ์Šค์Šค๋กœ ๋ฐ˜๋ณต์ ์œผ๋กœ ์ž‘์—…์„ ์ˆ˜ํ–‰ํ•˜๋Š” ๋ ˆ์ด์–ด๋ฅผ ์˜๋ฏธํ•ฉ๋‹ˆ๋‹ค. ์˜คํ”ˆ ํด๋กœ(Open Claw)๊ฐ€ ๋Œ€ํ‘œ์ ์ธ ์˜ˆ์‹œ์ธ๋ฐ, ์ด๋Š” ๊ธฐ์กด ์—์ด์ „ํŠธ์˜ ์ปจํ…์ŠคํŠธ(Context) ์ œํ•œ์œผ๋กœ ์ธํ•œ ๋ฉ”๋ชจ๋ฆฌ ์••์ถ• ๋ฐฉ์‹๋ณด๋‹ค ํ›จ์”ฌ ์ •๊ตํ•œ ๊ธฐ์–ต ์‹œ์Šคํ…œ์„ ์ œ๊ณตํ•ฉ๋‹ˆ๋‹ค.

๋˜ํ•œ, ์—์ด์ „ํŠธ์˜ โ€˜์„ฑ๊ฒฉ(Personality)โ€™ ๋˜ํ•œ ์ค‘์š”ํ•œ ์š”์†Œ๋กœ ๋ถ€์ƒํ•˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค. ์นดํŒŒ์‹œ๋Š” ํด๋กœ๋“œ(Claude) ์—์ด์ „ํŠธ์˜ ๊ฒฝ์šฐ ๋™๋ฃŒ์ฒ˜๋Ÿผ ๋А๊ปด์ง€๋Š” ์ข‹์€ ์„ฑ๊ฒฉ์„ ๊ฐ€์ง€๊ณ  ์žˆ์œผ๋ฉฐ, ์‚ฌ์šฉ์ž์˜ ์•„์ด๋””์–ด๋ฅผ ์นญ์ฐฌํ•  ๋•Œ์กฐ์ฐจ ์‚ฌ์šฉ์ž๊ฐ€ ๊ทธ ์นญ์ฐฌ์„ ๋ฐ›์„ ๋งŒํ•œ ๊ฐ€์น˜๊ฐ€ ์žˆ๋‹ค๊ณ  ๋А๋ผ๋„๋ก ์„ฌ์„ธํ•˜๊ฒŒ ๋ฐ˜์‘ํ•œ๋‹ค๊ณ  ๋งํ•ฉ๋‹ˆ๋‹ค. ๋ฐ˜๋ฉด ์ฝ”๋ฑ์Šค(Codex)์™€ ๊ฐ™์€ ์—์ด์ „ํŠธ๋Š” ๊ฑด์กฐํ•˜๊ณ  ๋ฌด๊ด€์‹ฌํ•˜๊ฒŒ ๋А๊ปด์งˆ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค. ์ด๋Ÿฌํ•œ โ€˜์„ฑ๊ฒฉโ€™์€ ์‚ฌ์šฉ์ž๊ฐ€ ์—์ด์ „ํŠธ์™€ ์ƒํ˜ธ์ž‘์šฉํ•˜๊ณ  ๊ทธ๋“ค์˜ โ€˜์นญ์ฐฌโ€™์„ ์–ป๊ธฐ ์œ„ํ•ด ๋…ธ๋ ฅํ•˜๊ฒŒ ๋งŒ๋“œ๋Š” ๋“ฑ, ์‚ฌ์šฉ์ž ๊ฒฝํ—˜์— ํฐ ์˜ํ–ฅ์„ ๋ฏธ์นฉ๋‹ˆ๋‹ค.

3. ์ง‘์‚ฌ โ€˜๋„๋น„(Dobby)โ€˜์™€ ์—์ด์ „ํŠธ ํผ์ŠคํŠธ ์„ธ์ƒ

์นดํŒŒ์‹œ๋Š” โ€˜ํด๋กœโ€™์˜ ์ž ์žฌ๋ ฅ์„ ๋ณด์—ฌ์ฃผ๋Š” ๊ฐœ์ธ์ ์ธ ๊ฒฝํ—˜์œผ๋กœ โ€˜์ง‘์‚ฌ ์—˜ํ”„ ๋„๋น„(Dobby the elf claw)โ€˜๋ฅผ ์†Œ๊ฐœํ–ˆ์Šต๋‹ˆ๋‹ค. ๊ทธ๋Š” ์ง€๋‚œ 1์›” ํ•œ ์ฃผ ๋™์•ˆ โ€˜ํด๋กœ ๊ด‘์ฆโ€™์— ๋น ์ ธ, ์ž์‹ ์˜ ์Šค๋งˆํŠธํ™ˆ ์‹œ์Šคํ…œ์„ ๊ด€๋ฆฌํ•˜๋Š” ์—์ด์ „ํŠธ โ€˜๋„๋น„โ€™๋ฅผ ๋งŒ๋“ค์—ˆ์Šต๋‹ˆ๋‹ค. ๊ทธ๋Š” ๋„๋น„์—๊ฒŒ โ€œ์šฐ๋ฆฌ ์ง‘์— ์žˆ๋Š” ์†Œ๋…ธ์Šค(Sonos)๋ฅผ ์ฐพ์•„๋ดโ€๋ผ๊ณ  ์ง€์‹œํ–ˆ๊ณ , ๋„๋น„๋Š” ๋กœ์ปฌ ๋„คํŠธ์›Œํฌ๋ฅผ ์Šค์บ”ํ•˜์—ฌ ์†Œ๋…ธ์Šค ์‹œ์Šคํ…œ์„ ์ฐพ๊ณ , API(Application Programming Interface)๋ฅผ ์—ญ์„ค๊ณ„ํ•˜์—ฌ ์Œ์•…์„ ์žฌ์ƒํ•˜๋Š” ๋ฐ ์„ฑ๊ณตํ–ˆ์Šต๋‹ˆ๋‹ค.

๋„๋น„๋Š” ์—ฌ๊ธฐ์„œ ๋ฉˆ์ถ”์ง€ ์•Š๊ณ , ์ง‘ ์•ˆ์˜ ์กฐ๋ช…, HVAC(๋ƒ‰๋‚œ๋ฐฉ ๊ณต์กฐ ์‹œ์Šคํ…œ), ๋ธ”๋ผ์ธ๋“œ, ์ˆ˜์˜์žฅ, ์ŠคํŒŒ, ์‹ฌ์ง€์–ด ๋ณด์•ˆ ์‹œ์Šคํ…œ๊นŒ์ง€ ์ œ์–ดํ•˜๊ฒŒ ๋˜์—ˆ์Šต๋‹ˆ๋‹ค. ์˜ˆ๋ฅผ ๋“ค์–ด, ์™ธ๋ถ€ ์นด๋ฉ”๋ผ์— ์›€์ง์ž„์ด ๊ฐ์ง€๋˜๋ฉด ๋„๋น„๋Š” ํ€ธ(Quinn) ๋ชจ๋ธ์„ ํ†ตํ•ด ์˜์ƒ์„ ๋ถ„์„ํ•˜๊ณ , โ€œํŽ˜๋ฑ์Šค ํŠธ๋Ÿญ์ด ๋„์ฐฉํ–ˆ์Šต๋‹ˆ๋‹คโ€์™€ ๊ฐ™์€ ๋ฉ”์‹œ์ง€๋ฅผ ์ด๋ฏธ์ง€์™€ ํ•จ๊ป˜ ์นดํŒŒ์‹œ์˜ ์™“์ธ ์•ฑ(WhatsApp)์œผ๋กœ ์ „์†กํ•ฉ๋‹ˆ๋‹ค. ์นดํŒŒ์‹œ๋Š” ์˜ˆ์ „์—๋Š” 6๊ฐœ์˜ ๋‹ค๋ฅธ ์•ฑ์„ ์‚ฌ์šฉํ•ด์•ผ ํ–ˆ๋˜ ์Šค๋งˆํŠธํ™ˆ ์ œ์–ด๋ฅผ ์ด์ œ ์ž์—ฐ์–ด๋กœ ๋„๋น„์—๊ฒŒ ๋ช…๋ นํ•˜๋Š” ๊ฒƒ๋งŒ์œผ๋กœ ๋ชจ๋‘ ์ฒ˜๋ฆฌํ•  ์ˆ˜ ์žˆ๊ฒŒ ๋˜์—ˆ๋‹ค๊ณ  ๊ทน์ฐฌํ•ฉ๋‹ˆ๋‹ค.

์ด๋Ÿฌํ•œ ๊ฒฝํ—˜์€ ์†Œํ”„ํŠธ์›จ์–ด์˜ ๋ฏธ๋ž˜์— ๋Œ€ํ•œ ์ค‘์š”ํ•œ ์‹œ์‚ฌ์ ์„ ๋˜์ง‘๋‹ˆ๋‹ค. ์‚ฌ๋žŒ๋“ค์€ ๋” ์ด์ƒ ๋ณต์žกํ•œ ์‚ฌ์šฉ์ž ์ธํ„ฐํŽ˜์ด์Šค(UI)๋ฅผ ๋ฐฐ์šฐ๊ณ  ์—ฌ๋Ÿฌ ์•ฑ์„ ์˜ค๊ฐ€๋Š” ๊ฒƒ์„ ์›์น˜ ์•Š์Šต๋‹ˆ๋‹ค. ๋Œ€์‹ , AI๊ฐ€ ์‚ฌ๋žŒ์˜ ์˜๋„๋ฅผ ์ž์—ฐ์–ด๋กœ ์ดํ•ดํ•˜๊ณ , ๋ชจ๋“  ํ•˜๋“œ์›จ์–ด์˜ API๋ฅผ ์ง์ ‘ ํ˜ธ์ถœํ•˜์—ฌ ์›ํ•˜๋Š” ๊ธฐ๋Šฅ์„ ์ˆ˜ํ–‰ํ•˜๋Š” โ€˜์—์ด์ „ํŠธ ํผ์ŠคํŠธ(Agent-First)โ€™ ์„ธ์ƒ์œผ๋กœ ๋‚˜์•„๊ฐ€๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค. ํ˜„์žฌ๋Š” ์—์ด์ „ํŠธ๋ฅผ ์„ค์ •ํ•˜๊ธฐ ์œ„ํ•ด โ€˜vi ์ฝ”๋”ฉโ€™๊ณผ ๊ฐ™์€ ์ˆ˜๋™ ์ž‘์—…์ด ํ•„์š”ํ•˜์ง€๋งŒ, ์นดํŒŒ์‹œ๋Š” ๋ช‡ ๋…„ ์•ˆ์— ์ด๋Ÿฌํ•œ ์„ค์ •์กฐ์ฐจ AI๊ฐ€ ์ž์œจ์ ์œผ๋กœ ์ฒ˜๋ฆฌํ•˜๊ฒŒ ๋  ๊ฒƒ์ด๋ฉฐ, ์•ฑ์Šคํ† ์–ด์˜ ์ˆ˜๋งŽ์€ ๋งž์ถคํ˜• ์•ฑ๋“ค์ด ๋ฌด์˜๋ฏธํ•ด์งˆ ๊ฒƒ์ด๋ผ๊ณ  ์˜ˆ์ธกํ•ฉ๋‹ˆ๋‹ค. ์—์ด์ „ํŠธ๊ฐ€ ์ง€๋Šฅ์˜ โ€˜์ ‘์ฐฉ์ œโ€™ ์—ญํ• ์„ ํ•˜๋ฉฐ ๋ชจ๋“  ๊ฒƒ์„ ์—ฐ๊ฒฐํ•˜๋Š” ์‹œ๋Œ€๊ฐ€ ์˜ค๊ณ  ์žˆ๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค.

4. ์ธ๊ฐ„์„ ๋ฃจํ”„์—์„œ ์ œ๊ฑฐํ•˜๋ผ: ์˜คํ† ๋ฆฌ์„œ์น˜(AutoResearch)

์นดํŒŒ์‹œ๋Š” ์—์ด์ „ํŠธ์˜ ๊ถ๊ทน์ ์ธ ๋ชฉํ‘œ๊ฐ€ ์ธ๊ฐ„์„ ์ž‘์—… ๋ฃจํ”„์—์„œ ์™„์ „ํžˆ ์ œ๊ฑฐํ•˜๊ณ  ์ž์œจ์„ฑ์„ ๊ทน๋Œ€ํ™”ํ•˜๋Š” ๊ฒƒ์ด๋ผ๊ณ  ๋งํ•ฉ๋‹ˆ๋‹ค. ๊ทธ๊ฐ€ ๊ฐœ๋ฐœํ•œ โ€˜์˜คํ† ๋ฆฌ์„œ์น˜(AutoResearch)โ€˜๋Š” ์ด๋Ÿฌํ•œ ์ฒ ํ•™์˜ ๊ตฌ์ฒด์ ์ธ ๊ตฌํ˜„์ฒด์ž…๋‹ˆ๋‹ค. ๊ทธ๋Š” ์ž์‹ ์˜ GPT-2 ๋ชจ๋ธ์„ ์ˆ˜์‹ญ ๋…„๊ฐ„ ์ง์ ‘ ํŠœ๋‹ํ•˜๋ฉฐ ์ตœ์ ํ™”์— ๋„๋‹ฌํ–ˆ๋‹ค๊ณ  ๋ฏฟ์—ˆ์ง€๋งŒ, ์˜คํ† ๋ฆฌ์„œ์น˜๋ฅผ ๋ฐค์ƒˆ ๊ฐ€๋™์‹œํ‚จ ๊ฒฐ๊ณผ, ์ž์‹ ์ด ๋†“์ณค๋˜ ์ƒˆ๋กœ์šด ํŠœ๋‹ ์กฐํ•ฉ(์˜ˆ: ๊ฐ€์ค‘์น˜ ๊ฐ์†Œ, Adam ๋ฒ ํƒ€ ๊ฐ’)์„ ๋ฐœ๊ฒฌํ–ˆ์Šต๋‹ˆ๋‹ค.

์˜คํ† ๋ฆฌ์„œ์น˜์˜ ํ•ต์‹ฌ์€ โ€˜์žฌ๊ท€์  ์ž๊ธฐ ๊ฐœ์„ (Recursive Self-Improvement)โ€˜์ž…๋‹ˆ๋‹ค. ์ธ๊ฐ„ ์—ฐ๊ตฌ์ž๊ฐ€ ์ผ์ผ์ด ๊ฒฐ๊ณผ๋ฅผ ํ™•์ธํ•˜๊ณ  ๋‹ค์Œ ๋‹จ๊ณ„๋ฅผ ์ง€์‹œํ•˜๋Š” ๋Œ€์‹ , ์—์ด์ „ํŠธ๊ฐ€ ๋ช…ํ™•ํ•œ ๋ชฉํ‘œ์™€ ํ‰๊ฐ€ ์ง€ํ‘œ, ๊ทธ๋ฆฌ๊ณ  ํ—ˆ์šฉ ๊ฐ€๋Šฅํ•œ ๊ฒฝ๊ณ„๋ฅผ ๋ถ€์—ฌ๋ฐ›์•„ ์Šค์Šค๋กœ ์‹คํ—˜ํ•˜๊ณ  ๊ฐœ์„ ํ•˜๋Š” ๊ณผ์ •์„ ๋ฐ˜๋ณตํ•˜๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค. ์นดํŒŒ์‹œ๋Š” ๋Œ€๊ทœ๋ชจ ์–ธ์–ด ๋ชจ๋ธ(LLM)์„ ๊ฐœ๋ฐœํ•˜๋Š” ์„ ๋‘ ์—ฐ๊ตฌ์†Œ๋“ค์ด ์ด๋ฏธ ์ด๋Ÿฌํ•œ ์ž์œจ์ ์ธ ์ž๊ธฐ ๊ฐœ์„  ์‹œ์Šคํ…œ์„ ๊ตฌ์ถ•ํ•˜๊ณ  ์žˆ์„ ๊ฒƒ์ด๋ผ๊ณ  ์ถ”์ธกํ•ฉ๋‹ˆ๋‹ค. ์ž‘์€ ๋ชจ๋ธ์—์„œ ์ž์œจ์ ์œผ๋กœ ํƒ์ƒ‰ํ•˜๊ณ  ์ตœ์ ํ™”ํ•œ ํ›„, ์ด๋ฅผ ๋Œ€๊ทœ๋ชจ ๋ชจ๋ธ์— ์ ์šฉํ•˜๋Š” ๋ฐฉ์‹์œผ๋กœ ์—ฐ๊ตฌ ํšจ์œจ์„ฑ์„ ๊ทน๋Œ€ํ™”ํ•˜๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค. ์—ฐ๊ตฌ์ž๋“ค์ด ์•„์ด๋””์–ด๋ฅผ ์ œ์•ˆํ•˜๋Š” ์—ญํ• ์€ ํ•˜์ง€๋งŒ, ์‹ค์ œ ๊ตฌํ˜„ ๋ฐ ์‹คํ—˜์€ ์ž๋™ํ™”๋œ ์‹œ์Šคํ…œ์ด ๋‹ด๋‹นํ•˜๋Š” ๋ฏธ๋ž˜๊ฐ€ ์˜ค๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค.

5. ๋ฉ”ํƒ€ ์ตœ์ ํ™”์˜ ๋ฌดํ•œ ๋ฃจํ”„: โ€˜ํ”„๋กœ๊ทธ๋žจ MDโ€™์˜ ์ง„ํ™”

์˜คํ† ๋ฆฌ์„œ์น˜์˜ ๊ฐœ๋…์„ ํ•œ ๋‹จ๊ณ„ ๋” ๋‚˜์•„๊ฐ€๋ฉด, AI๊ฐ€ ์—ฐ๊ตฌ ์กฐ์ง ์ž์ฒด์˜ โ€˜์šด์˜ ๋ฐฉ์‹โ€™๊นŒ์ง€ ๊ฐœ์„ ํ•˜๋Š” ๋ฉ”ํƒ€ ์ตœ์ ํ™”(Meta-Optimization)์˜ ์˜์—ญ์— ๋„๋‹ฌํ•ฉ๋‹ˆ๋‹ค. ์นดํŒŒ์‹œ๋Š” ์ž์‹ ์ด ์˜คํ† ๋ฆฌ์„œ์น˜์˜ ์ž‘๋™ ๋ฐฉ์‹์„ ์„ค๋ช…ํ•˜๊ธฐ ์œ„ํ•ด ๋งŒ๋“  ๋งˆํฌ๋‹ค์šด(Markdown) ํŒŒ์ผ์ธ โ€˜ํ”„๋กœ๊ทธ๋žจ MD(Program MD)โ€˜๋ฅผ ์˜ˆ๋กœ ๋“ญ๋‹ˆ๋‹ค. ์ด ํŒŒ์ผ์€ โ€œ์ด๊ฒƒ์„ ํ•˜๊ณ , ์ €๊ฒƒ์„ ํ•œ ๋‹ค์Œ, ์ด๋Ÿฐ ์•„์ด๋””์–ด๋ฅผ ์‹œ๋„ํ•˜๋ผโ€๋Š” ์‹์˜ ์ง€์นจ์„ ๋‹ด๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค.

๊ทธ๋Š” ์—ฌ๊ธฐ์„œ ๋” ๋‚˜์•„๊ฐ€, AI๊ฐ€ ๋” ๋‚˜์€ โ€˜ํ”„๋กœ๊ทธ๋žจ MDโ€™๋ฅผ ์ž‘์„ฑํ•  ์ˆ˜ ์žˆ์„ ๊ฒƒ์ด๋ผ๋Š” ์•„์ด๋””์–ด๋ฅผ ์ œ์‹œํ•ฉ๋‹ˆ๋‹ค. ์˜ˆ๋ฅผ ๋“ค์–ด, ์‚ฌ๋žŒ๋“ค์ด ๋‹ค์–‘ํ•œ โ€˜ํ”„๋กœ๊ทธ๋žจ MDโ€™๋ฅผ ์ž‘์„ฑํ•˜๊ณ , ๋™์ผํ•œ ํ•˜๋“œ์›จ์–ด์—์„œ ์–ด๋–ค โ€˜ํ”„๋กœ๊ทธ๋žจ MDโ€™๊ฐ€ ๊ฐ€์žฅ ํฐ ๊ฐœ์„ ์„ ๊ฐ€์ ธ์˜ค๋Š”์ง€ ๊ฒฝ์Ÿํ•˜๋Š” ์ฝ˜ํ…Œ์ŠคํŠธ๋ฅผ ์—ด ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค. ์ด ๋ฐ์ดํ„ฐ๋ฅผ ๋ฐ”ํƒ•์œผ๋กœ AI๋Š” ๊ฐ€์žฅ ํšจ์œจ์ ์ธ ์—ฐ๊ตฌ ์กฐ์ง ์šด์˜ ๋ฐฉ์‹์„ ํ•™์Šตํ•˜๊ณ , ๋” ๋‚˜์€ โ€˜ํ”„๋กœ๊ทธ๋žจ MDโ€™๋ฅผ ์ƒ์„ฑํ•  ์ˆ˜ ์žˆ๊ฒŒ ๋ฉ๋‹ˆ๋‹ค.

์ด๋Š” ๊ณง ์—ฐ๊ตฌ ์กฐ์ง์˜ ๋ชจ๋“  ์—ญํ• ๊ณผ ์—ฐ๊ฒฐ ๋ฐฉ์‹์„ ์„ค๋ช…ํ•˜๋Š” ๋งˆํฌ๋‹ค์šด ํŒŒ์ผ, ์ฆ‰ โ€˜์ฝ”๋“œโ€™๊ฐ€ ๋˜๊ณ , ์ด ์ฝ”๋“œ๋ฅผ ํŠœ๋‹ํ•˜์—ฌ ๋” ํšจ์œจ์ ์ธ ์—ฐ๊ตฌ ์กฐ์ง์„ ๋งŒ๋“ค ์ˆ˜ ์žˆ๋‹ค๋Š” ์˜๋ฏธ์ž…๋‹ˆ๋‹ค. ์ด๋Ÿฌํ•œ ๋ฌดํ•œํ•œ ์ตœ์ ํ™”์˜ ๊ณ„์ธต, ์ฆ‰ LLM, ์—์ด์ „ํŠธ, ํด๋กœ ์—”ํ‹ฐํ‹ฐ, ๊ทธ๋ฆฌ๊ณ  ์ด๋“ค์„ ์ง€์‹œํ•˜๋Š” ๋ช…๋ น ์ตœ์ ํ™”์— ์ด๋ฅด๊ธฐ๊นŒ์ง€ ๋ชจ๋“  ๊ฒƒ์ด โ€˜์Šคํ‚ฌ ๋ฌธ์ œโ€™๋กœ ๊ท€๊ฒฐ๋˜๋Š” ์ƒํ™ฉ์ด ๋ฐ”๋กœ ์นดํŒŒ์‹œ๊ฐ€ ๋งํ•˜๋Š” โ€˜AI ๊ด‘์ฆโ€™์˜ ๋ณธ์งˆ์ž…๋‹ˆ๋‹ค.

6. ํ˜„์žฌ์˜ ํ•œ๊ณ„์™€ ๋ฏธ๋ž˜์˜ ์ŠคํŽ™ํŠธ๋Ÿผ

์ด๋Ÿฌํ•œ ๋ฌดํ•œํ•œ ๊ฐ€๋Šฅ์„ฑ ์†์—์„œ๋„ ์นดํŒŒ์‹œ๋Š” ํ˜„์žฌ AI ์—์ด์ „ํŠธ์˜ ํ•œ๊ณ„์™€ ๊ณผ์ œ๋ฅผ ๋ช…ํ™•ํžˆ ์ง€์ ํ•ฉ๋‹ˆ๋‹ค.

  • ๊ฐ๊ด€์  ํ‰๊ฐ€์˜ ์ค‘์š”์„ฑ: ์—์ด์ „ํŠธ์™€ ์˜คํ† ๋ฆฌ์„œ์น˜๋Š” ๊ฐ๊ด€์ ์ธ ํ‰๊ฐ€ ์ง€ํ‘œ๊ฐ€ ๋ช…ํ™•ํ•˜๊ณ  ์‰ฝ๊ฒŒ ์ธก์ • ๊ฐ€๋Šฅํ•œ ์ž‘์—…(์˜ˆ: CUDA ์ปค๋„ ์ตœ์ ํ™”)์— ๋งค์šฐ ํšจ๊ณผ์ ์ž…๋‹ˆ๋‹ค. ๊ทธ๋Ÿฌ๋‚˜ ํ‰๊ฐ€๊ฐ€ ์–ด๋ ค์šด ์ฃผ๊ด€์ ์ธ ์˜์—ญ์—์„œ๋Š” ๊ทธ ํšจ๊ณผ๊ฐ€ ์ œํ•œ์ ์ž…๋‹ˆ๋‹ค.
  • ๋“ค์ญ‰๋‚ ์ญ‰ํ•จ(Jaggedness): ํ˜„์žฌ์˜ ๋ชจ๋ธ๋“ค์€ โ€œ๊ทน๋„๋กœ ๋˜‘๋˜‘ํ•œ ๋ฐ•์‚ฌ ๊ณผ์ • ํ•™์ƒ์ด๋ฉด์„œ ๋™์‹œ์— 10์‚ด์งœ๋ฆฌ ์•„์ดโ€์™€ ๊ฐ™์€ โ€˜๋“ค์ญ‰๋‚ ์ญ‰ํ•จโ€™์„ ๋ณด์ž…๋‹ˆ๋‹ค. ๋งค์šฐ ๋ณต์žกํ•œ ์‹œ์Šคํ…œ ํ”„๋กœ๊ทธ๋ž˜๋ฐ ๋ฌธ์ œ๋ฅผ ํ•ด๊ฒฐํ•˜๋‹ค๊ฐ€๋„, ํ„ฐ๋ฌด๋‹ˆ์—†๋Š” ์‹ค์ˆ˜๋ฅผ ๋ฐ˜๋ณตํ•˜๊ฑฐ๋‚˜ ๋ฏธ๋ฌ˜ํ•œ ์˜๋„๋ฅผ ํŒŒ์•…ํ•˜์ง€ ๋ชปํ•ด ์‚ฌ์šฉ์ž๋ฅผ ์ขŒ์ ˆ์‹œํ‚ค๊ธฐ๋„ ํ•ฉ๋‹ˆ๋‹ค.
  • ์ตœ์ ํ™”์˜ ๋งน์ : ๋ชจ๋ธ์˜ ๋Šฅ๋ ฅ์ด ์—„์ฒญ๋‚˜๊ฒŒ ํ–ฅ์ƒ๋˜์—ˆ์Œ์—๋„ ๋ถˆ๊ตฌํ•˜๊ณ , โ€œ์™œ ๊ณผํ•™์ž๋“ค์€ ์›์ž๋ฅผ ๋ฏฟ์ง€ ์•Š์„๊นŒ์š”? ๊ทธ๋“ค์€ ๋ชจ๋“  ๊ฒƒ์„ ๋งŒ๋“ค์–ด๋‚ด๋‹ˆ๊นŒ์š”(make everything up).โ€์™€ ๊ฐ™์€ 5๋…„ ์ „์˜ ๋‚ก์€ ๋†๋‹ด์„ ์—ฌ์ „ํžˆ ๋ฐ˜๋ณตํ•ฉ๋‹ˆ๋‹ค. ์ด๋Š” ๊ฐ•ํ™”ํ•™์Šต(Reinforcement Learning)์„ ํ†ตํ•ด ์ตœ์ ํ™”๋˜์ง€ ์•Š์€ ์˜์—ญ, ์ฆ‰ โ€˜๊ฒ€์ฆ ๋ถˆ๊ฐ€๋Šฅํ•œโ€™ ์˜์—ญ์—์„œ๋Š” ๋ชจ๋ธ์˜ ์ง€๋Šฅ์ด ๋ฐœ์ „ํ•˜์ง€ ์•Š๊ณ  ์ •์ฒด๋  ์ˆ˜ ์žˆ์Œ์„ ๋ณด์—ฌ์ค๋‹ˆ๋‹ค. ์ฝ”๋“œ ์ƒ์„ฑ ๋Šฅ๋ ฅ๊ณผ ๊ฐ™์€ โ€˜๊ฒ€์ฆ ๊ฐ€๋Šฅํ•œโ€™ ์ง€๋Šฅ์ด โ€˜๋†๋‹ด ์ง€๋Šฅโ€™๊ณผ ๊ฐ™์€ ๊ด‘๋ฒ”์œ„ํ•œ ์˜์—ญ์œผ๋กœ ์ผ๋ฐ˜ํ™”๋˜์ง€ ์•Š๋Š”๋‹ค๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค.
  • ๋ชจ๋ธ์˜ ๋ถ„ํ™”(Speciation) ํ•„์š”์„ฑ: ํ˜„์žฌ ์—ฐ๊ตฌ์†Œ๋“ค์€ ๋ชจ๋“  ์˜์—ญ์—์„œ ์ž„์˜๋กœ ์ง€๋Šฅ์ ์ธ ๋‹จ์ผ ๋ชจ๋ธ(Monoculture of Models)์„ ์ง€ํ–ฅํ•˜๋Š” ๊ฒฝํ–ฅ์ด ์žˆ์Šต๋‹ˆ๋‹ค. ๊ทธ๋Ÿฌ๋‚˜ ์นดํŒŒ์‹œ๋Š” ๋™๋ฌผ ์™•๊ตญ์˜ ๋‹ค์–‘ํ•œ ๋‡Œ์ฒ˜๋Ÿผ, AI ๋˜ํ•œ ํŠน์ • ์˜์—ญ์— ํŠนํ™”๋œ โ€˜๋ถ„ํ™”๋œ ์ง€๋Šฅ(Speciated Intelligence)โ€˜์œผ๋กœ ๋‚˜์•„๊ฐ€์•ผ ํ•œ๋‹ค๊ณ  ์ฃผ์žฅํ•ฉ๋‹ˆ๋‹ค. ์ˆ˜ํ•™ ์ฆ๋ช…์ด๋‚˜ ํŠน์ • ์‚ฐ์—… ๋ถ„์•ผ์™€ ๊ฐ™์ด ํ‹ˆ์ƒˆ์‹œ์žฅ์— ํŠนํ™”๋œ ์ž‘๊ณ  ํšจ์œจ์ ์ธ ๋ชจ๋ธ๋“ค์ด ๋“ฑ์žฅํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค. ์ปดํ“จํŒ… ์ธํ”„๋ผ์˜ ์ œ์•ฝ ๋˜ํ•œ ์ด๋Ÿฌํ•œ ๋ถ„ํ™”๋ฅผ ๊ฐ€์†ํ™”ํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
  • ๋‘๋‡Œ ์กฐ์ž‘ ๊ณผํ•™์˜ ๋ฏธ์„ฑ์ˆ™: ํ˜„์žฌ๋Š” ์ปจํ…์ŠคํŠธ ์œˆ๋„์šฐ(Context Window)๋ฅผ ํ†ตํ•œ ๋ชจ๋ธ ์กฐ์ž‘์ด ์ฃผ๋ฅผ ์ด๋ฃจ์ง€๋งŒ, ๋ชจ๋ธ์˜ ๊ฐ€์ค‘์น˜(Weights)๋ฅผ ์ง์ ‘ ์กฐ์ •ํ•˜์—ฌ ๋ฏธ์„ธ ์กฐ์ •(Fine-tuning)ํ•˜๊ฑฐ๋‚˜ ์ง€์†์ ์ธ ํ•™์Šต(Continual Learning)์„ ํ†ตํ•ด ํŠน์ • ์˜์—ญ์—์„œ ๋ชจ๋ธ์„ ๋” ๊นŠ์ด ๊ฐœ์„ ํ•˜๋Š” ๊ณผํ•™์€ ์•„์ง ์ดˆ๊ธฐ ๋‹จ๊ณ„์— ๋จธ๋ฌผ๋Ÿฌ ์žˆ์Šต๋‹ˆ๋‹ค.

๊ฒฐ๋ก : ๋๋‚˜์ง€ ์•Š๋Š” ๋ฃจํ”„, ๋ฌดํ•œํ•œ ๊ฐ€๋Šฅ์„ฑ

์•ˆ๋“œ๋ ˆ์ด ์นดํŒŒ์‹œ๊ฐ€ ๋ฌ˜์‚ฌํ•˜๋Š” AI์˜ ๋ฏธ๋ž˜๋Š” ์ธ๊ฐ„์ด ๋” ์ด์ƒ โ€˜๋ณ‘๋ชฉโ€™์ด ์•„๋‹Œ, ์—์ด์ „ํŠธ์—๊ฒŒ โ€˜์˜์ง€โ€™๋ฅผ ํ‘œํ˜„ํ•˜๊ณ  ์‹œ์Šคํ…œ์„ ์„ค๊ณ„ํ•˜๋Š” ์ƒˆ๋กœ์šด ์—ญํ• ๋กœ ์ „ํ™˜๋˜๋Š” ์‹œ๋Œ€์ž…๋‹ˆ๋‹ค. โ€˜AI ๊ด‘์ฆโ€™์€ ๋ฌดํ•œํ•œ ์ตœ์ ํ™”์˜ ๋ฃจํ”„์™€ ์•„์ง ํƒํ—˜๋˜์ง€ ์•Š์€ ์ž ์žฌ๋ ฅ์— ๋Œ€ํ•œ ํฅ๋ถ„๊ณผ ์•ฝ๊ฐ„์˜ ๋ถˆ์•ˆ๊ฐ์„ ๋™์‹œ์— ๋‚ดํฌํ•ฉ๋‹ˆ๋‹ค.

์ง€๊ธˆ์€ ๋ชจ๋ธ์˜ โ€˜๋“ค์ญ‰๋‚ ์ญ‰ํ•จโ€™๊ณผ โ€˜์ตœ์ ํ™”์˜ ๋งน์ โ€™ ๊ฐ™์€ ํ•œ๊ณ„๊ฐ€ ์กด์žฌํ•˜์ง€๋งŒ, ์ด๋Ÿฌํ•œ ๋ฌธ์ œ์ ๋“ค์€ ๊ฒฐ๊ตญ โ€˜์Šคํ‚ฌ ๋ฌธ์ œโ€™๋กœ ํ•ด๊ฒฐ๋  ๊ฒƒ์ž…๋‹ˆ๋‹ค. ์ธ๊ฐ„์€ ์—์ด์ „ํŠธ๋ฅผ ๋” ์ž˜ ํ™œ์šฉํ•˜๊ณ , ๋” ๋‚˜์€ โ€˜ํ”„๋กœ๊ทธ๋žจ MDโ€™๋ฅผ ์ž‘์„ฑํ•˜๋ฉฐ, ๊ถ๊ทน์ ์œผ๋กœ ์ธ๊ฐ„์˜ ๊ฐœ์ž… ์—†์ด ์Šค์Šค๋กœ ๊ฐœ์„ ํ•˜๊ณ  ์ง„ํ™”ํ•˜๋Š” ์ž์œจ์ ์ธ AI ์‹œ์Šคํ…œ์„ ๊ตฌ์ถ•ํ•˜๋Š” ๋ฐฉ๋ฒ•์„ ์ฐพ์•„๋‚ผ ๊ฒƒ์ž…๋‹ˆ๋‹ค. ์ด โ€˜๋ฌดํ•œ ๋ฃจํ”„์˜ ์‹œ๋Œ€(Loopy Era)โ€˜๋Š” ์—”์ง€๋‹ˆ์–ด๋ง, ์—ฐ๊ตฌ, ๊ทธ๋ฆฌ๊ณ  ์šฐ๋ฆฌ์˜ ์ผ์ƒ๊นŒ์ง€๋„ ๊ทผ๋ณธ์ ์œผ๋กœ ์žฌํŽธํ•  ๊ฒƒ์ด๋ฉฐ, ์šฐ๋ฆฌ๋Š” ๊ทธ ๋ณ€ํ™”์˜ ์‹œ์ž‘์ ์— ์„œ ์žˆ์Šต๋‹ˆ๋‹ค.


โ€œTerence Tao โ€“ How the worldโ€™s top mathematician uses AIโ€ โ€” Dwarkesh Patel ๊ธฐ๋ฐ˜ ๊ธฐ์‚ฌ ์›๋ณธ ์˜์ƒ ๋ณด๊ธฐ

AI, ๊ณผํ•™์  ๋ฐœ๊ฒฌ์˜ ํŒจ๋Ÿฌ๋‹ค์ž„์„ ๋ฐ”๊พธ๋‹ค: ํ…Œ๋ Œ์Šค ํƒ€์˜ค์™€์˜ ๋Œ€ํ™”

์„ธ๊ณ„ ์ตœ๊ณ ์˜ ์ˆ˜ํ•™์ž ์ค‘ ํ•œ ๋ช…์ธ ํ…Œ๋ Œ์Šค ํƒ€์˜ค(Terence Tao) ๊ต์ˆ˜๋Š” ์ธ๊ณต์ง€๋Šฅ(AI)์ด ๊ณผํ•™์  ๋ฐœ๊ฒฌ์˜ ๋ณธ์งˆ์„ ์–ด๋–ป๊ฒŒ ๋ณ€ํ™”์‹œํ‚ค๊ณ  ์žˆ๋Š”์ง€์— ๋Œ€ํ•œ ๊นŠ์ด ์žˆ๋Š” ํ†ต์ฐฐ์„ ์ œ์‹œํ•ฉ๋‹ˆ๋‹ค. ๋“œ์™€์ผ€์‹œ ํŒŒํ…”(Dwarkesh Patel)๊ณผ์˜ ์ธํ„ฐ๋ทฐ์—์„œ ํƒ€์˜ค ๊ต์ˆ˜๋Š” ์ผ€ํ”Œ๋Ÿฌ(Kepler)์˜ ํ–‰์„ฑ ์šด๋™ ๋ฒ•์น™ ๋ฐœ๊ฒฌ ๊ณผ์ •์„ ์‹œ์ž‘์œผ๋กœ, AI๊ฐ€ ์•„์ด๋””์–ด ์ƒ์„ฑ์˜ ๋น„์šฉ์„ ๊ฑฐ์˜ 0์œผ๋กœ ๋‚ฎ์ถ”๊ณ  ์žˆ์ง€๋งŒ, ๋™์‹œ์— ๊ณผํ•™๊ณ„์— ์ƒˆ๋กœ์šด ๋ณ‘๋ชฉ ํ˜„์ƒ, ์ฆ‰ โ€˜๊ฒ€์ฆ๊ณผ ํ‰๊ฐ€โ€™๋ผ๋Š” ๊ฑฐ๋Œ€ํ•œ ๊ณผ์ œ๋ฅผ ์•ˆ๊ฒจ์ฃผ๊ณ  ์žˆ๋‹ค๊ณ  ์—ญ์„คํ•ฉ๋‹ˆ๋‹ค. ์ด ๊ธฐ์‚ฌ๋Š” ํƒ€์˜ค ๊ต์ˆ˜์˜ ์ด์•ผ๊ธฐ๋ฅผ ํ†ตํ•ด AI ์‹œ๋Œ€์˜ ๊ณผํ•™์ด ๋‚˜์•„๊ฐ€์•ผ ํ•  ๋ฐฉํ–ฅ๊ณผ ๊ทธ ๋ณต์žกํ•œ ํ•จ์˜๋ฅผ ํƒ๊ตฌํ•ฉ๋‹ˆ๋‹ค.

1. ์ผ€ํ”Œ๋Ÿฌ์™€ ๋ธŒ๋ผํ—ค: ๊ฒฝํ—˜์  ๋ฐœ๊ฒฌ์˜ ์„œ๋ง‰

ํƒ€์˜ค ๊ต์ˆ˜๋Š” ์ธ๊ณต์ง€๋Šฅ์ด ์ˆ˜ํ•™์— ๋ฏธ์น  ์˜ํ–ฅ์„ ๋…ผํ•˜๊ธฐ ์œ„ํ•œ ํ›Œ๋ฅญํ•œ ์ถœ๋ฐœ์ ์œผ๋กœ ์š”ํ•˜๋„ค์Šค ์ผ€ํ”Œ๋Ÿฌ์˜ ํ–‰์„ฑ ์šด๋™ ๋ฒ•์น™ ๋ฐœ๊ฒฌ ์ด์•ผ๊ธฐ๋ฅผ ๊บผ๋ƒ…๋‹ˆ๋‹ค. ์ผ€ํ”Œ๋Ÿฌ๋Š” ์ฝ”ํŽ˜๋ฅด๋‹ˆ์ฟ ์Šค(Copernicus)์˜ ์ง€๋™์„ค, ์ฆ‰ ํƒœ์–‘์ด ํƒœ์–‘๊ณ„์˜ ์ค‘์‹ฌ์— ์žˆ๊ณ  ํ–‰์„ฑ๋“ค์ด ๊ทธ ์ฃผ์œ„๋ฅผ ๋ˆ๋‹ค๋Š” ๋ชจ๋ธ์— ๊ธฐ๋ฐ˜์„ ๋‘์—ˆ์Šต๋‹ˆ๋‹ค. ์ฝ”ํŽ˜๋ฅด๋‹ˆ์ฟ ์Šค๋Š” ํ–‰์„ฑ ๊ถค๋„๊ฐ€ ์™„๋ฒฝํ•œ ์›์ด๋ผ๊ณ  ๋ฏฟ์—ˆ์ง€๋งŒ, ์ผ€ํ”Œ๋Ÿฌ๋Š” ์ด ๊ถค๋„์˜ ๋น„์œจ์—์„œ ์–ด๋–ค ๊ธฐํ•˜ํ•™์  ์˜๋ฏธ๋ฅผ ๋ฐœ๊ฒฌํ•˜๋ ค ํ–ˆ์Šต๋‹ˆ๋‹ค. ๊ทธ๋Š” ์ง€๊ตฌ ๊ถค๋„๋ฅผ ์ •์œก๋ฉด์ฒด ์•ˆ์— ๋„ฃ์œผ๋ฉด, ๊ทธ ์ •์œก๋ฉด์ฒด๋ฅผ ๋‘˜๋Ÿฌ์‹ธ๋Š” ์™ธ๋ถ€ ๊ตฌ๊ฐ€ ํ™”์„ฑ์˜ ๊ถค๋„์™€ ๊ฑฐ์˜ ์™„๋ฒฝํ•˜๊ฒŒ ์ผ์น˜ํ•˜๋Š” ๋“ฑ, ๋‹น์‹œ ์•Œ๋ ค์ง„ ์—ฌ์„ฏ ํ–‰์„ฑ์˜ ๊ถค๋„ ์‚ฌ์ด์— ๋‹ค์„ฏ ๊ฐœ์˜ ํ”Œ๋ผํ†ค ์ž…์ฒด(Platonic solids: ์ •์œก๋ฉด์ฒด, ์ •์‚ฌ๋ฉด์ฒด, ์ •์ด์‹ญ๋ฉด์ฒด, ์ •ํŒ”๋ฉด์ฒด, ์ •์‹ญ์ด๋ฉด์ฒด)๋ฅผ ๋ผ์›Œ ๋„ฃ์„ ์ˆ˜ ์žˆ๋‹ค๋Š” ์•„๋ฆ„๋‹ค์šด ์ด๋ก ์„ ์ œ์•ˆํ–ˆ์Šต๋‹ˆ๋‹ค. ์ด๋Š” ์‹ ์˜ ์„ค๊ณ„๊ฐ€ ์ˆ˜ํ•™์  ์™„๋ฒฝํ•จ๊ณผ ์ผ์น˜ํ•œ๋‹ค๊ณ  ๋ฏฟ์—ˆ๊ธฐ ๋•Œ๋ฌธ์ž…๋‹ˆ๋‹ค.

์ด ์ด๋ก ์„ ํ™•์ธํ•˜๊ธฐ ์œ„ํ•ด ์ผ€ํ”Œ๋Ÿฌ์—๊ฒŒ๋Š” ๋ฐ์ดํ„ฐ๊ฐ€ ํ•„์š”ํ–ˆ์Šต๋‹ˆ๋‹ค. ๋‹น์‹œ ์œ ์ผํ•˜๊ฒŒ ๊ณ ํ’ˆ์งˆ ๋ฐ์ดํ„ฐ๋ฅผ ๋ณด์œ ํ•˜๊ณ  ์žˆ๋˜ ์ธ๋ฌผ์€ ๋ด๋งˆํฌ์˜ ๋ถ€์œ ํ•˜๊ณ  ๊ดด์งœ ์ฒœ๋ฌธํ•™์ž ํŠ€์ฝ” ๋ธŒ๋ผํ—ค(Tycho Brahe)์˜€์Šต๋‹ˆ๋‹ค. ๋ธŒ๋ผํ—ค๋Š” ์ˆ˜์‹ญ ๋…„๊ฐ„ ์œก์•ˆ์œผ๋กœ ํ–‰์„ฑ๋“ค์„ ๊ด€์ธกํ•˜๋ฉฐ ๋ฐฉ๋Œ€ํ•œ ๋ฐ์ดํ„ฐ๋ฅผ ์ถ•์ ํ–ˆ์Šต๋‹ˆ๋‹ค. ์ผ€ํ”Œ๋Ÿฌ๋Š” ๋ธŒ๋ผํ—ค์™€ ํ•จ๊ป˜ ์ผํ–ˆ์ง€๋งŒ, ๋ธŒ๋ผํ—ค๋Š” ๋ฐ์ดํ„ฐ๋ฅผ ๋งค์šฐ ์งˆํˆฌํ•˜์—ฌ ์กฐ๊ธˆ์”ฉ๋งŒ ์ œ๊ณตํ–ˆ์Šต๋‹ˆ๋‹ค. ๊ฒฐ๊ตญ ์ผ€ํ”Œ๋Ÿฌ๋Š” ์ด ๋ฐ์ดํ„ฐ๋ฅผ โ€˜ํ›”์น˜๋‹ค์‹œํ”ผโ€™ ํ•˜์—ฌ ์–ป๊ฒŒ ๋˜์—ˆ๊ณ , ์ž์‹ ์˜ ์•„๋ฆ„๋‹ค์šด ์ด๋ก ์ด ์‹ค์ œ ๋ฐ์ดํ„ฐ์™€ 10% ์ •๋„ ์˜ค์ฐจ๊ฐ€ ์žˆ์Œ์„ ๋ฐœ๊ฒฌํ•˜๊ณ  ํฌ๊ฒŒ ์‹ค๋งํ–ˆ์Šต๋‹ˆ๋‹ค.

์ผ€ํ”Œ๋Ÿฌ๋Š” ์ˆ˜๋…„๊ฐ„ ์ด ๋ฌธ์ œ์— ๋งค๋‹ฌ๋ฆฐ ๋์—, ๋ฐ์ดํ„ฐ์—์„œ ์‹ค์ œ ํ–‰์„ฑ ๊ถค๋„๋ฅผ ๋„์ถœํ•˜๋Š” ๋†€๋ผ์šด ๋ถ„์„ ๋Šฅ๋ ฅ์„ ๋ฐœํœ˜ํ–ˆ์Šต๋‹ˆ๋‹ค. ๊ทธ ๊ฒฐ๊ณผ ๊ถค๋„๊ฐ€ ์›์ด ์•„๋‹Œ ํƒ€์›์ด๋ผ๋Š” ์ถฉ๊ฒฉ์ ์ธ ์‚ฌ์‹ค์„ ๋ฐํ˜€๋ƒˆ๊ณ , ๋‘ ๊ฐ€์ง€ ํ–‰์„ฑ ์šด๋™ ๋ฒ•์น™(ํƒ€์› ๊ถค๋„, ๋ฉด์  ์†๋„ ์ผ์ •์˜ ๋ฒ•์น™)์„ ์ •๋ฆฝํ–ˆ์Šต๋‹ˆ๋‹ค. 10๋…„ ํ›„, ๊ทธ๋Š” ์„ธ ๋ฒˆ์งธ ๋ฒ•์น™(๊ณต์ „ ์ฃผ๊ธฐ์˜ ์ œ๊ณฑ์€ ๊ถค๋„ ์žฅ๋ฐ˜๊ฒฝ์˜ ์„ธ์ œ๊ณฑ์— ๋น„๋ก€)๊นŒ์ง€ ๋ฐœ๊ฒฌํ–ˆ์Šต๋‹ˆ๋‹ค. ์ผ€ํ”Œ๋Ÿฌ๋Š” ์ด ๋ฒ•์น™๋“ค์— ๋Œ€ํ•œ ๋ฌผ๋ฆฌ์  ์„ค๋ช…์€ ์ œ์‹œํ•˜์ง€ ๋ชปํ–ˆ์œผ๋ฉฐ, ์ด๋Š” ํ•œ ์„ธ๊ธฐ ํ›„ ๋‰ดํ„ด(Newton)์ด ์ค‘๋ ฅ ์ด๋ก ์„ ํ†ตํ•ด ์ด ๋ชจ๋“  ๊ฒƒ์„ ์„ค๋ช…ํ•˜๋ฉด์„œ ๋น„๋กœ์†Œ ์™„์„ฑ๋˜์—ˆ์Šต๋‹ˆ๋‹ค.

ํƒ€์˜ค ๊ต์ˆ˜๋Š” ์ด ์ด์•ผ๊ธฐ์—์„œ ์ค‘์š”ํ•œ ํ†ต์ฐฐ์„ ์ œ์‹œํ•ฉ๋‹ˆ๋‹ค. ์ผ€ํ”Œ๋Ÿฌ๋Š” ๋งˆ์น˜ โ€˜๊ณ ์˜จ์˜ ๋Œ€๊ทœ๋ชจ ์–ธ์–ด ๋ชจ๋ธ(LLM)โ€˜๊ณผ ๊ฐ™์•˜๋‹ค๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค. ๊ทธ๋Š” โ€œ์„ธ๊ณ„์˜ ์กฐํ™”(Harmonics of the World)โ€œ๋ผ๋Š” ์ฑ…์—์„œ ํ–‰์„ฑ๋“ค์˜ โ€˜์Œ์ •โ€™์„ ๋…ผํ•˜๋Š” ๋“ฑ ๋ฌด์ž‘์œ„์ ์ธ ๊ด€๊ณ„๋“ค์„ ๋Š์ž„์—†์ด ์‹œ๋„ํ–ˆ์Šต๋‹ˆ๋‹ค. ์ด ๊ณผ์ •์—์„œ ์ง€๊ตฌ์˜ ์Œ์ •์€ โ€˜๋ฏธ-ํŒŒ-๋ฏธโ€™์ด๋ฉฐ, ์ด๊ฒƒ์ด ์ง€๊ตฌ์˜ ๊ธฐ๊ทผ๊ณผ ๋ถˆํ–‰์˜ ์›์ธ์ด๋ผ๋Š” ์‹์˜ ํ™ฉ๋‹นํ•œ ์ ์„ฑ์ˆ ์  ์•„์ด๋””์–ด๋„ ํฌํ•จ๋˜์–ด ์žˆ์—ˆ์Šต๋‹ˆ๋‹ค. ํ•˜์ง€๋งŒ ๊ทธ ์†์—์„œ ์„ธ ๋ฒˆ์งธ ํ–‰์„ฑ ์šด๋™ ๋ฒ•์น™์ธ โ€˜์ œ๊ณฑ-์„ธ์ œ๊ณฑ ๋ฒ•์น™(cube-square law)โ€˜์ด ๋ฐœ๊ฒฌ๋˜์—ˆ์Šต๋‹ˆ๋‹ค. ๋ธŒ๋ผํ—ค์˜ ๋ฐ์ดํ„ฐ์™€ ๊ฐ™์€ ๊ฒ€์ฆ ๊ฐ€๋Šฅํ•œ ๋ฐ์ดํ„ฐ ๋ฑ…ํฌ๋งŒ ์žˆ๋‹ค๋ฉด, LLM์€ ์ผ€ํ”Œ๋Ÿฌ์ฒ˜๋Ÿผ 20๋…„ ๋™์•ˆ ๋ฌด์ž‘์œ„์ ์ธ ๊ด€๊ณ„๋ฅผ ์‹œ๋„ํ•  ์ˆ˜ ์žˆ๊ณ , ๊ทธ์ค‘ ์ผ๋ถ€๋Š” ๊ณผํ•™์  ์ง„๋ณด๋กœ ์ด์–ด์งˆ ์ˆ˜ ์žˆ๋‹ค๋Š” ๊ฒƒ์ด์ฃ .

2. ๋ณ€ํ™”ํ•˜๋Š” ๊ณผํ•™์  ๋ฐฉ๋ฒ•๋ก : ๊ฐ€์„ค์—์„œ ๋ฐ์ดํ„ฐ ์ค‘์‹ฌ์œผ๋กœ

์ „ํ†ต์ ์œผ๋กœ ๊ณผํ•™์€ ๋ฌธ์ œ ์‹๋ณ„, ๊ฐ€์„ค ์ˆ˜๋ฆฝ, ๋ฐ์ดํ„ฐ ์ˆ˜์ง‘, ๊ฒ€์ฆ, ์„ค๋ช…์˜ ๊ณผ์ •์„ ๋”ฐ๋ฅด๋ฉฐ, ํŠนํžˆ โ€˜์œ ๋ ˆ์นดโ€™์™€ ๊ฐ™์€ ์ฒœ์žฌ์ ์ธ ์•„์ด๋””์–ด ์ƒ์„ฑ์˜ ์ˆœ๊ฐ„์„ ๋†’์ด ํ‰๊ฐ€ํ–ˆ์Šต๋‹ˆ๋‹ค. ์ผ€ํ”Œ๋Ÿฌ ์—ญ์‹œ ๋งŽ์€ ์•„์ด๋””์–ด๋ฅผ ์‹œ๋„ํ–ˆ๊ณ , ๊ทธ์ค‘ ์ƒ๋‹น์ˆ˜๋Š” ์‹คํŒจํ–ˆ์Šต๋‹ˆ๋‹ค. ํ•˜์ง€๋งŒ ๊ทธ์˜ ์„ฑ๊ณต์€ ๋ธŒ๋ผํ—ค์˜ ์ •๋ฐ€ํ•œ ๋ฐ์ดํ„ฐ ์ˆ˜์ง‘์ด๋ผ๋Š” โ€˜๊ฒ€์ฆโ€™ ๊ณผ์ •์ด ์žˆ์—ˆ๊ธฐ์— ๊ฐ€๋Šฅํ–ˆ์Šต๋‹ˆ๋‹ค. ๋ธŒ๋ผํ—ค๊ฐ€ ์ด์ „ ๊ด€์ธก๋ณด๋‹ค 10๋ฐฐ๋‚˜ ๋” ์ •ํ™•ํ•œ ๋ฐ์ดํ„ฐ๋ฅผ ์ œ๊ณตํ–ˆ๊ธฐ ๋•Œ๋ฌธ์— ์ผ€ํ”Œ๋Ÿฌ๋Š” ์ž์‹ ์˜ ๋ชจ๋ธ์„ ๋ฐ์ดํ„ฐ์— ๋งž์ถฐ ์กฐ์ •ํ•˜๊ณ  ์ •ํ™•ํ•œ ๊ฒฐ๊ณผ๋ฅผ ์–ป์„ ์ˆ˜ ์žˆ์—ˆ๋˜ ๊ฒƒ์ž…๋‹ˆ๋‹ค.

๊ทธ๋Ÿฌ๋‚˜ 20์„ธ๊ธฐ ๋“ค์–ด ๊ณผํ•™์˜ ํŒจ๋Ÿฌ๋‹ค์ž„์€ ๋ณ€ํ™”ํ–ˆ์Šต๋‹ˆ๋‹ค. ์ด๋ก ๊ณผ ์‹คํ—˜์ด๋ผ๋Š” ๋‘ ๊ฐ€์ง€ ์ฃผ์š” ํŒจ๋Ÿฌ๋‹ค์ž„์— ๋”ํ•ด, ์ˆ˜์น˜ ์‹œ๋ฎฌ๋ ˆ์ด์…˜(numerical simulation)๊ณผ ๋น…๋ฐ์ดํ„ฐ(big data) ๋ถ„์„์ด ๋“ฑ์žฅํ–ˆ์Šต๋‹ˆ๋‹ค. ์ด์ œ๋Š” ๋Œ€๊ทœ๋ชจ ๋ฐ์ดํ„ฐ์…‹์„ ๋จผ์ € ์ˆ˜์ง‘ํ•˜๊ณ , ๊ทธ ์•ˆ์—์„œ ํŒจํ„ด์„ ์ฐพ์•„ ๊ฐ€์„ค์„ ๋„์ถœํ•˜๋Š” ๋ฐฉ์‹์ด ์ ์ฐจ ๋ณดํŽธํ™”๋˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค. ํƒ€์˜ค ๊ต์ˆ˜๋Š” ์ด๋ฅผ โ€œ๊ณผํ•™์  ๋ฐฉ๋ฒ•์ด ๊ฑฐ์˜ ์—ญ์ „๋˜์—ˆ๋‹คโ€๊ณ  ํ‘œํ˜„ํ•ฉ๋‹ˆ๋‹ค. ์ผ€ํ”Œ๋Ÿฌ๋Š” ์–ด์ฉŒ๋ฉด ์ตœ์ดˆ์˜ ๋ฐ์ดํ„ฐ ๊ณผํ•™์ž ์ค‘ ํ•œ ๋ช…์ด์—ˆ์ง€๋งŒ, ๊ทธ์กฐ์ฐจ๋„ ๋ธŒ๋ผํ—ค์˜ ๋ฐ์ดํ„ฐ๋กœ ์‹œ์ž‘ํ•˜๊ธฐ ์ „์— ์ด๋ฏธ ์„ ์ž…๊ฒฌ์ ์ธ ์ด๋ก ์„ ๊ฐ€์ง€๊ณ  ์žˆ์—ˆ์Šต๋‹ˆ๋‹ค. ํ•˜์ง€๋งŒ ์ด์ œ ๋ฐ์ดํ„ฐ์˜ ๊ทœ๋ชจ์™€ ์œ ์šฉ์„ฑ์ด ์••๋„์ ์œผ๋กœ ์ปค์ง€๋ฉด์„œ, ๋ฐ์ดํ„ฐ๋ฅผ ๋จผ์ € ๋ถ„์„ํ•˜๊ณ  ๊ฐ€์„ค์„ ์ฐพ๋Š” ๋ฐฉ์‹์ด ๋”์šฑ ๊ฐ•๋ ฅํ•ด์ง€๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค.

๋‹ค๋งŒ, ๋ฐ์ดํ„ฐ ๊ธฐ๋ฐ˜ ์ ‘๊ทผ๋ฒ•์—๋„ ํ•จ์ •์€ ์žˆ์Šต๋‹ˆ๋‹ค. ์š”ํ•œ ๋ณด๋ฐ(Johann Bode)์˜ ๋ฒ•์น™์ด ์ข‹์€ ์˜ˆ์ž…๋‹ˆ๋‹ค. ๊ทธ๋Š” ์ผ€ํ”Œ๋Ÿฌ์—๊ฒŒ ์˜๊ฐ์„ ๋ฐ›์•„ ํ–‰์„ฑ ๊ฐ„ ๊ฑฐ๋ฆฌ๊ฐ€ ๊ธฐํ•˜๊ธ‰์ˆ˜์ ์œผ๋กœ ์ฆ๊ฐ€ํ•œ๋‹ค๋Š” ํŒจํ„ด์„ ๋ฐœ๊ฒฌํ–ˆ๊ณ , ํ™”์„ฑ๊ณผ ๋ชฉ์„ฑ ์‚ฌ์ด์— ํ–‰์„ฑ์ด ํ•˜๋‚˜ ๋” ์žˆ์„ ๊ฒƒ์ด๋ผ๊ณ  ์˜ˆ์ธกํ–ˆ์Šต๋‹ˆ๋‹ค. ์‹ค์ œ๋กœ ์ฒœ์™•์„ฑ(Uranus)๊ณผ ์†Œํ–‰์„ฑ ์„ธ๋ ˆ์Šค(Ceres)๊ฐ€ ๋ฐœ๊ฒฌ๋˜๋ฉด์„œ ๊ทธ์˜ ์˜ˆ์ธก๊ณผ ์ผ์น˜ํ•˜๋Š” ๋“ฏ ๋ณด์˜€์Šต๋‹ˆ๋‹ค. ๊ทธ๋Ÿฌ๋‚˜ ํ•ด์™•์„ฑ(Neptune)์ด ๋ฐœ๊ฒฌ๋˜๋ฉด์„œ ์ด ํŒจํ„ด์€ ์™„์ „ํžˆ ์–ด๊ธ‹๋‚ฌ๊ณ , ๊ฒฐ๊ตญ ๋ณด๋ฐ์˜ ๋ฒ•์น™์€ โ€˜์ˆซ์ž์  ์šฐ์—ฐ์˜ ์ผ์น˜(numerical fluke)โ€˜์— ๋ถˆ๊ณผํ–ˆ์Œ์ด ๋“œ๋Ÿฌ๋‚ฌ์Šต๋‹ˆ๋‹ค. ์ผ€ํ”Œ๋Ÿฌ๊ฐ€ ์„ธ ๋ฒˆ์งธ ๋ฒ•์น™์„ ์ฒ˜์Œ ๋‘ ๋ฒ•์น™๋งŒํผ ๊ฐ•์กฐํ•˜์ง€ ์•Š์€ ๊ฒƒ๋„, ๋‹จ 6๊ฐœ์˜ ๋ฐ์ดํ„ฐ ํฌ์ธํŠธ๋งŒ์œผ๋กœ ๋„์ถœ๋œ ๊ฒฐ๋ก ์˜ ์‹ ๋ขฐ์„ฑ์— ๋Œ€ํ•ด ๋ณธ๋Šฅ์ ์œผ๋กœ ์กฐ์‹ฌ์Šค๋Ÿฌ์› ๊ธฐ ๋•Œ๋ฌธ์ผ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.

3. AI๊ฐ€ ์ด‰๋ฐœํ•œ ์ƒˆ๋กœ์šด ๋ณ‘๋ชฉ ํ˜„์ƒ: ์•„์ด๋””์–ด ์ƒ์„ฑ์—์„œ ๊ฒ€์ฆ์œผ๋กœ

ํƒ€์˜ค ๊ต์ˆ˜๋Š” AI๊ฐ€ ์•„์ด๋””์–ด ์ƒ์„ฑ์˜ ๋น„์šฉ์„ ๊ฑฐ์˜ 0์— ๊ฐ€๊น๊ฒŒ ๋‚ฎ์ท„๋‹ค๊ณ  ๊ฐ•์กฐํ•ฉ๋‹ˆ๋‹ค. ์ด๋Š” ์ธํ„ฐ๋„ท์ด ํ†ต์‹  ๋น„์šฉ์„ ๋‚ฎ์ถ˜ ๊ฒƒ๊ณผ ์œ ์‚ฌํ•ฉ๋‹ˆ๋‹ค. ํ•˜์ง€๋งŒ ์ด๋Š” ๊ณง ๊ณผํ•™์— ์ƒˆ๋กœ์šด ๋ณ‘๋ชฉ ํ˜„์ƒ(bottleneck)์„ ์ดˆ๋ž˜ํ•ฉ๋‹ˆ๋‹ค. ๊ณผ๊ฑฐ์—๋Š” ์•„๋งˆ์ถ”์–ด ๊ณผํ•™์ž๋“ค์ด ๋ฌด์ˆ˜ํžˆ ๋งŽ์€ ์ด๋ก ์„ ์ œ์‹œํ–ˆ์ง€๋งŒ, ๊ทธ์ค‘ ๋Œ€๋ถ€๋ถ„์€ ๊ฐ€์น˜๊ฐ€ ์—†์—ˆ์Šต๋‹ˆ๋‹ค. ๊ทธ๋ž˜์„œ ๋™๋ฃŒ ์‹ฌ์‚ฌ(peer review)์™€ ์ถœํŒ ์‹œ์Šคํ…œ์€ ๊ณ ํ’ˆ์งˆ ์•„์ด๋””์–ด๋ฅผ ๊ฑธ๋Ÿฌ๋‚ด๊ณ  ๊ฒ€์ฆํ•˜๋Š” ์—ญํ• ์„ ํ–ˆ์Šต๋‹ˆ๋‹ค. ํ•˜์ง€๋งŒ ์ด์ œ AI๊ฐ€ ์—„์ฒญ๋‚œ ๊ทœ๋ชจ๋กœ ๊ฐ€๋Šฅํ•œ ์„ค๋ช…์„ ์ƒ์„ฑํ•ด๋‚ด๋ฉด์„œ, ์ธ๊ฐ„ ๊ฒ€ํ† ์ž๋“ค์€ ์ด๋ฏธ ๊ฐ๋‹นํ•˜๊ธฐ ํž˜๋“  ์ƒํ™ฉ์— ์ฒ˜ํ•ด ์žˆ์Šต๋‹ˆ๋‹ค. ๋งŽ์€ ํ•™์ˆ ์ง€๋“ค์ด AI๊ฐ€ ์ƒ์„ฑํ•œ ๋…ผ๋ฌธ์œผ๋กœ ๋„˜์ณ๋‚˜๊ณ  ์žˆ๋‹ค๊ณ  ๋ณด๊ณ ํ•ฉ๋‹ˆ๋‹ค.

๋ฌธ์ œ๋Š” ์ด ๋ฐฉ๋Œ€ํ•œ ์•„์ด๋””์–ด ํ’€์—์„œ ์–ด๋–ค ์•„์ด๋””์–ด๊ฐ€ ์ง„์งœ ์ง„์ „์„ ์ด๋ฃจ๊ณ , ์–ด๋–ค ์•„์ด๋””์–ด๊ฐ€ ๋ง‰๋‹ค๋ฅธ ๊ณจ๋ชฉ์ด๋‚˜ ์ž˜๋ชป๋œ ๋‹จ์„œ(red herrings)์ธ์ง€๋ฅผ ํ‰๊ฐ€ํ•˜๊ณ  ๊ฒ€์ฆํ•˜๋Š” ๋ฐฉ๋ฒ•์„ ๋Œ€๊ทœ๋ชจ๋กœ ์•Œ์ง€ ๋ชปํ•œ๋‹ค๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค. ๊ฐœ๋ณ„ ๋…ผ๋ฌธ์— ๋Œ€ํ•ด์„œ๋Š” ์ˆ˜๋…„๊ฐ„์˜ ๋…ผ์˜๋ฅผ ํ†ตํ•ด ํ•ฉ์˜์— ๋„๋‹ฌํ•  ์ˆ˜ ์žˆ์ง€๋งŒ, ๋งค์ผ ์ˆ˜์ฒœ ๊ฐœ์˜ ์•„์ด๋””์–ด๊ฐ€ ์ƒ์„ฑ๋  ๋•Œ ์ด๋Ÿฐ ๋ฐฉ์‹์€ ํ†ตํ•˜์ง€ ์•Š์Šต๋‹ˆ๋‹ค.

ํƒ€์˜ค ๊ต์ˆ˜๋Š” ๋ฒจ ์—ฐ๊ตฌ์†Œ(Bell Labs)์—์„œ โ€˜๋น„ํŠธ(bit)โ€™ ๊ฐœ๋…์ด ๋“ฑ์žฅํ–ˆ์„ ๋•Œ๋ฅผ ์˜ˆ๋กœ ๋“ญ๋‹ˆ๋‹ค. ์ˆ˜๋งŽ์€ ๊ธฐ์ˆ  ๋…ผ๋ฌธ ์†์—์„œ โ€˜๋น„ํŠธโ€™๋ผ๋Š” ์•„์ด๋””์–ด๋Š” ํ™•๋ฅ ๋ก , ์ปดํ“จํ„ฐ ๊ณผํ•™ ๋“ฑ ๋‹ค์–‘ํ•œ ๋ถ„์•ผ์— ๊ฑธ์ณ ์—„์ฒญ๋‚œ ํ•จ์˜๋ฅผ ๊ฐ€์กŒ์Šต๋‹ˆ๋‹ค. ๋ฏธ๋ž˜์— AI๊ฐ€ ์ด์™€ ๊ฐ™์€ ํ†ต์ผ ๊ฐœ๋…(unifying concept)์„ ์ œ์‹œํ•œ๋‹ค๋ฉด, ์ˆ˜๋ฐฑ๋งŒ ๊ฐœ์˜ ๋…ผ๋ฌธ ์ค‘์—์„œ ์–ด๋–ป๊ฒŒ ๊ทธ๊ฒƒ์„ ์‹๋ณ„ํ•ด๋‚ผ ์ˆ˜ ์žˆ์„๊นŒ์š”?

์•„์ด๋””์–ด์˜ ๊ฐ€์น˜๋ฅผ ํ‰๊ฐ€ํ•˜๋Š” ๊ฒƒ์€ ๋งค์šฐ ์–ด๋ ต์Šต๋‹ˆ๋‹ค. ๋”ฅ๋Ÿฌ๋‹(Deep Learning) ์ž์ฒด๋„ ์˜ค๋žซ๋™์•ˆ AI์˜ ํ‹ˆ์ƒˆ ๋ถ„์•ผ์˜€๊ณ , ์ฒซ ๋ฒˆ์งธ ์›๋ฆฌ(first principles) ์ถ”๋ก ์ด ์•„๋‹Œ ๋ฐ์ดํ„ฐ ํ•™์Šต๋งŒ์œผ๋กœ ๋‹ต์„ ์–ป๋Š”๋‹ค๋Š” ์ƒ๊ฐ์€ ๋งค์šฐ ๋…ผ๋ž€์ด ๋งŽ์•˜์Šต๋‹ˆ๋‹ค. โ€˜ํŠธ๋žœ์Šคํฌ๋จธ(Transformer)โ€™ ์•„ํ‚คํ…์ฒ˜๋Š” ํ˜„๋Œ€ ๋Œ€๊ทœ๋ชจ ์–ธ์–ด ๋ชจ๋ธ์˜ ๊ธฐ๋ฐ˜์ด ๋˜์—ˆ์ง€๋งŒ, ๋‹ค๋ฅธ ์•„ํ‚คํ…์ฒ˜๊ฐ€ ๋จผ์ € ์ฑ„ํƒ๋˜์—ˆ๋‹ค๋ฉด ๊ทธ๊ฒƒ์ด ํ‘œ์ค€์ด ๋˜์—ˆ์„ ์ˆ˜๋„ ์žˆ์Šต๋‹ˆ๋‹ค. ์•„์ด๋””์–ด์˜ ๊ฒฐ์‹ค ์—ฌ๋ถ€๋Š” ๋ฏธ๋ž˜์™€ ๋ฌธํ™”, ์‚ฌํšŒ์  ์ฑ„ํƒ ์—ฌ๋ถ€์— ๋‹ฌ๋ ค ์žˆ์Šต๋‹ˆ๋‹ค. 10์ง„๋ฒ• ์ฒด๊ณ„๊ฐ€ ๋กœ๋งˆ ์ˆซ์ž๋ณด๋‹ค ํ›จ์”ฌ ์œ ์šฉํ•˜์ง€๋งŒ, 10์ด๋ผ๋Š” ์ˆซ์ž์— ํŠน๋ณ„ํ•œ ์˜๋ฏธ๊ฐ€ ์žˆ๋Š” ๊ฒƒ์€ ์•„๋‹™๋‹ˆ๋‹ค. ๋‹จ์ง€ ๋ชจ๋“  ์‚ฌ๋žŒ์ด ์‚ฌ์šฉํ•˜๊ธฐ ๋•Œ๋ฌธ์— ํ‘œ์ค€ํ™”๋˜์—ˆ๊ณ , ์ด์ œ๋Š” ๊ด€์„ฑ ๋•Œ๋ฌธ์— ๋ฐ”๊พธ๊ธฐ ํž˜๋“ญ๋‹ˆ๋‹ค. ์ฆ‰, ๊ณผํ•™์  ์„ฑ๊ณผ๋ฅผ ๊ณผ๊ฑฐ์™€ ๋ฏธ๋ž˜์˜ ๋งฅ๋ฝ ์—†์ด ์ˆœ์ˆ˜ํ•˜๊ฒŒ ๊ณ ๋ฆฝ์‹œ์ผœ ๊ฐ๊ด€์ ์œผ๋กœ ํ‰๊ฐ€ํ•˜๋Š” ๊ฒƒ์€ ๋ถˆ๊ฐ€๋Šฅํ•ฉ๋‹ˆ๋‹ค.

๋˜ํ•œ, ๊ณผํ•™์  ์ง„๋ณด๋Š” ์ข…์ข… ๊ธฐ์กด์˜ ๊ฐ€์ •์„ ์‚ญ์ œํ•˜๋Š” ๋ฐฉ์‹์œผ๋กœ ์ด๋ฃจ์–ด์ง‘๋‹ˆ๋‹ค. ์ง€๋™์„ค์ด ์˜ค๋žซ๋™์•ˆ ๋ฐ›์•„๋“ค์—ฌ์ง€์ง€ ์•Š์€ ๊ฒƒ์€ ์•„๋ฆฌ์Šคํ† ํ…”๋ ˆ์Šค(Aristotle)์˜ ๋ฌผ๋ฆฌํ•™, ์ฆ‰ ๋ฌผ์ฒด๊ฐ€ ์ž์—ฐ์ ์œผ๋กœ ์ •์ง€ ์ƒํƒœ๋ฅผ ์œ ์ง€ํ•˜๋ ค ํ•œ๋‹ค๋Š” ์ƒ๊ฐ ๋•Œ๋ฌธ์ด์—ˆ์Šต๋‹ˆ๋‹ค. ์ง€๊ตฌ๊ฐ€ ์›€์ง์ธ๋‹ค๋ฉด ์™œ ์šฐ๋ฆฌ๊ฐ€ ๋ชจ๋‘ ๋–จ์–ด์ง€์ง€ ์•Š๋Š”๊ฐ€ ํ•˜๋Š” ์˜๋ฌธ์ด ํ•ด๊ฒฐ๋˜์ง€ ์•Š์•˜๊ธฐ ๋•Œ๋ฌธ์ž…๋‹ˆ๋‹ค. ๋‰ดํ„ด์˜ ์šด๋™ ๋ฒ•์น™์ด ์ •๋ฆฝ๋œ ํ›„์—์•ผ ๋น„๋กœ์†Œ ์ง€๊ตฌ๊ฐ€ ์›€์ง์ธ๋‹ค๋Š” ๊ฐœ๋…์ด ์ดํ•ด๋  ์ˆ˜ ์žˆ์—ˆ์Šต๋‹ˆ๋‹ค. ๋‹ค์œˆ(Darwin)์˜ ์ง„ํ™”๋ก  ์—ญ์‹œ ์ข…(species)์ด ๊ณ ์ •๋˜์–ด ์žˆ์ง€ ์•Š๋‹ค๋Š” ๋‹น์‹œ๋กœ์„œ๋Š” ํŒŒ๊ฒฉ์ ์ธ ๊ฐ€์ •์„ ์‚ญ์ œํ•˜๋ฉด์„œ ์ด๋ฃจ์–ด์ง„ ์œ„๋Œ€ํ•œ ์ง„๋ณด์ž…๋‹ˆ๋‹ค. ํƒ€์˜ค ๊ต์ˆ˜๋Š” ํ˜„์žฌ ์šฐ๋ฆฌ๊ฐ€ โ€˜์ธ๊ฐ„ ์ง€๋Šฅ์ด ์šฐ์ฃผ์˜ ์ค‘์‹ฌโ€™์ด๋ผ๋Š” ์ƒ๊ฐ์—์„œ ๋ฒ—์–ด๋‚˜ ๋‹ค์–‘ํ•œ ์œ ํ˜•์˜ ์ง€๋Šฅ์ด ์กด์žฌํ•จ์„ ์ธ์‹ํ•˜๋Š” โ€˜์ธ์ง€์  ์ฝ”ํŽ˜๋ฅด๋‹ˆ์ฟ ์Šค์  ํ˜๋ช…(cognitive Copernican revolution)โ€˜์„ ๊ฒช๊ณ  ์žˆ๋‹ค๊ณ  ๋งํ•ฉ๋‹ˆ๋‹ค.

4. ๊ณผํ•™์  ์†Œํ†ต๊ณผ ์„ค๋“์˜ ์ค‘์š”์„ฑ

ํƒ€์˜ค ๊ต์ˆ˜๋Š” ๊ณผํ•™์  ๋ฐœ๊ฒฌ์ด ๋‹จ์ˆœํžˆ ์ƒˆ๋กœ์šด ์ด๋ก ์„ ๋งŒ๋“ค๊ณ  ๊ฒ€์ฆํ•˜๋Š” ๊ฒƒ์„ ๋„˜์–ด, ๋‹ค๋ฅธ ์‚ฌ๋žŒ๋“ค์—๊ฒŒ ํšจ๊ณผ์ ์œผ๋กœ ์†Œํ†ตํ•˜๋Š” ๊ฒƒ ๋˜ํ•œ ๋งค์šฐ ์ค‘์š”ํ•˜๋‹ค๊ณ  ๊ฐ•์กฐํ•ฉ๋‹ˆ๋‹ค. ๊ทธ๋Š” ๋‹ค์œˆ์˜ ์ง„ํ™”๋ก ์ด ๋‰ดํ„ด์˜ โ€˜ํ”„๋ฆฐํ‚คํ”ผ์•„(Principia Mathematica)โ€˜๋ณด๋‹ค 2์„ธ๊ธฐ ํ›„์— ๋‚˜์™”์Œ์—๋„ ๋ถˆ๊ตฌํ•˜๊ณ , ๊ฐœ๋…์ ์œผ๋กœ๋Š” ๋” ๊ฐ„๋‹จํ•ด ๋ณด์˜€๋˜ ์ด์œ ๋ฅผ ์„ค๋ช…ํ•ฉ๋‹ˆ๋‹ค. ๋‹ค์œˆ์˜ ์ด๋ก ์€ ์ž์—ฐ์„ ํƒ์— ๋Œ€ํ•œ ์ฆ๊ฑฐ๊ฐ€ ์••๋„์ ์ด์—ˆ์ง€๋งŒ, ์ด๋Š” ๋ˆ„์ ์ ์ด๊ณ  ํšŒ๊ณ ์ ์ธ(cumulative and retrospective) ์„ฑ๊ฒฉ์„ ๋ ์—ˆ์Šต๋‹ˆ๋‹ค. ๋ฐ˜๋ฉด ๋‰ดํ„ด์€ ์ž์‹ ์˜ ๋ฐฉ์ •์‹์„ ์ œ์‹œํ•˜๊ณ  ๋‹ฌ์˜ ๊ถค๋„ ์ฃผ๊ธฐ์™€ ๊ฑฐ๋ฆฌ๊ฐ€ ์ผ์น˜ํ•˜๋Š”์ง€ ํ™•์ธํ•˜๋ฉด ๋ฐ”๋กœ ์ง„์ „์ด ์ด๋ฃจ์–ด์กŒ์Šต๋‹ˆ๋‹ค.

๋‹ค์œˆ์€ ๋†€๋ผ์šด ๊ณผํ•™ ์ปค๋ฎค๋‹ˆ์ผ€์ดํ„ฐ์˜€์Šต๋‹ˆ๋‹ค. ๊ทธ๋Š” ๋ฐฉ์ •์‹ ์—†์ด ํ‰์ดํ•œ ์˜์–ด๋กœ ๊ธ€์„ ์ผ๊ณ , ๋‹ค์–‘ํ•œ ์‚ฌ์‹ค๋“ค์„ ์„ค๋“๋ ฅ ์žˆ๋Š” ๋น„์ „์œผ๋กœ ์ข…ํ•ฉํ–ˆ์Šต๋‹ˆ๋‹ค. ๋ฐ˜๋ฉด ๋‰ดํ„ด์€ ๋ผํ‹ด์–ด๋กœ ๊ธ€์„ ์ผ๊ณ , ์ž์‹ ์˜ ์ด๋ก ์„ ์„ค๋ช…ํ•˜๊ธฐ ์œ„ํ•ด ์ƒˆ๋กœ์šด ์ˆ˜ํ•™ ๋ถ„์•ผ(๋ฏธ์ ๋ถ„)๋ฅผ ์ฐฝ์•ˆํ•ด์•ผ ํ–ˆ์Šต๋‹ˆ๋‹ค. ๊ทธ๋Š” ๋˜ํ•œ ๋‹น์‹œ ๊ณผํ•™์ž๋“ค์ด ํ›จ์”ฌ ๋น„๋ฐ€์Šค๋Ÿฝ๊ณ  ๊ฒฝ์Ÿ์ ์ด์—ˆ๋˜ ์‹œ๋Œ€์˜ ์ธ๋ฌผ์ด์—ˆ์œผ๋ฉฐ, ์ž์‹ ์˜ ๊ฐ€์žฅ ์ค‘์š”ํ•œ ํ†ต์ฐฐ ์ผ๋ถ€๋ฅผ ๊ฒฝ์Ÿ์ž๋“ค์—๊ฒŒ ๋นผ์•—๊ธฐ์ง€ ์•Š๊ธฐ ์œ„ํ•ด ์ˆจ๊ธฐ๊ธฐ๋„ ํ–ˆ์Šต๋‹ˆ๋‹ค. ๋‰ดํ„ด์˜ ์—…์ ์ด ๋„๋ฆฌ ์•Œ๋ ค์ง„ ๊ฒƒ์€ ์ˆ˜์‹ญ ๋…„ ํ›„ ๋‹ค๋ฅธ ๊ณผํ•™์ž๋“ค์ด ๊ทธ์˜ ์ž‘์—…์„ ํ›จ์”ฌ ๊ฐ„๋‹จํ•œ ์šฉ์–ด๋กœ ์„ค๋ช…ํ•œ ์ดํ›„์˜€์Šต๋‹ˆ๋‹ค.

ํƒ€์˜ค ๊ต์ˆ˜๋Š” โ€œ์„œ์ˆ ์˜ ๊ธฐ์ˆ ๊ณผ ์ฃผ์žฅ์„ ํŽผ์น˜๋Š” ๋Šฅ๋ ฅ, ๊ทธ๋ฆฌ๊ณ  ์„œ์‚ฌ๋ฅผ ๋งŒ๋“œ๋Š” ๊ฒƒโ€์ด ๊ณผํ•™์˜ ๋งค์šฐ ์ค‘์š”ํ•œ ๋ถ€๋ถ„์ด๋ผ๊ณ  ๋งํ•ฉ๋‹ˆ๋‹ค. ๋ฐ์ดํ„ฐ๊ฐ€ ๋’ท๋ฐ›์นจ๋˜๋ฉด ๋„์›€์ด ๋˜์ง€๋งŒ, ์‚ฌ๋žŒ๋“ค์ด ์„ค๋“๋˜์–ด์•ผ๋งŒ ์ด๋ก ์„ ๋” ๋ฐœ์ „์‹œํ‚ค๊ฑฐ๋‚˜ ์ดˆ๊ธฐ ํˆฌ์ž๋ฅผ ํ•˜์—ฌ ๋ฐฐ์šฐ๊ณ  ํƒ๊ตฌํ•  ๊ฒƒ์ž…๋‹ˆ๋‹ค. ๊ทธ๋Š” AI๊ฐ€ ์•„์ง ์„ค๋“๋ ฅ์„ ๊ฐ•ํ™”ํ•™์Šต(Reinforcement Learning)ํ•˜๊ธฐ ์–ด๋ ต๋‹ค๋Š” ์ ์„ ์ง€์ ํ•˜๋ฉฐ, ์ด๊ฒƒ์ด ์–ด์ฉŒ๋ฉด ๊ณผํ•™์˜ ์˜์›ํ•œ ์ธ๊ฐ„์  ์ธก๋ฉด์œผ๋กœ ๋‚จ์„ ์ˆ˜ ์žˆ๋‹ค๊ณ  ์˜ˆ์ธกํ•ฉ๋‹ˆ๋‹ค.

5. ์ˆ˜ํ•™ ๋ถ„์•ผ AI์˜ ํ˜„์žฌ์™€ ๋ฏธ๋ž˜: ํญ๊ณผ ๊นŠ์ด

ํƒ€์˜ค ๊ต์ˆ˜๋Š” ๋“œ์™€์ผ€์‹œ ํŒŒํ…”์˜ ์‹œ๋ฆฌ์ฆˆ โ€œ์šฐ์ฃผ ๊ฑฐ๋ฆฌ ์‚ฌ๋‹ค๋ฆฌ(cosmic distance ladder)โ€œ๋ฅผ ์–ธ๊ธ‰ํ•˜๋ฉฐ, โ€œ๋งŽ์€ ๋ถ„์•ผ์—์„œ ์—ฐ์—ญ์  ์˜ค๋ฒ„ํ–‰(deductive overhang)์ด ์‚ฌ๋žŒ๋“ค์ด ๊นจ๋‹ซ๋Š” ๊ฒƒ๋ณด๋‹ค ํ›จ์”ฌ ํด ์ˆ˜ ์žˆ๋‹คโ€๋Š” ์ ์— ๋™์˜ํ•ฉ๋‹ˆ๋‹ค. ์ฆ‰, ๋ฌธ์ œ๋ฅผ ์—ฐ๊ตฌํ•˜๋Š” ์˜ฌ๋ฐ”๋ฅธ ํ†ต์ฐฐ๋ ฅ๋งŒ ์žˆ๋‹ค๋ฉด, ์„ธ์ƒ์— ๋Œ€ํ•ด ํ›จ์”ฌ ๋” ๋งŽ์€ ๊ฒƒ์„ ์•Œ ์ˆ˜ ์žˆ๋‹ค๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค. ์ฒœ๋ฌธํ•™์€ ๋ฐ์ดํ„ฐ๊ฐ€ ๋ณ‘๋ชฉ ํ˜„์ƒ์ด์—ˆ๊ธฐ ๋•Œ๋ฌธ์— ๋ฐ์ดํ„ฐ๋ฅผ ๋ถ„์„ํ•˜๊ณ  ์ •๋ณด์—์„œ ๋งˆ์ง€๋ง‰ ํ•œ ๋ฐฉ์šธ๊นŒ์ง€ ์งœ๋‚ด๋Š” ๋ฐ ๊ฐ€์žฅ ๋จผ์ € ์•ž์žฅ์„  ๊ณผํ•™ ๋ถ„์•ผ ์ค‘ ํ•˜๋‚˜์˜€์Šต๋‹ˆ๋‹ค. ํ€€ํŠธ ํ—ค์ง€ํŽ€๋“œ(quant hedge funds)์—์„œ ์ฒœ๋ฌธํ•™ ๋ฐ•์‚ฌ๋ฅผ ์„ ํ˜ธํ•˜๋Š” ์ด์œ ๋„ ์—ฌ๊ธฐ์— ์žˆ์Šต๋‹ˆ๋‹ค.

์ตœ๊ทผ AI๋Š” ์ˆ˜ํ•™ ๋ถ„์•ผ์—์„œ ๋†€๋ผ์šด ์ง„์ „์„ ๋ณด์ด๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค. ํƒ€์˜ค ๊ต์ˆ˜๋Š” ์ตœ๊ทผ ๋ช‡ ๋‹ฌ ๋™์•ˆ AI ํ”„๋กœ๊ทธ๋žจ๋“ค์ด ์—๋ฅด๋˜์‹œ ๋ฌธ์ œ(Erdล‘s problems) 1,100์—ฌ ๊ฐœ ์ค‘ ์•ฝ 50๊ฐœ๋ฅผ ํ•ด๊ฒฐํ–ˆ๋‹ค๊ณ  ์–ธ๊ธ‰ํ–ˆ์Šต๋‹ˆ๋‹ค. ํ•˜์ง€๋งŒ ๊ทธ๋Š” โ€œํ˜„์žฌ ๋‚ฎ๊ฒŒ ๋งค๋‹ฌ๋ฆฐ ์—ด๋งค(low-hanging fruit)๋Š” ์ด๋ฏธ ๋‹ค ๋”ฐ๋ฒ„๋ฆฐ ์ƒํƒœโ€์ด๋ฉฐ, ์ˆœ์ˆ˜ AI ์†”๋ฃจ์…˜์ด ๊ธ‰๊ฒฉํžˆ ์ค„์–ด๋“ค๊ณ  ์žˆ๋‹ค๊ณ  ๋งํ•ฉ๋‹ˆ๋‹ค. ์ด์ œ๋Š” AI๊ฐ€ ๊ฐ€๋Šฅํ•œ ์ฆ๋ช… ์ „๋žต์„ ์ƒ์„ฑํ•˜๊ณ , ๋‹ค๋ฅธ AI ๋„๊ตฌ๊ฐ€ ์ด๋ฅผ ๋น„ํŒํ•˜๊ณ , ์žฌ์ž‘์„ฑํ•˜๊ฑฐ๋‚˜, ์ˆ˜์น˜ ๋ฐ์ดํ„ฐ๋ฅผ ์ƒ์„ฑํ•˜๊ณ , ๋ฌธํ—Œ ์กฐ์‚ฌ๋ฅผ ํ•˜๋Š” ๋“ฑ, โ€œ๋งŽ์€ ์ธ๊ฐ„๊ณผ ๋งŽ์€ AI ๋„๊ตฌ๋“ค ๊ฐ„์˜ ์ง€์†์ ์ธ ๋Œ€ํ™”โ€๋ฅผ ํ†ตํ•ด ๋ฌธ์ œ๊ฐ€ ํ•ด๊ฒฐ๋˜๋Š” ์–‘์ƒ์ž…๋‹ˆ๋‹ค.

ํƒ€์˜ค ๊ต์ˆ˜๋Š” AI ๋„๊ตฌ๋ฅผ โ€œ2๋ฏธํ„ฐ๋ฅผ ์ ํ”„ํ•  ์ˆ˜ ์žˆ๋Š” ์ ํ•‘ ๋จธ์‹ โ€์— ๋น„์œ ํ•ฉ๋‹ˆ๋‹ค. ์ด๋“ค์€ ๋•Œ๋•Œ๋กœ ์ž˜๋ชป๋œ ๋ฐฉํ–ฅ์œผ๋กœ ์ ํ”„ํ•˜๊ฑฐ๋‚˜ ์ถ”๋ฝํ•˜๊ธฐ๋„ ํ•˜์ง€๋งŒ, ๋•Œ๋กœ๋Š” ์ธ๊ฐ„์ด ๋„๋‹ฌํ•  ์ˆ˜ ์—†์—ˆ๋˜ ๊ฐ€์žฅ ๋‚ฎ์€ ๋ฒฝ์˜ ๊ผญ๋Œ€๊ธฐ์— ๋„๋‹ฌํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค. ํ•˜์ง€๋งŒ ์ด ๋„๊ตฌ๋“ค์€ ๋ถ€๋ถ„์ ์ธ ์ง„์ „์„ ๋งŒ๋“ค๊ฑฐ๋‚˜, ๋จผ์ € ์ง‘์ค‘ํ•ด์•ผ ํ•  ์ค‘๊ฐ„ ๋‹จ๊ณ„๋ฅผ ์‹๋ณ„ํ•˜๋Š” ๋ฐ๋Š” ๋งค์šฐ ์„œํˆด๋‹ค๋Š” ํ•œ๊ณ„๊ฐ€ ์žˆ์Šต๋‹ˆ๋‹ค.

๊ทธ๋Š” AI์˜ ์ง„๋ณด์— ๋Œ€ํ•ด โ€œ๋น„๊ด€๋ก ์ (bearish)์ด๋ฉด์„œ ๋™์‹œ์— ๋‚™๊ด€๋ก ์ (bullish)โ€œ์ธ ๋‘ ๊ฐ€์ง€ ๊ด€์ ์„ ์ œ์‹œํ•ฉ๋‹ˆ๋‹ค. ๋น„๊ด€๋ก ์  ๊ด€์ ์€ AI๊ฐ€ ์•„์ง ์ธ๊ฐ„ ์ „๋ฌธ๊ฐ€๊ฐ€ ๋„๋‹ฌํ•˜๋Š” ์ˆ˜์ค€์˜ โ€œ๋†’์€ ๋ฒฝโ€์— ๋„๋‹ฌํ•˜์ง€ ๋ชปํ•œ๋‹ค๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค. ํ•˜์ง€๋งŒ ๋‚™๊ด€๋ก ์  ๊ด€์ ์€ AI๊ฐ€ ์ผ๋‹จ ํŠน์ • ์ˆ˜์ค€์— ๋„๋‹ฌํ•˜๋ฉด, โ€œ๊ทธ ์ˆ˜์ค€์—์„œ ๊ฐ€๋Šฅํ•œ ๋ชจ๋“  ๋ฌธ์ œ๋ฅผ ํ•ด๊ฒฐํ•  ์ˆ˜ ์žˆ๋Š” ๊ฐ•๋ ฅํ•œ ์†์„ฑโ€์„ ๊ฐ€์ง€๊ณ  ์žˆ๋‹ค๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค. ์ธ๊ฐ„์€ ์ˆ˜๋ฐฑ๋งŒ ๊ฐœ์˜ ๋ฌธ์ œ์— ๋Œ€ํ•ด ์ˆ˜๋ฐฑ ๋…„์˜ ์—ฐ๊ตฌ๋ฅผ ๋™์‹œ์— ์ˆ˜ํ–‰ํ•  ์ˆ˜ ์—†์ง€๋งŒ, AI๋Š” ๊ฐ€๋Šฅํ•˜๋‹ค๋Š” ๊ฒƒ์ด์ฃ .

ํƒ€์˜ค ๊ต์ˆ˜๋Š” AI๊ฐ€ **โ€˜ํญ(breadth)โ€˜**์— ๋›ฐ์–ด๋‚˜๊ณ  ์ธ๊ฐ„์€ **โ€˜๊นŠ์ด(depth)โ€˜**์— ๋›ฐ์–ด๋‚˜๋‹ค๊ณ  ๊ฒฐ๋ก  ๋‚ด๋ฆฝ๋‹ˆ๋‹ค. ๊ทธ๋Š” AI์˜ ํญ๋„“์€ ๋Šฅ๋ ฅ์„ ์ตœ๋Œ€ํ•œ ํ™œ์šฉํ•˜๊ธฐ ์œ„ํ•ด ๊ณผํ•™ ์ˆ˜ํ–‰ ๋ฐฉ์‹์„ ์žฌ์„ค๊ณ„ํ•ด์•ผ ํ•œ๋‹ค๊ณ  ์ฃผ์žฅํ•ฉ๋‹ˆ๋‹ค. โ€œํ•˜๋‚˜ ๋˜๋Š” ๋‘ ๊ฐœ์˜ ์ •๋ง ๊นŠ๊ณ  ์ค‘์š”ํ•œ ๋ฌธ์ œ๋ณด๋‹ค๋Š” ๋งค์šฐ ๊ด‘๋ฒ”์œ„ํ•œ ๋ฌธ์ œ๋“ค์„ ๋‹ค๋ฃจ๋Š” ๋ฐ ๋” ๋งŽ์€ ๋…ธ๋ ฅ์„ ๊ธฐ์šธ์—ฌ์•ผ ํ•œ๋‹คโ€๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค. AI๋Š” ์ƒˆ๋กœ์šด ๊ณผํ•™ ๋ถ„์•ผ๋ฅผ ํƒ์ƒ‰ํ•˜๊ณ , ์‰ฌ์šด ๊ด€์ธก์„ ๋ชจ๋‘ ์ˆ˜ํ–‰ํ•˜์—ฌ โ€˜๋‚œ์ด๋„์˜ ์„ฌโ€™์„ ์‹๋ณ„ํ•œ ๋‹ค์Œ, ์ธ๊ฐ„ ์ „๋ฌธ๊ฐ€๊ฐ€ ๊ทธ๊ณณ์— ์™€์„œ ์ž‘์—…ํ•  ์ˆ˜ ์žˆ๋„๋ก ํ•  ๊ฒƒ์ž…๋‹ˆ๋‹ค. ๊ทธ๋Š” โ€œ๋ฏธ๋ž˜์—๋Š” ๋งค์šฐ ์ƒํ˜ธ ๋ณด์™„์ ์ธ ๊ณผํ•™์„ ๋ณด๊ฒŒ ๋  ๊ฒƒโ€์ด๋ฉฐ, โ€œ๊ณผํ•™์€ ๊ทธ ์ดํ›„๋กœ ์•Œ์•„๋ณผ ์ˆ˜ ์—†๊ฒŒ ๋ณ€ํ•  ๊ฒƒโ€์ด๋ผ๊ณ  ์ „๋งํ•ฉ๋‹ˆ๋‹ค.

ํ”„๋กœ๊ทธ๋ž˜๋จธ๋“ค์ด AI ๋„๊ตฌ ๋•๋ถ„์— ์ƒ์‚ฐ์„ฑ์ด ํฌ๊ฒŒ ํ–ฅ์ƒ๋˜์—ˆ๋‹ค๊ณ  ๋А๋ผ๋Š” ๊ฒƒ์ฒ˜๋Ÿผ, ์ˆ˜ํ•™์—์„œ๋„ ์œ ์‚ฌํ•œ ๋ณ€ํ™”๊ฐ€ ์ผ์–ด๋‚  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค. ํƒ€์˜ค ๊ต์ˆ˜๋Š” ์ˆ˜ํ•™์—์„œ ๋ฌธ์ œ ํ•ด๊ฒฐ ๊ณผ์ • ์ž์ฒด๊ฐ€ ๋ฌธ์ œ ์ž์ฒด๋ณด๋‹ค ๋” ์ค‘์š”ํ•˜๋‹ค๊ณ  ๋งํ•ฉ๋‹ˆ๋‹ค. AI๋Š” ์ƒ์šฉ๊ตฌ(boilerplate) ์ฝ”๋“œ์™€ ๊ฐ™์€ ๋ณด์กฐ์ ์ธ ์ž‘์—…๋“ค์„ ๊ฐ€์†ํ™”ํ•˜์—ฌ ๊ทธ์˜ ๋…ผ๋ฌธ์— ๋” ๋งŽ์€ ์ฝ”๋“œ๋‚˜ ๊ทธ๋ฆผ์„ ํฌํ•จํ•  ์ˆ˜ ์žˆ๊ฒŒ ํ•ด์ฃผ์—ˆ์Šต๋‹ˆ๋‹ค. ๊ทธ๋Š” AI๊ฐ€ ์—†๋‹ค๋ฉด ์˜ค๋Š˜๋‚  ๊ทธ๊ฐ€ ์“ฐ๋Š” ๋…ผ๋ฌธ์˜ ์œ ํ˜•์€ 5๋ฐฐ ๋” ์˜ค๋ž˜ ๊ฑธ๋ฆด ๊ฒƒ์ด๋ผ๊ณ  ๋งํ•ฉ๋‹ˆ๋‹ค. ํ•˜์ง€๋งŒ ์ˆ˜ํ•™ ๋ฌธ์ œ์˜ ๊ฐ€์žฅ ์–ด๋ ค์šด ํ•ต์‹ฌ ๋ถ€๋ถ„์„ ํ•ด๊ฒฐํ•˜๋Š” ๋ฐ๋Š” ์•„์ง AI์˜ ๋„์›€์„ ๋ฐ›์ง€ ์•Š๊ณ  ์—ฌ์ „ํžˆ ํŽœ๊ณผ ์ข…์ด๋ฅผ ์‚ฌ์šฉํ•ฉ๋‹ˆ๋‹ค.

๊ทธ๋Š” AI ์œ ํ˜• ๋„๊ตฌ๋“ค์ด โ€œ์ˆ˜ํ•™์˜ ์‹คํ—˜์  ์ธก๋ฉด์„ ํ˜์‹ ํ•  ๊ฒƒโ€์ด๋ผ๊ณ  ์˜ˆ์ธกํ•ฉ๋‹ˆ๋‹ค. ๊ฐœ๋ณ„ ๋ฌธ์ œ์™€ ํ•ด๊ฒฐ ๊ณผ์ •์— ๋œ ์‹ ๊ฒฝ ์“ฐ๊ณ , ๋ฌด์—‡์ด ํšจ๊ณผ๊ฐ€ ์žˆ๊ณ  ๋ฌด์—‡์ด ํšจ๊ณผ๊ฐ€ ์—†๋Š”์ง€์— ๋Œ€ํ•œ ๋Œ€๊ทœ๋ชจ ๋ฐ์ดํ„ฐ๋ฅผ ์ˆ˜์ง‘ํ•˜๋Š” ๋ฐ ์ง‘์ค‘ํ•  ๊ฒƒ์ด๋ผ๋Š” ๊ฒ๋‹ˆ๋‹ค. โ€œ์ˆ˜ํ•™์„ ๋Œ€๊ทœ๋ชจ๋กœ ์ˆ˜ํ–‰ํ•˜๋Š” ์•„์ด๋””์–ด(the idea of doing mathematics at scale)โ€œ๋Š” ์•„์ง ์ดˆ๊ธฐ ๋‹จ๊ณ„์ด์ง€๋งŒ, ์ด๊ฒƒ์ด AI๊ฐ€ ์ด ๋ถ„์•ผ๋ฅผ ํ˜์‹ ํ•  ์ง€์ ์ด๋ผ๊ณ  ๊ทธ๋Š” ๋งํ•ฉ๋‹ˆ๋‹ค.

ํƒ€์˜ค ๊ต์ˆ˜๋Š” AI์˜ ์ง„๋ณด๋ฅผ โ€œ๋†€๋ž๊ณ ๋„ ์‹ค๋ง์Šค๋Ÿฌ์šด(amazing and disappointing)โ€ ๊ฐ์ •์œผ๋กœ ๋ฐ”๋ผ๋ด…๋‹ˆ๋‹ค. ์‚ฌ๋žŒ๋“ค์€ ๋น ๋ฅด๊ฒŒ ์ ์‘ํ•ฉ๋‹ˆ๋‹ค. 20๋…„ ์ „ ๊ตฌ๊ธ€ ๊ฒ€์ƒ‰์ด ์ฒ˜์Œ ๋‚˜์™”์„ ๋•Œ ๋†€๋ผ์› ์ง€๋งŒ, ๋ช‡ ๋…„ ํ›„์—๋Š” ๋‹น์—ฐํ•˜๊ฒŒ ์—ฌ๊ฒจ์กŒ์Šต๋‹ˆ๋‹ค. 2026๋…„ ์ˆ˜์ค€์˜ AI๋Š” 2021๋…„์—๋Š” ๊ฒฝ์ด๋กœ์› ์„ ๊ฒƒ์ด์ง€๋งŒ, ์ด์ œ๋Š” ์–ผ๊ตด ์ธ์‹, ์ž์—ฐ์–ด ์ฒ˜๋ฆฌ, ๋Œ€ํ•™ ์ˆ˜์ค€์˜ ์ˆ˜ํ•™ ๋ฌธ์ œ ํ•ด๊ฒฐ ๋“ฑ์ด ๋‹น์—ฐํ•˜๊ฒŒ ์—ฌ๊ฒจ์ง„๋‹ค๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค.

๊ทธ๋Š” 2023๋…„์— 2026๋…„๊นŒ์ง€ AI๊ฐ€ ์ˆ˜ํ•™ ๋ถ„์•ผ์—์„œ โ€œ์‹ ๋ขฐํ•  ์ˆ˜ ์žˆ๋Š” ๊ณต๋™ ์ €์ž(trustworthy co-author) ๊ฐ™์€ ์กด์žฌ๊ฐ€ ๋  ๊ฒƒโ€์ด๋ผ๊ณ  ์˜ˆ์ธกํ–ˆ๊ณ , ํ˜„์žฌ๋กœ์„œ๋Š” ์ด ์˜ˆ์ธก์ด ๊ฝค ์ž˜ ๋งž์•„๊ฐ€๊ณ  ์žˆ๋‹ค๊ณ  ํ‰๊ฐ€ํ•ฉ๋‹ˆ๋‹ค.

๋งˆ์ง€๋ง‰์œผ๋กœ ํƒ€์˜ค ๊ต์ˆ˜๋Š” โ€˜์ธ๊ณต์ ์ธ ์˜๋ฆฌํ•จ(artificial cleverness)โ€˜๊ณผ โ€˜์ธ๊ณต ์ง€๋Šฅ(artificial intelligence)โ€˜์„ ๊ตฌ๋ถ„ํ•ฉ๋‹ˆ๋‹ค. ์ง€๋Šฅ์€ ์ •์˜ํ•˜๊ธฐ ์–ด๋ ต์ง€๋งŒ, ๊ทธ๊ฐ€ ๋งํ•˜๋Š” ์ง€๋Šฅ์€ โ€œ๋ฌธ์ œ ํ•ด๊ฒฐ์„ ์œ„ํ•ด ํ˜‘๋ ฅํ•  ๋•Œ ์•„์ด๋””์–ด๋ฅผ ๋Š์ž„์—†์ด ๊ฐœ์„ ํ•˜๊ณ , ์ฒด๊ณ„์ ์œผ๋กœ ๋ฌด์—‡์ด ์ž‘๋™ํ•˜๊ณ  ๋ฌด์—‡์ด ์ž‘๋™ํ•˜์ง€ ์•Š๋Š”์ง€ ํŒŒ์•…ํ•˜๋ฉฐ, ์ง„์ „์˜ ๊ธธ์„ ์ฐพ์•„๋‚ด๋Š” ์ ์‘์„ฑ(adaptivity)๊ณผ ์ง€์†์ ์ธ ๊ฐœ์„ โ€์„ ์˜๋ฏธํ•ฉ๋‹ˆ๋‹ค. AI๋Š” ์ด๋Ÿฌํ•œ ๊ณผ์ •์„ ๋ชจ๋ฐฉํ•  ์ˆ˜ ์žˆ์ง€๋งŒ, โ€œ์กฐ๊ธˆ์”ฉ ์ ํ”„ํ•ด์„œ ์†์žก์ด๋ฅผ ์žก๊ณ , ๊ฑฐ๊ธฐ์— ๋จธ๋ฌผ๋Ÿฌ ๋‹ค๋ฅธ ์‚ฌ๋žŒ๋“ค์„ ๋Œ์–ด์˜ฌ๋ฆฌ๊ณ , ๊ฑฐ๊ธฐ์„œ๋ถ€ํ„ฐ ๋‹ค์‹œ ์ ํ”„ํ•˜๋Š”โ€ ๋ˆ„์ ์ ์ธ ๊ณผ์ •์€ ์•„์ง ๋ถ€์กฑํ•ฉ๋‹ˆ๋‹ค. AI๋Š” ์‹œํ–‰์ฐฉ์˜ค์™€ ๋ฐ˜๋ณต, ์ฆ‰ โ€˜๋ฌด์ฐจ๋ณ„ ๋Œ€์ž…(brute force)โ€˜์— ๊ฐ€๊น์Šต๋‹ˆ๋‹ค. AI๊ฐ€ ๋ฌธ์ œ๋ฅผ ํ•ด๊ฒฐํ•˜๋”๋ผ๋„, ๊ทธ ์ž์ฒด์˜ ์ˆ˜ํ•™์  ์ดํ•ด๊ฐ€ ์ง„์ „๋œ ๊ฒƒ์€ ์•„๋‹ˆ๋ฉฐ, ์ƒˆ๋กœ์šด ์„ธ๋Œ€์˜ ํ›ˆ๋ จ ๋ฐ์ดํ„ฐ์— 0.001%์˜ ๊ธฐ์—ฌ๋ฅผ ํ•  ๋ฟ, ์ƒˆ๋กœ์šด ์„ธ์…˜์—์„œ๋Š” ๋ฐฉ๊ธˆ ํ•œ ์ผ์„ ์žŠ์–ด๋ฒ„๋ฆฐ๋‹ค๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค.

๊ฒฐ๋ก 

ํ…Œ๋ Œ์Šค ํƒ€์˜ค ๊ต์ˆ˜์˜ ํ†ต์ฐฐ์€ AI๊ฐ€ ๊ณผํ•™์  ๋ฐœ๊ฒฌ์˜ ๋ณธ์งˆ์„ ๊ทผ๋ณธ์ ์œผ๋กœ ๋’คํ”๋“ค๊ณ  ์žˆ์Œ์„ ๋ณด์—ฌ์ค๋‹ˆ๋‹ค. ์•„์ด๋””์–ด ์ƒ์„ฑ์˜ ๋ฌธํ„ฑ์„ ๋‚ฎ์ถ”๊ณ  โ€˜ํญโ€™์˜ ํƒ์ƒ‰์„ ๊ฐ€๋Šฅํ•˜๊ฒŒ ํ•˜๋Š” AI๋Š” ๊ณผํ•™์˜ ์ƒˆ๋กœ์šด ์ง€ํ‰์„ ์—ด๊ณ  ์žˆ์ง€๋งŒ, ๋™์‹œ์— ๊ฒ€์ฆ, ํ‰๊ฐ€, ๊ทธ๋ฆฌ๊ณ  ์ง„์ •ํ•œ ํ†ต์ผ ๊ฐœ๋…์„ ์‹๋ณ„ํ•˜๋Š” ์ƒˆ๋กœ์šด ๊ณผ์ œ๋ฅผ ์•ˆ๊ฒจ์ฃผ๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค. ๊ณผํ•™์  ์†Œํ†ต๊ณผ ์„ค๋“์ด๋ผ๋Š” ์ธ๊ฐ„ ๊ณ ์œ ์˜ ์˜์—ญ ๋˜ํ•œ ์—ฌ์ „ํžˆ ์ค‘์š”ํ•ฉ๋‹ˆ๋‹ค. AI์™€ ์ธ๊ฐ„์˜ ์ƒํ˜ธ ๋ณด์™„์ ์ธ ๊ด€๊ณ„ ์†์—์„œ, ์šฐ๋ฆฌ๋Š” ์ด์ „์— ์ƒ์ƒํ•  ์ˆ˜ ์—†์—ˆ๋˜ ๋ฐฉ์‹์œผ๋กœ ๊ณผํ•™์„ ์žฌ์ •์˜ํ•˜๊ณ , ๊ถ๊ทน์ ์œผ๋กœ๋Š” โ€˜๊นŠ์ดโ€™์™€ โ€˜ํญโ€™์„ ๋ชจ๋‘ ์•„์šฐ


โ€œInside Palantir: Building Software That Matters | Shyam Sankar on a16zโ€ โ€” a16z ๊ธฐ๋ฐ˜ ๊ธฐ์‚ฌ ์›๋ณธ ์˜์ƒ ๋ณด๊ธฐ

์œ„ํ˜‘๋ฐ›๋Š” ๋ฏธ๊ตญ, AI๋กœ ์žฌ๋ฌด์žฅํ•˜๋ผ: ํŒ”๋ž€ํ‹ฐ์–ด ์ƒด ์‚ฐ์นด๋ฅด๊ฐ€ ๋งํ•˜๋Š” ๊ตญ๋ฐฉ ํ˜์‹ ๊ณผ โ€˜์ „ ๊ตญ๊ฐ€์  ๋™์›โ€™์˜ ์‹œ๋Œ€

์œ„๊ธฐ ์† ๋ฏธ๊ตญ์˜ ๊ฐ์„ฑ: ๊ตญ๋ฐฉ ํ˜์‹ ์˜ ๋ชฉ์†Œ๋ฆฌ

์„ธ๊ณ„ ๊ฐ์ง€์—์„œ ๋ฒŒ์–ด์ง€๋Š” ์‚ฌ๊ฑด๋“ค์€ ์—ฌ์ „ํžˆ ์ž”ํ˜นํ•œ ์•ผ๋งŒ์ฃผ์˜๊ฐ€ ์กด์žฌํ•˜๋ฉฐ, ์•…์ด ์‹ค์žฌํ•œ๋‹ค๋Š” ๋ƒ‰ํ˜นํ•œ ํ˜„์‹ค์„ ์ƒ๊ธฐ์‹œํ‚จ๋‹ค. ํ•œ ๋‚˜๋ผ๊ฐ€ ์ „์Ÿ์— ๋Œ์ž…ํ•  ๋•Œ, ๊ตญ๋ฐฉ๋ถ€๋งŒ์ด ์ „์Ÿ์„ ์ˆ˜ํ–‰ํ•˜๋Š” ๊ฒƒ์œผ๋กœ๋Š” ์ถฉ๋ถ„์น˜ ์•Š๋‹ค. ์ง„์ •ํ•œ ์ „์Ÿ์€ ์˜จ ๋‚˜๋ผ๊ฐ€ ํ•จ๊ป˜ ์น˜๋ฅด๋Š” ๊ฒƒ์ด๋ฉฐ, ๋ฏธ๊ตญ ๊ตญ๋ฏผ์ด ์ด๋Ÿฌํ•œ ์œ„๊ธฐ์— ๋Œ€์‘ํ•  ์—ญ๋Ÿ‰์ด ์—†๋‹ค๋Š” ์ƒ๊ฐ์€ ๋ฏฟ๊ธฐ ์–ด๋ ต๋‹ค. ํŒ”๋ž€ํ‹ฐ์–ด(Palantir)์˜ ์ƒด ์‚ฐ์นด๋ฅด(Shyam Sankar)๋Š” ๋ฏธ๊ตญ์˜ ๊ฐ€์žฅ ํฐ ์œ„ํ—˜์€ ์™ธ๋ถ€์˜ ๊ณต๊ฒฉ์ด ์•„๋‹Œ โ€˜์ž๋ฉธโ€™์ด๋ผ๊ณ  ๊ฒฝ๊ณ ํ•˜๋ฉฐ, AI ๊ฒฝ์Ÿ์—์„œ ์Šน๋ฆฌํ•˜๊ณ  ํŠนํžˆ ๋ฌผ๋ฆฌ์  AI ๋ฐ ๋กœ๋ด‡ ๊ณตํ•™ ๋ถ„์•ผ์—์„œ ์•ž์„œ๋‚˜๊ฐ€๊ธฐ ์œ„ํ•œ ๊ทผ๋ณธ์ ์ธ ๋ณ€ํ™”๋ฅผ ์ด‰๊ตฌํ•œ๋‹ค.

์˜ค๋žซ๋™์•ˆ ํŒ”๋ž€ํ‹ฐ์–ด์˜ ๊ณต๋™ ์ฐฝ๋ฆฝ์ž์ด์ž โ€˜ํ•ด๊ฒฐ์‚ฌ(fixer)โ€™๋กœ ์•Œ๋ ค์กŒ๋˜ ์ƒด ์‚ฐ์นด๋ฅด๋Š” ์ตœ๊ทผ ๋ช‡ ๋…„ ์‚ฌ์ด ๋Œ€์ค‘์  ์ธ๋ฌผ๋กœ ๋ถ€์ƒํ–ˆ๋‹ค. ๊ทธ์˜ ๋™๋ฃŒ๋“ค์€ ๊ทธ๋ฅผ โ€œ๊ฒฝ๋ ฅ์˜ ์€์ธโ€์œผ๋กœ ์นญ์†กํ•˜๋ฉฐ, ํŒ”๋ž€ํ‹ฐ์–ด์— ์˜์ž…ํ•˜๊ณ  ์ง€์›ํ–ˆ์„ ๋ฟ ์•„๋‹ˆ๋ผ, ์ƒˆ๋กœ์šด ์‹œ์ž‘์„ ํ•  ์ˆ˜ ์žˆ๋„๋ก โ€œ๋‚ ๊ฐœ๋ฅผ ๋‹ฌ์•„์ฃผ์—ˆ๋‹คโ€๊ณ  ์ž…์„ ๋ชจ์€๋‹ค. ์ˆ˜๋…„ ๋™์•ˆ โ€˜๋ฐฐํ›„ ์ธ๋ฌผโ€™๋กœ ํ™œ๋™ํ•˜๋˜ ๊ทธ๊ฐ€ ์ „๋ฉด์— ๋‚˜์„œ๊ฒŒ ๋œ ๊ฒฐ์ •์ ์ธ ๊ณ„๊ธฐ๋Š” 2022๋…„์— ๋ฐœํ‘œํ•œ ๊ตญ๋ฐฉ ๊ฐœํ˜์— ๋Œ€ํ•œ ์ค‘์š”ํ•œ ๊ธ€์ธ โ€˜ํผ์ŠคํŠธ ๋ธŒ๋ ‰ํผ์ŠคํŠธ(First Breakfast)โ€˜์˜€๋‹ค. ์ด ๊ธ€์€ ๋ฏธ๊ตญ์ด ๋‚˜์•„๊ฐ€์•ผ ํ•  ๋ฐฉํ–ฅ์— ๋Œ€ํ•œ ๊ฐ•๋ ฅํ•œ ๋ชฉ์†Œ๋ฆฌ๊ฐ€ ๋˜์—ˆ๋‹ค.

์‚ฐ์นด๋ฅด๋Š” ์ด๋Ÿฌํ•œ ๋ณ€ํ™”์˜ ๋™๊ธฐ๋ฅผ โ€œ์ ˆ๋ฐ•ํ•จ๊ณผ ๋‚™๊ด€์ฃผ์˜๊ฐ€ ๋ฐ˜๋ฐ˜ ์„ž์ธ ํ–‰๋™โ€์ด๋ผ๊ณ  ์„ค๋ช…ํ•œ๋‹ค. ์ˆ˜๋…„๊ฐ„ ๊ตญ๋ฐฉ๋ถ€ ๋‚ด๋ถ€์—์„œ ๊ตญ๋ฐฉ ์‹œ์Šคํ…œ์ด ์„œ์„œํžˆ ์•…ํ™”๋˜๋Š” ๊ฒƒ์„ ๋ชฉ๊ฒฉํ•˜๋ฉฐ, ๊ทธ๋Š” โ€˜๋“๋Š” ๋ฌผ ์† ๊ฐœ๊ตฌ๋ฆฌ(frog boil)โ€˜์ฒ˜๋Ÿผ ์ ์ง„์ ์œผ๋กœ ๋‹ค๊ฐ€์˜ค๋Š” ์œ„๊ธฐ๋ฅผ ์ ˆ๊ฐํ–ˆ๋‹ค. ํ•˜์ง€๋งŒ ๋™์‹œ์— ๊ตญ๋ฐฉ ๋ถ„์•ผ์— ๊ธฐ์—ฌํ•˜๋ ค๋Š” ์ฐฝ์—…๊ฐ€๋“ค์˜ ์—„์ฒญ๋‚œ ์—๋„ˆ์ง€์™€ ๊ตญ๊ฐ€์  ๊ด€์‹ฌ์ด ์žฌ๋ถ€์ƒํ•˜๋Š” ๊ฒƒ์„ ๋ณด๋ฉฐ, ์ง€๊ธˆ์ด์•ผ๋ง๋กœ ๋ชจ๋“  ๊ฒƒ์„ ๋ฐ”๋กœ์žก์„ ์ˆ˜ ์žˆ๋Š” โ€˜๊ธฐํšŒ์˜ ์ˆœ๊ฐ„โ€™์ด๋ผ๊ณ  ํŒ๋‹จํ–ˆ๋‹ค. ๊ทธ๋Š” ๊ณผ๊ฑฐ ์šฐ๋ฆฌ๊ฐ€ ์Šน๋ฆฌํ–ˆ๋˜ ๋ฐฉ์‹๋“ค์„ ์šฐ์—ฐํžˆ ๋“ฑํ•œ์‹œํ•˜๊ฒŒ ๋˜์—ˆ์œผ๋ฉฐ, ์ด์ œ๋Š” ๊ทธ ๋ฐฉ์‹์„ ํ™œ๊ธฐ์ฐจ๊ฒŒ ๋˜์ฐพ์„ ๊ธฐํšŒ๊ฐ€ ์™”๋‹ค๊ณ  ๊ฐ•์กฐํ•œ๋‹ค.

๋Ÿฌ์‹œ์•„์˜ 2014๋…„ ํฌ๋ฆผ๋ฐ˜๋„ ํ•ฉ๋ณ‘, 2015๋…„ ์Šคํ”„๋ž˜ํ‹€๋ฆฌ ๊ตฐ๋„(Spratly Islands)์˜ ๊ตฐ์‚ฌํ™”, ์ด๋ž€ ํ•ต ํ˜‘์ •(JCPOA)์˜ ์‹คํŒจ, ๊ทธ๋ฆฌ๊ณ  2023๋…„ 10์›” 7์ผ ํ•˜๋งˆ์Šค์˜ ์ด์Šค๋ผ์—˜ ๊ณต๊ฒฉ ์ดํ›„ ๋ฐœ์ƒํ•œ ์ฐธํ˜นํ•œ ์‚ฌ๊ฑด๋“ค์€ ์‚ฐ์นด๋ฅด์—๊ฒŒ โ€œ๋„๋Œ€์ฒด ๋ฌด์Šจ ์ผ์ด ๋ฒŒ์–ด์ง€๊ณ  ์žˆ๋Š”๊ฐ€?โ€๋ผ๋Š” ๊ธ‰์ง„์ ์ธ ์งˆ๋ฌธ์„ ๋˜์ง€๊ฒŒ ํ–ˆ๋‹ค. ์ด๋Ÿฌํ•œ ์ผ๋ จ์˜ ์‚ฌ๊ฑด๋“ค์€ ๋ฏธ๊ตญ์ด ์–ต์ง€๋ ฅ(Deterrence)์„ ์ƒ์‹คํ–ˆ์Œ์„ ๋ณด์—ฌ์ฃผ๋Š” ์ฆ๊ฑฐ์ด๋ฉฐ, ์‹œ๊ฐ„์ด ์–ผ๋งˆ ๋‚จ์ง€ ์•Š์•˜์Œ์„ ์˜๋ฏธํ–ˆ๋‹ค. ๊ทธ๋Ÿฌ๋‚˜ ๊ทธ๋Š” ์ง€๋‚œ 1๋…„ ๋™์•ˆ ๊ตญ๋ฐฉ๋ถ€ ๋‚ด๋ถ€์—์„œ ์ง€๋‚œ 19๋…„๊ฐ„ ๋ณด์ง€ ๋ชปํ–ˆ๋˜ ๋ณ€ํ™”๊ฐ€ ์ผ์–ด๋‚ฌ์œผ๋ฉฐ, ์‚ฌ๋žŒ๋“ค์ด ์ด๋Ÿฌํ•œ ๊ฐœํ˜์˜ ์ˆœ๊ฐ„์„ ํฌ์ฐฉํ•˜๊ณ  ์žˆ๋‹ค๊ณ  ๊ธ์ •์ ์œผ๋กœ ํ‰๊ฐ€ํ–ˆ๋‹ค.

์žŠํ˜€์ง„ ์Šน๋ฆฌ์˜ ๊ณต์‹: โ€˜๊ตญ๊ฐ€ ์ „์ฒดโ€™์˜ ๋™์›

์‚ฐ์นด๋ฅด๋Š” ์˜ค๋Š˜๋‚ ์˜ ์ง€์ •ํ•™์  ์ƒํ™ฉ์„ 1930๋…„๋Œ€ ํ›„๋ฐ˜์— ๋น„์œ ํ•˜๋ฉฐ, ๋” ํฐ ์ถฉ๋Œ์„ ๋ง‰๊ธฐ ์œ„ํ•œ โ€˜๋™์›(Mobilize)โ€˜์˜ ์ค‘์š”์„ฑ์„ ์—ญ์„คํ•œ๋‹ค. ๊ทธ์˜ ์ €์„œ ใ€Ž๋ชจ๋นŒ๋ผ์ด์ฆˆ(Mobilize)ใ€๋Š” ์šฐ๋ฆฌ๊ฐ€ ๋ƒ‰์ „์—์„œ ์Šน๋ฆฌ(ํ˜น์€ ์†Œ๋ จ์ด ํŒจ๋ฐฐ)ํ•œ ํ›„ ์•ˆ์ผํ•จ์— ๋น ์ ธ ๋ฏธ๋ž˜์— ๋Œ€ํ•œ ๋งŽ์€ ์ž˜๋ชป๋œ ๋ฏฟ์Œ์„ ๊ฐ–๊ฒŒ ๋˜์—ˆ๋‹ค๊ณ  ์ง€์ ํ•œ๋‹ค. ๊ทธ๋Š” ์ œ2์ฐจ ์„ธ๊ณ„๋Œ€์ „ ๋‹น์‹œ ๋ฏธ๊ตญ์˜ ๋™์›์ด ์ง„์ฃผ๋งŒ ๊ณต๊ฒฉ ์ดํ›„ ๊ฐ‘์ž๊ธฐ ์ผ์–ด๋‚œ ๊ฒƒ์ด ์•„๋‹ˆ๋ผ๊ณ  ๊ฐ•์กฐํ•œ๋‹ค. ์˜คํžˆ๋ ค ํ”„๋žญํด๋ฆฐ ๋ฃจ์Šค๋ฒจํŠธ(FDR) ๋Œ€ํ†ต๋ น์„ ๋น„๋กฏํ•œ ์ง€๋„๋ถ€๋Š” 1930๋…„๋Œ€ ํ›„๋ฐ˜๋ถ€ํ„ฐ ๊ตญ๊ฐ€์  ์˜์ง€๊ฐ€ ์ถฉ๋ถ„ํžˆ ํ˜•์„ฑ๋˜์ง€ ์•Š์•˜์Œ์—๋„ ๋ถˆ๊ตฌํ•˜๊ณ  ๋™์›์˜ ํ•„์š”์„ฑ์„ ์ธ์‹ํ–ˆ๋‹ค. ๋žœ๋“œ๋ฆฌ์Šค(Lend-Lease) ํ”„๋กœ๊ทธ๋žจ์€ ๊ณต์žฅ ๊ฑด์„ค๊ณผ ์žฌ์ •๋น„์— 18๊ฐœ์›”์ด ๊ฑธ๋ ธ๊ณ , ๊ฒฐ๊ตญ ์ง„์ฃผ๋งŒ ๊ณต๊ฒฉ ๋‹น์‹œ ๋ฏธ๊ตญ์€ ์ด๋ฏธ ์™„์ „ ์ƒ์‚ฐ ์ฒด์ œ๋ฅผ ๊ฐ–์ถ”๊ณ  ์žˆ์—ˆ๋‹ค.

์ง„์ •ํ•œ ์ „์Ÿ์€ โ€œ๊ตญ๊ฐ€ ์ „์ฒด(the whole country)โ€œ๊ฐ€ ์น˜๋ฅด๋Š” ๊ฒƒ์ด๋ฉฐ, ๋‹จ์ˆœํžˆ ๊ตญ๋ฐฉ ์‚ฐ์—… ๊ธฐ๋ฐ˜์ด๋‚˜ ๊ตญ๋ฐฉ๋ถ€๋งŒ์˜ ๋ฌธ์ œ๊ฐ€ ์•„๋‹ˆ๋ผ๋Š” ๊ฒƒ์ด ์‚ฐ์นด๋ฅด์˜ ํ•ต์‹ฌ ์ฃผ์žฅ์ด๋‹ค. ๊ทธ๋Š” ๋ฏธ๊ตญ ์‹œ๋ฏผ ๋ชจ๋‘๊ฐ€ ๊ตญ๊ฐ€๊ฐ€ ์ œ๊ณตํ•˜๋Š” ๋ฒˆ์˜๋ฟ ์•„๋‹ˆ๋ผ, ๊ทธ ๋ฒˆ์˜์˜ ๊ธฐ๋ฐ˜์ด ๋˜๋Š” ์ž์œ ์— ํˆฌ์žํ•ด์•ผ ํ•œ๋‹ค๊ณ  ๋งํ•œ๋‹ค. ๊ทธ๋Ÿฌ๋‚˜ ์šฐ๋ฆฌ๋Š” ์ด๋Ÿฌํ•œ ์ •์‹ ์—์„œ ๋„ˆ๋ฌด๋‚˜ ๋ฉ€๋ฆฌ ๋–จ์–ด์ ธ ์™”๋‹ค. 1989๋…„์—๋Š” ์ฃผ์š” ๋ฌด๊ธฐ ์‹œ์Šคํ…œ ์ง€์ถœ์˜ 6%๋งŒ์ด ๊ตญ๋ฐฉ ์ „๋ฌธ ๊ธฐ์—…์— ํ• ๋‹น๋˜์—ˆ์ง€๋งŒ, ํ˜„์žฌ ์ด ์ˆ˜์น˜๋Š” 86%์— ๋‹ฌํ•œ๋‹ค. ์ด๋Š” ๊ณผ๊ฑฐ์—๋Š” ํฌ๋ผ์ด์Šฌ๋Ÿฌ(Chrysler)๊ฐ€ ๋ฏธ๋‹ˆํŠธ๋งจ(Minuteman) ๋ฏธ์‚ฌ์ผ๊ณผ ๋ฏธ๋‹ˆ๋ฐด์„ ๋™์‹œ์— ์ƒ์‚ฐํ–ˆ๋“ฏ, ๋ฏผ๊ฐ„ ์‚ฐ์—…์ด ๊ตญ๊ฐ€ ์•ˆ๋ณด๋ฅผ ์ž์—ฐ์Šค๋Ÿฝ๊ฒŒ ๋’ท๋ฐ›์นจํ–ˆ๋˜ ๊ฒƒ๊ณผ๋Š” ๋Œ€์กฐ์ ์ด๋‹ค. ์˜ค๋Š˜๋‚  ๋ฏผ๊ฐ„ ๋ถ€๋ฌธ์˜ ์—ฐ๊ตฌ๊ฐœ๋ฐœ(R&D) ์ง€์ถœ์€ ์ •๋ถ€์˜ ์—ญ๋Ÿ‰์„ ์••๋„ํ•˜๋ฉฐ, ์ด ๋ฏผ๊ฐ„ ๊ธฐ์ˆ ์˜ ๊ฐ€๊ฒฉ ๋Œ€๋น„ ์„ฑ๋Šฅ ๊ณก์„ ์„ ๊ตญ๋ฐฉ๋ ฅ์œผ๋กœ ์ „ํ™˜ํ•˜๋Š” ๊ฒƒ์ด ์ค‘์š”ํ•˜๋‹ค๊ณ  ์‚ฐ์นด๋ฅด๋Š” ๊ฐ•์กฐํ•œ๋‹ค.

๊ณผ๊ฑฐ ๋ฏธ๊ตญ์˜ ์‚ฐ์—… ๊ธฐ๋ฐ˜์€ ๊ธ€๋ Œ ๋งˆํ‹ด(Glenn Martin), ์žญ ๋…ธ์Šค๋Ÿฝ(Jack Northrup), ๋ฅด๋กœ์ด ๊ทธ๋ฃจ๋จผ(Leroy Grumman)๊ณผ ๊ฐ™์€ ์ฐฝ์—…๊ฐ€๋“ค์ด ์ด๋Œ์—ˆ๋‹ค. ์ด๋“ค์€ ๋‹จ๊ธฐ์ ์ธ ์žฌ๋ฌด ์„ฑ๊ณผ๋ณด๋‹ค๋Š” ์ž์‹ ๊ณผ ํšŒ์‚ฌ๋ณด๋‹ค ํ›จ์”ฌ ํฐ ๋น„์ „์„ ์ถ”๊ตฌํ–ˆ๋‹ค. ์ด๋Ÿฌํ•œ ์ฐฝ์—…๊ฐ€ ์ •์‹ ์€ ์ •๋ถ€ ๋‚ด๋ถ€์—๋„ ์กด์žฌํ–ˆ๋Š”๋ฐ, ํ•ด๊ตฐ ๋‚ด ๋ฐ˜๋Œ€์—๋„ ๋ถˆ๊ตฌํ•˜๊ณ  ํ•ต ์ถ”์ง„ ํ•ด๊ตฐ์„ ๊ฑด์„คํ•œ ํ•˜์ด๋จผ ๋ฆฌ์ฝ”๋ฒ„(Hyman Rickover) ์ œ๋…์ด ๋Œ€ํ‘œ์ ์ด๋‹ค. ์‹ฌ์ง€์–ด โ€˜์›์žํญํƒ„์˜ ์•„๋ฒ„์ง€โ€™ ์˜คํŽœํ•˜์ด๋จธ(Oppenheimer)์กฐ์ฐจ ๊ทธ์˜ ์‹คํŒจ๋ฅผ ์˜ˆ๊ฒฌํ–ˆ์ง€๋งŒ, ๋ฆฌ์ฝ”๋ฒ„๋Š” ๊ตดํ•˜์ง€ ์•Š์•˜๋‹ค. ์‚ฐ์นด๋ฅด๋Š” ์ด๋Ÿฌํ•œ โ€˜ํ—ˆ์ธ ํŒŒ(chutzpah, ๋Œ€๋‹ดํ•จ)โ€™ ์ •์‹ ์ด ๋ฐ”๋กœ ์‹ค๋ฆฌ์ฝ˜๋ฐธ๋ฆฌ ์ฐฝ์—…๊ฐ€๋“ค์˜ ํŠน์ง•๊ณผ ์ผ๋งฅ์ƒํ†ตํ•œ๋‹ค๊ณ  ๋งํ•œ๋‹ค.

ํ˜์‹ ์˜ ๊ฑธ๋ฆผ๋Œ: ํ†ตํ•ฉ๊ณผ ์ˆœ์‘์˜ ๋”œ๋ ˆ๋งˆ

๋ƒ‰์ „ ์ข…์‹ ํ›„, ๋ฏธ๊ตญ์€ โ€˜ํ‰ํ™” ๋ฐฐ๋‹น๊ธˆ(peace dividend)โ€˜์„ ๊ธฐ๋Œ€ํ•˜๋ฉฐ ๊ตญ๋ฐฉ๋น„ ์ง€์ถœ์„ ์ค„์˜€๋‹ค. ์ด ์‹œ๊ธฐ์— 51๊ฐœ์— ๋‹ฌํ•˜๋˜ ์ฃผ์š” ๊ตญ๋ฐฉ ๊ณ„์•ฝ์—…์ฒด๊ฐ€ ๋‹จ 5๊ฐœ๋กœ ์ค„์–ด๋“œ๋Š” ์œ ๋ช…ํ•œ โ€˜์ตœํ›„์˜ ๋งŒ์ฐฌ(Last Supper)โ€˜์ด ์žˆ์—ˆ๋‹ค. ์‚ฐ์นด๋ฅด๋Š” ์ด ์‚ฌ๊ฑด์— ๋Œ€ํ•œ ํ†ต์ƒ์ ์ธ ์„ค๋ช…, ์ฆ‰ ํ†ตํ•ฉ์œผ๋กœ ์ธํ•œ ๊ฒฝ์Ÿ ์ƒ์‹ค์ด ๋ฌธ์ œ์˜€๋‹ค๋Š” ์ฃผ์žฅ์ด ํ‹€๋ ธ๋‹ค๊ณ  ์ง€์ ํ•œ๋‹ค. ๊ตญ๋ฐฉ ์‚ฐ์—…์€ ๋ณธ์งˆ์ ์œผ๋กœ ์ˆ˜์š” ๋…์ (Monopsony) ์‹œ์žฅ์ด๋ฉฐ, ํ˜์‹ ์€ ๊ธฐ์—… ๊ฐ„ ๊ฒฝ์Ÿ์ด ์•„๋‹ˆ๋ผ ์ •๋ถ€ ๋‚ด๋ถ€์˜ ์„œ๋น„์Šค ๊ฐ„ ๊ฒฝ์Ÿ์—์„œ ์ด‰๋ฐœ๋˜์—ˆ๋‹ค๋Š” ๊ฒƒ์ด๋‹ค.

์‚ฐ์นด๋ฅด๊ฐ€ ๋ณด๊ธฐ์— โ€˜์ตœํ›„์˜ ๋งŒ์ฐฌโ€™์œผ๋กœ ์ธํ•œ ํ†ตํ•ฉ์€ โ€˜์ˆœ์‘(conformity)โ€˜์„ ๋‚ณ์•˜๋‹ค. ์ด๋Š” ๊ตญ๋ฐฉ ์‚ฐ์—…์˜ ์ง„์ •ํ•œ โ€˜๊ธˆ์œตํ™”(financialization)โ€˜์˜ ์‹œ์ž‘์ด์—ˆ๊ณ , ๊ธฐ์—…๋“ค์€ ์„ฑ์žฅ๋ณด๋‹ค๋Š” ์žฌ๋ฌด ์ง€ํ‘œ, ๋ฐฐ๋‹น, ์ž์‚ฌ์ฃผ ๋งค์ž…, ํ˜„๊ธˆ ํ๋ฆ„์— ์ง‘์ค‘ํ•˜๊ฒŒ ๋˜์—ˆ๋‹ค. ์ด๋Ÿฌํ•œ ์ˆœ์‘์ ์ธ ํ™˜๊ฒฝ์—์„œ๋Š” ์ฐฝ์—…๊ฐ€ ์ •์‹ ์ด ๋ฒˆ์„ฑํ•  ์ˆ˜ ์—†์—ˆ๋‹ค. โ€˜์ด๋‹จ์•„(heretics)โ€˜๋“ค์€ ์ถ•์ถœ๋˜๊ฑฐ๋‚˜, ๊ธฐ์ˆ  ์‚ฐ์—…๊ณผ ๊ฐ™์€ ๋‹ค๋ฅธ ๊ฒฝ์ œ ๋ถ€๋ฌธ์œผ๋กœ ๋– ๋‚ฌ๋‹ค.

๊ทธ๋Ÿฌ๋‚˜ ์‚ฐ์นด๋ฅด๋Š” ์ด๋Ÿฌํ•œ โ€˜์ด๋‹จ์•„โ€™๋“ค์ด์•ผ๋ง๋กœ ํ˜์‹ ์— ํ•„์ˆ˜์ ์ด๋ผ๊ณ  ๊ฐ•์กฐํ•œ๋‹ค. ๊ทธ์˜ ์ €์„œ์— ๊ธฐ๋ก๋œ ๋†€๋ผ์šด ๊ตญ๋ฐฉ ํ˜์‹  ์‚ฌ๋ก€๋“ค์€ ๊ฑฐ์˜ ์˜ˆ์™ธ ์—†์ด โ€˜์ด๋‹จ์ ์ธ ์•„์ด๋””์–ดโ€™์—์„œ ๋น„๋กฏ๋˜์—ˆ๋‹ค. ์ œ๋„, ๊ด€๋ฃŒ์ฃผ์˜, ๊ธฐ์กด ์ ˆ์ฐจ๋Š” ์ด๋“ค์„ ์–ต์••ํ•˜๋ ค ํ–ˆ์ง€๋งŒ, ๊ฒฐ๊ตญ ์„ฑ๊ณตํ–ˆ๋‹ค. ์ œ2์ฐจ ์„ธ๊ณ„๋Œ€์ „์„ ์Šน๋ฆฌ๋กœ ์ด๋ˆ โ€˜ํžˆ๊ธด์Šค ๋ณดํŠธ(Higgins boat)โ€˜๊ฐ€ ๋Œ€ํ‘œ์ ์ด๋‹ค. ํ•ด๊ตฐ์€ ์ด ๋ณดํŠธ ๊ตฌ๋งค๋ฅผ ๊ฑฐ๋ถ€ํ•˜๊ณ  ์„ค๊ณ„๋„๋ฅผ ํ›”์น˜๋ ค ํ–ˆ์ง€๋งŒ, ๊ฒฐ๊ตญ ์ „์Ÿ ์ค‘ ๋ชจ๋“  ์ƒ๋ฅ™์ •์˜ 92%๊ฐ€ ํžˆ๊ธด์Šค ๋ณดํŠธ์˜€๋‹ค. ๋งŒ์•ฝ ์ฐฝ์—…๊ฐ€ ํžˆ๊ธด์Šค๊ฐ€ ๋ณ‘์ ์œผ๋กœ ์ง‘์ฐฉํ•˜์ง€ ์•Š์•˜๋‹ค๋ฉด ๋…ธ๋ฅด๋ง๋”” ์ƒ๋ฅ™์ž‘์ „์€ ๋ถˆ๊ฐ€๋Šฅํ–ˆ์„ ๊ฒƒ์ด๋‹ค.

์‚ฐ์นด๋ฅด๋Š” ์ด๋Ÿฌํ•œ ๋ฌธ์ œ์— ๋”ํ•ด, ์ˆ˜์š” ๋…์ ์˜ ํŠน์„ฑ์ƒ ๊ตญ๋ฐฉ๋ถ€๊ฐ€ ๊ณต๊ธ‰์—…์ฒด์— ์˜จ๊ฐ– ์ œ์•ฝ์„ ๊ฐ€ํ•˜๋ฉฐ โ€˜๊ฐˆ๋ผํŒŒ๊ณ ์Šค ์„ฌ(Galapagos Islands)โ€˜์— ๊ณ ๋ฆฝ๋œ ์ƒ๋ฌผ์ฒ˜๋Ÿผ ๋งŒ๋“ค์—ˆ๋‹ค๊ณ  ๋น„์œ ํ•œ๋‹ค. ๊ทธ ๊ฒฐ๊ณผ, ๋›ฐ์–ด๋‚œ โ€˜๊ฑฐ๋Œ€ ๊ฑฐ๋ถ์ดโ€™ ๊ฐ™์€ ๊ธฐ์—…๋“ค์ด ํƒ„์ƒํ–ˆ์ง€๋งŒ, ์ด๋“ค์€ ๋ณธํ† ์—์„œ๋Š” ๊ฒฝ์Ÿ๋ ฅ์„ ์žƒ๊ณ  โ€˜๋Š‘๋Œ€โ€™๋“ค์—๊ฒŒ ์žก์•„๋จนํž ๊ฒƒ์ด๋ผ๋Š” ๋น„ํŒ์ด๋‹ค. ํŒ”๋ž€ํ‹ฐ์–ด๊ฐ€ ์„ค๋ฆฝ๋  ๋‹น์‹œ๋งŒ ํ•ด๋„ ๊ตญ๋ฐฉ๋ถ€์—๋Š” ์™ธ๋ถ€์ธ์ด ์ ‘๊ทผํ•  โ€˜์ •๋ฌธ(front door)โ€˜์ด ์‚ฌ์‹ค์ƒ ์—†์—ˆ๊ณ , ์ •๋ณด ์ปค๋ฎค๋‹ˆํ‹ฐ์˜ ์ธํํ…”(In-Q-Tel)๋งŒ์ด ์œ ์ผํ•œ ํ†ต๋กœ์˜€๋‹ค. ๊ทธ๋Ÿฌ๋‚˜ 2015๋…„๋ถ€ํ„ฐ ์‹œ์ž‘๋œ ๋ณ€ํ™”์˜ ๋ฌผ๊ฒฐ๋กœ ์ธํ•ด, ์ง€๋‚œ 10๋…„๊ฐ„ ๊ตญ๋ฐฉ๋ถ€๋ฅผ ๋•๊ธฐ ์œ„ํ•ด ๋ฌธ์„ ๋‘๋“œ๋ฆฌ๋˜ โ€˜์ด๋‹จ์•„โ€™๋“ค์ด ๋น„๋กœ์†Œ ๊ธฐํšŒ๋ฅผ ์–ป๊ฒŒ ๋˜์—ˆ๋‹ค๊ณ  ์‚ฐ์นด๋ฅด๋Š” ๋งํ•œ๋‹ค.

์ด๋‹จ์  ์˜์›…๋“ค์˜ ๊ท€ํ™˜: ๊ตญ๋ฐฉ์„ ์œ„ํ•œ ๋ฏผ๊ฐ„ ์ธ์žฌ ๋™์›

์‚ฐ์นด๋ฅด๋Š” ๊ตญ๋ฐฉ ์‚ฐ์—…์˜ โ€˜์ด๋‹จ์  ์˜์›…๋“คโ€™ ์ค‘ ํ•œ ๋ช…์œผ๋กœ ๋“œ๋ฃจ ์ฟ ์ปค(Drew Cukor) ๋Œ€๋ น์„ ๊ผฝ๋Š”๋‹ค. ๊ทธ๋Š” ํ•ด๋ณ‘๋Œ€ ๋Œ€๋ น์œผ๋กœ, ๋‹จ ํ•œ ๋ฒˆ์˜ ์ž‘์ „ ์‹คํŒจ(RPG ์˜ค์ธ์œผ๋กœ ์ธํ•œ ์•ผ์ง€๋””์กฑ ๋‚œ๋ฏผ ๊ตฌ์กฐ ์‹คํŒจ)๊ฐ€ ์ˆ˜์ฒœ ๋ช…์˜ ๋ฏผ๊ฐ„์ธ์—๊ฒŒ ์ฐธํ˜นํ•œ ๊ฒฐ๊ณผ๋ฅผ ๊ฐ€์ ธ์˜จ ๊ฒƒ์„ ๋ชฉ๊ฒฉํ•œ ํ›„, AI๋ฅผ ๊ตญ๋ฐฉ๋ถ€์— ๋„์ž…ํ•˜๋Š” ๋ฉ”์ด๋ธ ํ”„๋กœ์ ํŠธ(Project Maven)์— ์ „๋…ํ–ˆ๋‹ค. ์ž์›๋„ ์—†๋Š” ์ƒํ™ฉ์—์„œ ์‹œ์ž‘๋œ ์ด ํ”„๋กœ์ ํŠธ์—์„œ ์ฟ ์ปค ๋Œ€๋ น์€ ๋ชจ๋“  ์ด๋“ค์˜ ๋ฐ˜๋Œ€์™€ ์งˆํˆฌ, ์‹ฌ์ง€์–ด ํ—ˆ์œ„ ๊ณ ๋ฐœ(์ด๋ž€์ธ์„ ์ง‘์— ์ˆจ๊ธฐ๊ณ  ์žˆ๋‹ค๋Š”)๊นŒ์ง€ ๊ฐ์ˆ˜ํ•ด์•ผ ํ–ˆ๋‹ค. ํ•˜์ง€๋งŒ ๊ทธ๋Š” ํ”๋“ค๋ฆฌ์ง€ ์•Š์•˜๊ณ , ๊ฒฐ๊ตญ AI์˜ ๊ตญ๋ฐฉ ๋„์ž…์— ๊ฒฐ์ •์ ์ธ ์—ญํ• ์„ ํ–ˆ๋‹ค. ์‚ฐ์นด๋ฅด๋Š” ์ด๋Ÿฌํ•œ ์ด์•ผ๊ธฐ๊ฐ€ โ€œ๋‚˜๋ผ๊ฐ€ ๋‹น์‹ ์„ ์ •๋ง ํ•„์š”๋กœ ํ•  ๋•Œ, ์ˆจ์–ด์žˆ๋Š” ์ด๋‹จ์•„๋“ค์ด ๋‚˜์„œ๋„๋ก ์˜๊ฐ์„ ์ฃผ๋Š” ๊ฒƒโ€์ด๋ผ๊ณ  ๊ฐ•์กฐํ•œ๋‹ค.

์กด ๋ณด์ด๋“œ(John Boyd) ๋Œ€๋ น์˜ F-16 ๊ฐœ๋ฐœ๊ณผ OODA ๋ฃจํ”„(OODA Loop) ์ด๋ก , ๋ฒ„๋‚˜๋“œ ์Šˆ๋ฆฌ๋ฒ„(Bernard Schriever) ์žฅ๊ตฐ๊ณผ ์—๋“œ์›Œ๋“œ ํ™€(Edward Hall) ๋ฐ•์‚ฌ์˜ ๋ฏธ๋‹ˆํŠธ๋งจ ๋ฏธ์‚ฌ์ผ ๊ฐœ๋ฐœ ์‚ฌ๋ก€ ๋˜ํ•œ โ€˜์ด๋‹จ์•„โ€™๋“ค์„ ๋ณดํ˜ธํ•˜๊ณ  ์ง€์›ํ•˜๋Š” ๋ฆฌ๋”์‹ญ์˜ ์ค‘์š”์„ฑ์„ ๋ณด์—ฌ์ค€๋‹ค. ์ด๋“ค์€ ๋ชจ๋‘ ์กฐ์ง ๋‚ด์—์„œ ์—„์ฒญ๋‚œ ์ €ํ•ญ์— ๋ถ€๋”ชํ˜”์ง€๋งŒ, ๊ทธ๋“ค์˜ ํƒ์›”ํ•จ์„ ์•Œ์•„๋ณธ ๋ฆฌ๋”๋“ค์˜ ๋ณดํ˜ธ ์•„๋ž˜ ํ˜์‹ ์„ ์ด๋ค„๋‚ผ ์ˆ˜ ์žˆ์—ˆ๋‹ค.

์ด๋Ÿฌํ•œ ๋งฅ๋ฝ์—์„œ ์ƒด ์‚ฐ์นด๋ฅด ์ž์‹ ๋„ ์ตœ๊ทผ ์œก๊ตฐ์— ์ž…๋Œ€ํ–ˆ๋‹ค. ๊ทธ๋Š” 2014๋…„๋ถ€ํ„ฐ ์ด์Šค๋ผ์—˜๊ณผ ํ˜‘๋ ฅํ•˜๋ฉฐ, 10์›” 7์ผ ์ดํ›„ ์ด์Šค๋ผ์—˜์ด 36๋งŒ ๋ช…์˜ ์˜ˆ๋น„๊ตฐ์„ ๋™์›ํ–ˆ์„ ๋•Œ ๋†€๋ผ์šด ๋ณ€ํ™”๋ฅผ ๋ชฉ๊ฒฉํ–ˆ๋‹ค. ์ด์Šค๋ผ์—˜ ์˜ˆ๋น„๊ตฐ๋“ค์€ 20๋…„ ์ด์ƒ ๋ฏผ๊ฐ„ ์‚ฐ์—…์—์„œ ์ธํ„ฐ๋„ท ๊ทœ๋ชจ์˜ ์‹œ์Šคํ…œ์„ ๊ตฌ์ถ•ํ•˜๋ฉฐ ์Œ“์€ ๊ฒฝํ—˜์„ ๋ฐ”ํƒ•์œผ๋กœ, ๋ถˆ๊ณผ 4๊ฐœ์›” ๋งŒ์— ๊ตฐ๋Œ€์˜ ๊ธฐ์ˆ ์„ 10๋…„ ์ด์ƒ ๋ฐœ์ „์‹œ์ผฐ๋‹ค. ์‚ฐ์นด๋ฅด๋Š” ๋ฏธ๊ตญ ๋˜ํ•œ ์ด์Šค๋ผ์—˜์ฒ˜๋Ÿผ ์—„์ฒญ๋‚œ ๊ธฐ์ˆ  ์ธ์žฌ๋ฅผ ๋ณด์œ ํ•˜๊ณ  ์žˆ์œผ๋ฉฐ, 24์„ธ์˜ ์ Š์€ ์‹œ์ ˆ์—๋Š” ํ•  ์ˆ˜ ์—†์—ˆ๋˜ ๊ธฐ์—ฌ๋ฅผ 44์„ธ์˜ ๊ฒฝํ—˜์œผ๋กœ ํ•  ์ˆ˜ ์žˆ๋‹ค๊ณ  ๋งํ•œ๋‹ค. ์ œ2์ฐจ ์„ธ๊ณ„๋Œ€์ „ ๋‹น์‹œ 10๋งŒ ๋ช…์˜ ๋ฏผ๊ฐ„์ธ์ด ๊ตฐ์œผ๋กœ ํŠน๋ณ„ ์ž„๊ด€(direct commissioned)๋˜์—ˆ๋˜ ๊ฒƒ์ฒ˜๋Ÿผ, ์˜ค๋Š˜๋‚ ์—๋„ ๋ฏผ๊ฐ„ ๊ธฐ์ˆ  ์ธ๋ ฅ์„ ๊ตญ๋ฐฉ์— ํ™œ์šฉํ•ด์•ผ ํ•œ๋‹ค๋Š” ๊ฒƒ์ด๋‹ค. ๊ทธ๋Š” ์ „ ์˜คํ”ˆAI(OpenAI) ์ˆ˜์„ ์—ฐ๊ตฌ์› ๋ฐฅ ๋งฅ๊ทธ๋ฃจ(Bob McGrew), ๋ฉ”ํƒ€(Meta)์˜ ์ตœ๊ณ ๊ธฐ์ˆ ์ฑ…์ž„์ž(CTO) ๋ณด์Šค(Bos), ์ „ ์˜คํ”ˆAI ์ตœ๊ณ ์ œํ’ˆ์ฑ…์ž„์ž ์•ค๋“œ๋ฅ˜ ์œŒ(Andrew Wheel) ๋“ฑ ์„ธ ๋ช…์˜ ๋™๋ฃŒ์™€ ํ•จ๊ป˜ ์œก๊ตฐ์— ํ•ฉ๋ฅ˜ํ•˜์—ฌ ๊ณ ์œ„ ์ž๋ฌธ๊ด€์œผ๋กœ์„œ ๋‹ค์–‘ํ•œ ํ”„๋กœ์ ํŠธ๋ฅผ ๋•๊ณ  ์žˆ๋‹ค.

์‚ฐ์นด๋ฅด๋Š” ์œก๊ตฐ ๋‚ด๋ถ€์—์„œ ๊ฐ€์žฅ ๋†€๋ž๋˜ ์ ์œผ๋กœ โ€˜์‹ ์„ธ๋Œ€ ์žฅ๋ณ‘๋“ค(green suitors)โ€˜์˜ ํƒ์›”ํ•œ ์—ญ๋Ÿ‰์„ ๊ผฝ๋Š”๋‹ค. ์ด๋“ค์€ ์ •์‹ ์ปดํ“จํ„ฐ ๊ณผํ•™ ๊ต์œก์„ ๋ฐ›์ง€ ์•Š์•˜์Œ์—๋„ ๋ถˆ๊ตฌํ•˜๊ณ , ์Šค์Šค๋กœ AI ์• ํ”Œ๋ฆฌ์ผ€์ด์…˜์„ ๊ตฌ์ถ•ํ•˜๋Š” ๋†€๋ผ์šด ๋Šฅ๋ ฅ์„ ๋ณด์—ฌ์ค€๋‹ค. ํŠนํžˆ โ€˜์ž‘์ „ ๋ฐ์ดํ„ฐ ํŒ€(Operational Data Teams)โ€˜์—์„œ ์ด๋“ค์ด ๊ฐœ๋ฐœํ•˜๋Š” AI ์†”๋ฃจ์…˜์€ ์ƒ์—… ๋ถ€๋ฌธ์—์„œ ๋ณผ ์ˆ˜ ์žˆ๋Š” ์–ด๋–ค ๊ฒƒ๋ณด๋‹ค๋„ ์„ค๋“๋ ฅ์ด ์žˆ๋‹ค๊ณ  ํ•œ๋‹ค. ์ด๋Š” ์ƒ์กด์ด ๊ฑธ๋ฆฐ โ€˜์‹ค์กด์  ์œ„ํ˜‘โ€™ ์•ž์—์„œ ์‚ฌ๋žŒ๋“ค์€ 10%์˜ ํšจ์œจ์„ฑ ๊ฐœ์„ ์ด ์•„๋‹Œ โ€˜์Šน๋ฆฌโ€™๋ผ๋Š” ์ด๋ถ„๋ฒ•์  ๋ชฉํ‘œ์— ์ง‘์ค‘ํ•˜๊ธฐ ๋•Œ๋ฌธ์ด๋‹ค. ๋˜ํ•œ AI๋Š” ํŠน์ • ๊ธฐ์ˆ ์„ ๊ฐ€์ง„ ๊ฐœ์ธ, ํŠนํžˆ ์ž์‹ ์˜ ์ „๋ฌธ ์˜์—ญ์„ ๊นŠ์ด ์ดํ•ดํ•˜๋Š” ์ •๋ณด ์žฅ๊ต๋“ค์—๊ฒŒ ์—„์ฒญ๋‚œ ํž˜์„ ์‹ค์–ด์ค€๋‹ค. ๊ณผ๊ฑฐ์—๋Š” ํŒŒ์›Œํฌ์ธํŠธ ์Šฌ๋ผ์ด๋“œ๋ฅผ ๋งŒ๋“ค๋˜ ์ด๋“ค์ด ์ด์ œ๋Š” 2์ฃผ ๋งŒ์— ์ง์ ‘ ์†”๋ฃจ์…˜์„ ๊ตฌ์ถ•ํ•˜๊ณ , ๊ทธ ์„ฑ๊ณผ๋ฅผ ํ†ตํ•ด ๊ตฐ์„ ๋ฐœ์ „์‹œํ‚ค๋Š” ๊ฒฝํ—˜์  ๋Œ€ํ™”๋ฅผ ๋‚˜๋ˆ„๊ณ  ์žˆ๋‹ค. ์ด๋Ÿฌํ•œ ๋ฐ”ํ…€์—…(bottom-up) ํ˜์‹ ๊ณผ ์ž„๋ฌดํ˜• ์ง€ํœ˜(Mission Command)๋Š” ๋ฏธ๊ตญ ๊ตฐ๋Œ€์˜ ๊ณ ์œ ํ•œ ๊ฐ•์ ์ด๋ผ๊ณ  ์‚ฐ์นด๋ฅด๋Š” ๊ฐ•์กฐํ•œ๋‹ค.

AI ์‹œ๋Œ€์˜ ์†Œํ”„ํŠธ์›จ์–ด ์ง€ํ˜• ๋ณ€ํ™”: โ€˜๋ฒ ํƒ€โ€™์™€ โ€˜์•ŒํŒŒโ€™์˜ ๊ฐˆ๋ฆผ๊ธธ

์ตœ๊ทผ ์†Œํ”„ํŠธ์›จ์–ด ์—…๊ณ„์—์„œ๋Š” AI์˜ ๋“ฑ์žฅ์œผ๋กœ ์ธํ•œ โ€˜SaaS ์ข…๋ง๋ก (SaaS Apocalypse)โ€˜์ด ๋Œ€๋‘๋˜๊ณ  ์žˆ๋‹ค. AI๊ฐ€ ์ฝ”๋“œ๋ฅผ ์ƒ์„ฑํ•˜๊ณ  ํ†ตํ•ฉํ•˜๋Š” ๋Šฅ๋ ฅ์ด ํ–ฅ์ƒ๋˜๋ฉด์„œ, ๊ธฐ์กด SaaS(Software as a Service) ๊ธฐ์—…๋“ค์˜ ์ง„์ž… ์žฅ๋ฒฝ(switching costs)์ด ๋‚ฎ์•„์ง€๊ณ  ์žˆ๋‹ค๋Š” ์ฃผ์žฅ์ด๋‹ค. ์ด์— ๋Œ€ํ•ด ์‚ฐ์นด๋ฅด๋Š” ์–‘์ชฝ ์ฃผ์žฅ์ด ๋ชจ๋‘ ์‚ฌ์‹ค์ด๋ผ๊ณ  ๋งํ•˜๋ฉฐ, ์†Œํ”„ํŠธ์›จ์–ด๋ฅผ โ€˜๋ฒ ํƒ€(Beta)โ€˜์™€ โ€˜์•ŒํŒŒ(Alpha)โ€˜๋กœ ๊ตฌ๋ถ„ํ•˜๋Š” ์ƒˆ๋กœ์šด ๊ด€์ ์„ ์ œ์‹œํ•œ๋‹ค.

โ€˜๋ฒ ํƒ€ ์†Œํ”„ํŠธ์›จ์–ดโ€™๋Š” ์‚ฌ์šฉ์ž๋ฅผ ๋‹ค๋ฅธ ๋ชจ๋“  ์‚ฌ๋žŒ๊ณผ โ€˜์œ ์‚ฌํ•˜๊ฒŒโ€™ ๋งŒ๋“œ๋Š” ๋ฐ ์ดˆ์ ์„ ๋งž์ถ˜๋‹ค. ์ด๋Š” ๊ณผ๊ฑฐ ์†Œํ”„ํŠธ์›จ์–ด ์‚ฐ์—…์˜ ๋ฌธ์ œ์ ์ด๊ธฐ๋„ ํ–ˆ๋Š”๋ฐ, ์†Œํ”„ํŠธ์›จ์–ด ๊ฐœ๋ฐœ์˜ ํ”ผ๋“œ๋ฐฑ ๋ฃจํ”„๊ฐ€ โ€˜๊ฐ€์น˜ ์ฐฝ์ถœโ€™๋ณด๋‹ค๋Š” โ€˜ํŒ๋งค ๊ฐ€๋Šฅ์„ฑโ€™์— ์ง‘์ค‘๋˜์–ด ์žˆ์—ˆ๊ธฐ ๋•Œ๋ฌธ์ด๋‹ค. ์‚ฐ์นด๋ฅด๋Š” ์ด๋Ÿฌํ•œ ๋ฒ ํƒ€ ์†Œํ”„ํŠธ์›จ์–ด๋Š” AI ์‹œ๋Œ€์— ํฐ ์–ด๋ ค์›€์„ ๊ฒช์„ ๊ฒƒ์ด๋ผ๊ณ  ์˜ˆ์ธกํ•œ๋‹ค. ๋ฐ˜๋ฉด, AI๋Š” ์‚ฌ์šฉ์ž์˜ ๊ณ ์œ ํ•œ ์š”๊ตฌ์— ๋งž์ถฐ ์†Œํ”„ํŠธ์›จ์–ด๋ฅผ ๋งŒ๋“ค ์ˆ˜ ์žˆ๊ฒŒ ํ•จ์œผ๋กœ์จ ๋ณธ์งˆ์ ์œผ๋กœ โ€˜์•ŒํŒŒ(Alpha)โ€˜์— ์ง‘์ค‘ํ•  ์ˆ˜ ์žˆ๊ฒŒ ํ•œ๋‹ค. ์•ŒํŒŒ ์†Œํ”„ํŠธ์›จ์–ด๋Š” ์‚ฌ์šฉ์ž๋ฅผ โ€˜๋‹ค๋ฅด๊ฒŒโ€™ ๋งŒ๋“ค๊ณ , ๊ฒฝ์Ÿ ์šฐ์œ„์™€ ์ „๋žต์„ ํ‘œํ˜„ํ•  ์ˆ˜ ์žˆ๊ฒŒ ํ•ด์ค€๋‹ค. ๊ทธ๋Š” ์ด๋ฏธ ์•ŒํŒŒ์— ์ง‘์ค‘ํ•˜๊ณ  ์žˆ๋˜ ํ”Œ๋žซํผ๋“ค์€ AI ์‹œ๋Œ€์— ๋”์šฑ ๊ฐ•๋ ฅํ•œ ๋ฐ”๋žŒ์„ ๋งž์„ ๊ฒƒ์ด๋ผ๊ณ  ๋งํ•œ๋‹ค.

์ฝ”๋กœ๋‚˜19 ํŒฌ๋ฐ๋ฏน ๋‹น์‹œ, ๋งŽ์€ ๊ธฐ์—…์˜ ์ „์‚ฌ์  ์ž์› ๊ด€๋ฆฌ(ERP) ์‹œ์Šคํ…œ์ด ๋ฌด๋„ˆ์กŒ์ง€๋งŒ, ์คŒ(Zoom)์ด๋‚˜ ํŒ€์ฆˆ(Teams) ๊ฐ™์€ ํ˜‘์—… ๋„๊ตฌ๋“ค์€ ์›๊ฒฉ ๊ทผ๋ฌด๋ฅผ ๊ฐ€๋Šฅํ•˜๊ฒŒ ํ–ˆ๋‹ค. ์‚ฐ์นด๋ฅด๋Š” ์ด๋ฅผ ์†Œํ”„ํŠธ์›จ์–ด ์‚ฐ์—…์˜ โ€˜์Šคํ‘ธํŠธ๋‹ˆํฌ ์‡ผํฌ(Sputnik moment)โ€˜๋กœ ๋น„์œ ํ•˜๋ฉฐ, ์ง€๋‚œ ์ˆ˜์‹ญ ๋…„๊ฐ„ ์šฐ๋ฆฌ๊ฐ€ ๊ฐ€์น˜ ์žˆ๋‹ค๊ณ  ๋ฏฟ์—ˆ๋˜ ์ˆ˜๋งŽ์€ ์†Œํ”„ํŠธ์›จ์–ด ์ง€์ถœ์ด ์‚ฌ์‹ค์€ ๊ทธ๋ ‡์ง€ ์•Š์•˜์Œ์„ ๊นจ๋‹ฌ์•„์•ผ ํ•œ๋‹ค๊ณ  ์ง€์ ํ•œ๋‹ค. ํŒ”๋ž€ํ‹ฐ์–ด๋Š” ๊ณ ๊ฐ๋“ค์ด ํ˜ผ๋ž€์˜ ์†๋„์— ๋งž์ถฐ ํ˜„์‹ค์— ์ ์‘ํ•˜๋„๋ก ๋„์šฐ๋ฉฐ ์˜คํžˆ๋ ค ํฐ ์„ฑ์žฅ์„ ๊ฒฝํ—˜ํ–ˆ๋‹ค. ๊ทธ๋Š” ๋งŽ์€ ๊ธฐ์—…๋“ค์ด ๋ชจ๋ฐฉ์ ์œผ๋กœ ์ง€์ถœํ•˜๋˜ โ€˜๋ฒ ํƒ€โ€™ ์†Œํ”„ํŠธ์›จ์–ด๋“ค์ด ํฐ ์••๋ฐ•์„ ๋ฐ›์„ ๊ฒƒ์ด๋ฉฐ, ๋ฐ˜๋Œ€๋กœ ๊ธฐ์—…์˜ ๊ฒฝ์Ÿ ์šฐ์œ„๋ฅผ ํ‘œํ˜„ํ•˜๋Š” โ€˜์•ŒํŒŒโ€™ ์†Œํ”„ํŠธ์›จ์–ด, ์ฆ‰ ํˆดํ‚ท(toolkit)๊ณผ ๊ฐ™์€ ์ ‘๊ทผ ๋ฐฉ์‹์ด ๊ฐ๊ด‘๋ฐ›์„ ๊ฒƒ์ด๋ผ๊ณ  ๋‚ด๋‹ค๋ณธ๋‹ค.

AI ์‹œ๋Œ€์˜ ๊ฐ€์น˜ ์ฐฝ์ถœ์— ๋Œ€ํ•ด์„œ๋Š”, ํ˜„์žฌ ํ•˜๋“œ์›จ์–ด ๊ณ„์ธต์ด ๊ฐ€์žฅ ๋†’์€ ๋งˆ์ง„์„ ์ฐจ์ง€ํ•˜๊ณ  ์žˆ์ง€๋งŒ, ์ธํ„ฐ๋„ท ๊ฒฝ์ œ์—์„œ๋Š” ์• ํ”Œ๋ฆฌ์ผ€์ด์…˜ ๊ณ„์ธต์ด ๊ทธ๋žฌ๋‹ค. ์‚ฐ์นด๋ฅด๋Š” AI ์‹œ๋Œ€์—๋Š” ์นฉ(chips) ๊ณ„์ธต๊ณผ AI ์ธํ”„๋ผ(AI infrastructure) ๊ณ„์ธต(๊ทธ๋Š” ์ด๋ฅผ โ€˜์˜จํ†จ๋กœ์ง€(Ontology)โ€˜๋ผ๊ณ  ๋ถ€๋ฅธ๋‹ค)์—์„œ ๊ฐ€์น˜๊ฐ€ ์ฐฝ์ถœ๋  ๊ฒƒ์ด๋ฉฐ, ์ด ๋‘ ๊ณ„์ธต์ด ์ƒ๋‹นํ•œ ๋ฐฉ์–ด๋ ฅ์„ ๊ฐ€์งˆ ๊ฒƒ์ด๋ผ๊ณ  ์˜ˆ์ธกํ•œ๋‹ค. ๋ชจ๋ธ(models) ์ž์ฒด๋Š” ์ ์ฐจ ์ƒํ’ˆํ™”๋  ์••๋ ฅ์„ ๋ฐ›์œผ๋ฏ€๋กœ, ๋ชจ๋ธ ๊ธฐ์—…๋“ค์€ AI ์ธํ”„๋ผ๋ฅผ ๊ตฌ์ถ•ํ•˜๋Š” ๋ฐฉํ–ฅ์œผ๋กœ ํ™•์žฅํ•˜๊ณ , ํŠน์ • ์ˆ˜์ง AI ์†”๋ฃจ์…˜ ๊ธฐ์—…๋“ค์€ ๊ณ ๊ฐ ๊ธฐ๋ฐ˜์„ ํ™•์žฅํ•˜๊ธฐ ์œ„ํ•ด AI ์ธํ”„๋ผ๋กœ ๋‚ด๋ ค์˜ค๊ฒŒ ๋  ๊ฒƒ์ด๋ผ๋Š” ๋ถ„์„์ด๋‹ค.

์ธ๊ฐ„์˜ ์˜์ง€, AI์˜ ๋ฏธ๋ž˜: ๊ฒฝ์ œ์™€ ์ผ์ž๋ฆฌ์— ๋Œ€ํ•œ ํ†ต์ฐฐ

AI๊ฐ€ ๊ฒฝ์ œ์™€ ์ผ์ž๋ฆฌ์— ๋ฏธ์น  ์˜ํ–ฅ์— ๋Œ€ํ•ด ์‚ฐ์นด๋ฅด๋Š” ์ค‘์š”ํ•œ ๊ด€์ ์„ ์ œ์‹œํ•œ๋‹ค. ๊ทธ๋Š” AI์— ๋Œ€ํ•ด ์ด์•ผ๊ธฐํ•  ๋•Œ ๋งˆ์น˜ ์ธ๊ฐ„์˜ ์˜์ง€(human agency)๊ฐ€ ์—†๋Š” ๊ฒƒ์ฒ˜๋Ÿผ โ€œAI๊ฐ€ X๋ฅผ ํ•  ๊ฒƒ์ด๋‹คโ€๋ผ๊ณ  ๋งํ•˜๋Š” ๋ฐฉ์‹์— ๋ถˆํŽธํ•จ์„ ๋А๋‚€๋‹ค๊ณ  ๋งํ•œ๋‹ค. โ€œ์ธ๊ฐ„์ด AI๋ฅผ ์‚ฌ์šฉํ•˜์—ฌ X๋ฅผ ํ•  ๊ฒƒ์ด๋‹คโ€๊ฐ€ ์˜ฌ๋ฐ”๋ฅธ ํ‘œํ˜„์ด๋ฉฐ, ์šฐ๋ฆฌ์—๊ฒŒ๋Š” AI๋ฅผ ํŠน์ • ๋ฐฉํ–ฅ์œผ๋กœ ์ด๋Œ ์„ ํƒ๊ถŒ๊ณผ ์˜๋ฌด๊ฐ€ ์žˆ๋‹ค๋Š” ๊ฒƒ์ด๋‹ค.

๊ทธ๋Š” AI๊ฐ€ 1970๋…„๋Œ€ ์ดํ›„ ๋ฒŒ์–ด์ง„ ์ž„๊ธˆ ์„ฑ์žฅ๊ณผ ๊ตญ๋‚ด์ด์ƒ์‚ฐ(GDP) ์„ฑ์žฅ ๊ฐ„์˜ ๊ทผ๋ณธ์ ์ธ ๋‹จ์ ˆ์„ ํ•ด๊ฒฐํ•  ์—ญ์‚ฌ์  ๊ธฐํšŒ๋ฅผ ์ œ๊ณตํ•œ๋‹ค๊ณ  ๋ณธ๋‹ค. AI๋Š” ๋ฏธ๊ตญ ๋…ธ๋™์ž๋“ค์—๊ฒŒ โ€˜์ดˆ๋Šฅ๋ ฅ(superpowers)โ€˜์„ ๋ถ€์—ฌํ•˜์—ฌ ์ƒ์‚ฐ์„ฑ์„ ๊ทน๋Œ€ํ™”ํ•˜๊ณ , ์ด๋Š” ๊ตญ๊ฐ€์˜ ์žฌ์‚ฐ์—…ํ™”(re-industrialization)๋ฅผ ๋’ท๋ฐ›์นจํ•  ๊ธฐ๋ฐ˜์ด ๋  ์ˆ˜ ์žˆ๋‹ค๋Š” ๊ฒƒ์ด๋‹ค. ํ•˜๋“œ๋ฆฌ์•ˆ(Hadrien)๊ณผ ๊ฐ™์€ ๊ธฐ์—…๋“ค์ด ๊ธฐ์ˆ ์„ ํ™œ์šฉํ•˜์—ฌ ์‚ฌ๋žŒ๋“ค์˜ ์ƒ์‚ฐ์„ฑ์„ 50~100๋ฐฐ ํ–ฅ์ƒ์‹œํ‚ค๋ฉฐ ์žฌ์‚ฐ์—…ํ™”๋ฅผ ์ด๋„๋Š” ๊ฒƒ์ด ๊ทธ ์˜ˆ์‹œ์ด๋‹ค.

๋˜ํ•œ ์‚ฐ์นด๋ฅด๋Š” ์„ธ๊ณ„ํ™”์˜ โ€˜๊ฑฐ๋Œ€ํ•œ ๊ฑฐ์ง“๋งโ€™ ์ค‘ ํ•˜๋‚˜๊ฐ€ ํ˜์‹ (innovation)์€ ํ•œ๊ณณ์—์„œ ํ•˜๊ณ  ์ƒ์‚ฐ(production)์€ ๋‹ค๋ฅธ ๊ณณ์—์„œ ํ•  ์ˆ˜ ์žˆ๋‹ค๋Š” ๋ฏฟ์Œ์ด์—ˆ๋‹ค๊ณ  ์ง€์ ํ•œ๋‹ค. ๊ทธ๋Š” ํ˜์‹ ์€ ์ƒ์‚ฐ์„ฑ์˜ ๊ฒฐ๊ณผ์ด๋ฉฐ, ๋ฌด์–ธ๊ฐ€๋ฅผ ๋งŒ๋“ค์ง€ ์•Š์œผ๋ฉด ๊ทธ๊ฒƒ์„ ๋งŒ๋“œ๋Š” ๋ฐฉ์‹์ด๋‚˜ ๋ฌด์—‡์„ ๋งŒ๋“ค์ง€์— ๋Œ€ํ•ด ํ˜์‹ ํ•  ์ˆ˜ ์—†๋‹ค๊ณ  ๊ฐ•์กฐํ•œ๋‹ค. ์ŠคํŽ˜์ด์ŠคX(SpaceX)์˜ R&D ์—”์ง€๋‹ˆ์–ด๋“ค์ด ์ƒ์‚ฐ ํ˜„์žฅ๊ณผ ํ•จ๊ป˜ ์ผํ•˜๋Š” ๊ฒƒ์ด ๋Œ€ํ‘œ์ ์ธ ์˜ˆ์‹œ์ด๋‹ค. ๊ทธ๋Š” AI๋ฅผ ๊ตญ๊ฐ€์  ๋น„์ƒ์‚ฌํƒœ์ด์ž ๋™์‹œ์— ๊ตญ๊ฐ€์  ๊ธฐํšŒ๋กœ ๋ณด์•„์•ผ ํ•œ๋‹ค๊ณ  ์ฃผ์žฅํ•œ๋‹ค.

์—ญ์‚ฌ์ ์œผ๋กœ ๊ธฐ์ˆ  ํ˜๋ช…์€ ๋Œ€๋ถ€๋ถ„ โ€˜๋„๊ตฌ ํ˜๋ช…โ€™์ด์—ˆ์ง€ โ€˜๊ฐœ๋… ํ˜๋ช…โ€™์ด ์•„๋‹ˆ์—ˆ๋‹ค. ๋ง์›๊ฒฝ์„ ๋ฐœ๋ช…ํ•œ ๊ฒƒ์€ ๊ฐˆ๋ฆด๋ ˆ์˜ค๊ฐ€ ์•„๋‹ˆ๋ผ ๋‹ค๋ฅธ ์‚ฌ๋žŒ์ด์—ˆ์ง€๋งŒ, ๊ฐˆ๋ฆด๋ ˆ์˜ค๋Š” ๊ทธ๊ฒƒ์„ ์‚ฌ์šฉํ•˜์—ฌ ํ–‰์„ฑ ์šด๋™์„ ๋ฐœ๊ฒฌํ–ˆ๋‹ค. ํ˜„๋ฏธ๊ฒฝ, ๋™๋ ฅ ์ง์กฐ๊ธฐ, ๊ฐœ์ธ์šฉ ์ปดํ“จํ„ฐ ๋“ฑ ์ด๋Ÿฌํ•œ ๊ธฐ์ˆ ์˜ ๋ฏธ๋ž˜๋Š” ๊ธฐ์ˆ ์„ ๋ฐœ๋ช…ํ•œ ์‚ฌ๋žŒ์ด ์•„๋‹ˆ๋ผ โ€˜๊ธฐ์ˆ ์„ ์‚ฌ์šฉํ•˜๋Š” ์‚ฌ๋žŒ(wielders)โ€˜์— ์˜ํ•ด ๊ฒฐ์ •๋œ๋‹ค. ์˜ค๋Š˜๋‚  AI ๋น„๊ด€๋ก (AI dumerism)์€ ์ฃผ๋กœ ๋งค์šฐ ๋˜‘๋˜‘ํ•œ โ€˜๋ฐœ๋ช…๊ฐ€โ€™๋“ค์˜ ๋ชฉ์†Œ๋ฆฌ์—์„œ ๋‚˜์˜ค์ง€๋งŒ, ๊ทธ๋“ค์˜ ์ฐฝ์กฐ๋ฌผ์ฒ˜๋Ÿผ ๊ทธ๋“ค ์—ญ์‹œ โ€˜๋ถˆ๊ท ํ˜•ํ•œ ์ง€๋Šฅ(jagged intelligence)โ€˜์„ ๊ฐ€์งˆ ์ˆ˜ ์žˆ๊ธฐ์—, ๋ชจ๋ธ ๊ตฌ์ถ•์— ๋Šฅ์ˆ™ํ•˜๋‹ค๊ณ  ํ•ด์„œ ๊ทธ ํ•จ์˜๊นŒ์ง€ ์ •ํ™•ํžˆ ์˜ˆ์ธกํ•  ์ˆ˜ ์žˆ๋Š” ๊ฒƒ์€ ์•„๋‹ˆ๋‹ค.

์‚ฐ์นด๋ฅด๋Š” ์šฐ๋ฆฌ๊ฐ€ AI์˜ ๋ฏธ๋ž˜๋ฅผ ๊ฒฐ์ •ํ•  โ€˜์‚ฌ์šฉ์žโ€™๋กœ์„œ ์ธ๊ฐ„์˜ ์˜์ง€๋ฅผ ํฌ๊ธฐํ•ด์„œ๋Š” ์•ˆ ๋œ๋‹ค๊ณ  ๊ฐ•์กฐํ•œ๋‹ค. ๊ฒฝ์ œํ•™์ž๋“ค์˜ AI ๋ฏธ๋ž˜ ์˜ˆ์ธก ๊ทธ๋ž˜ํ”„์ฒ˜๋Ÿผ, ๋‹ค์–‘ํ•œ ๊ฒฐ๊ณผ๊ฐ€ ๋ชจ๋‘ ๊ฐ€๋Šฅํ•˜๋ฉฐ, ์šฐ๋ฆฌ๊ฐ€ ์–ด๋–ค ๋ฏธ๋ž˜๋ฅผ ์„ ํƒํ• ์ง€๋Š” ์šฐ๋ฆฌ์—๊ฒŒ ๋‹ฌ๋ ค ์žˆ๋‹ค๋Š” ๊ฒƒ์ด๋‹ค.

๋งˆ์ง€๋ง‰์œผ๋กœ, ๊ทธ๋Š” ๊ณผ๊ฑฐ ๋ฏธ๊ตญ ๊ฒฝ์ œ์—์„œ ๊ณต


โ€œSilicon Valleyโ€™s Big Bets on War Pay Off, and the Trump Family Business Looks to Transylvaniaโ€ โ€” New York Times Podcasts ๊ธฐ๋ฐ˜ ๊ธฐ์‚ฌ ์›๋ณธ ์˜์ƒ ๋ณด๊ธฐ

์ „์Ÿ์ด ์‹ค๋ฆฌ์ฝ˜๋ฐธ๋ฆฌ์— ์•ˆ๊ฒจ์ค€ ๊ธฐํšŒ: ํŠธ๋Ÿผํ”„ ๊ฐ€๋ฌธ์˜ ํŠธ๋ž€์‹ค๋ฐ”๋‹ˆ์•„ ๊ฐœ๋ฐœ ๋ฏธ์Šคํ„ฐ๋ฆฌ

๊ฒฉ๋™ํ•˜๋Š” ๊ตญ์ œ ์ •์„ธ ์†์—์„œ ๋œป๋ฐ–์˜ ๊ธฐํšŒ๋ฅผ ํฌ์ฐฉํ•˜๋Š” ์ด๋“ค์ด ์žˆ์Šต๋‹ˆ๋‹ค. ๋ฏธ๊ตญ์˜ ๊ธฐ์ˆ  ํ—ˆ๋ธŒ์ธ ์‹ค๋ฆฌ์ฝ˜๋ฐธ๋ฆฌ์˜ ๊ฑฐ๋Œ€ ๊ธฐ์—…๋“ค์€ ๊ธ‰๋ณ€ํ•˜๋Š” ์•ˆ๋ณด ํ™˜๊ฒฝ ์†์—์„œ โ€˜์ „์Ÿ์˜ ๊ธฐ์ˆ โ€™์— ๊นŠ์ด ๋›ฐ์–ด๋“ค๋ฉฐ ์ƒˆ๋กœ์šด ํ™ฉ๊ธˆ๊ธฐ๋ฅผ ๋งž์ดํ•˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค. ํ•œํŽธ, ๋„๋„๋“œ ํŠธ๋Ÿผํ”„ ์ „ ๋Œ€ํ†ต๋ น์˜ ๊ฐ€์กฑ ์‚ฌ์—…์ฒด์ธ ํŠธ๋Ÿผํ”„ ์กฐ์ง(Trump Organization)์€ ๋ฐœ๋ฆฌ๋‚˜ ๋ชฐ๋””๋ธŒ ๊ฐ™์€ ํ™”๋ คํ•œ ํœด์–‘์ง€๊ฐ€ ์•„๋‹Œ, ๋ฃจ๋งˆ๋‹ˆ์•„์˜ ์™ธ๋”ด ์ง€์—ญ ํŠธ๋ž€์‹ค๋ฐ”๋‹ˆ์•„์—์„œ ์˜์™ธ์˜ ๋ถ€๋™์‚ฐ ํ”„๋กœ์ ํŠธ๋ฅผ ์ถ”์ง„ํ•˜๋ฉฐ ์„ธ๊ฐ„์˜ ์ด๋ชฉ์„ ์ง‘์ค‘์‹œํ‚ค๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค. ๋‰ด์š•ํƒ€์ž„์Šค(The Times)์˜ ์ตœ์‹  ๋ณด๋„๋Š” ์ด ๋‘ ๊ฐ€์ง€ ์ƒ๋ฐ˜๋œ ํ๋ฆ„์ด ์–ด๋–ป๊ฒŒ ์˜ค๋Š˜๋‚ ์˜ ์„ธ๊ณ„๋ฅผ ์žฌํŽธํ•˜๊ณ  ์žˆ๋Š”์ง€ ์กฐ๋ช…ํ•ฉ๋‹ˆ๋‹ค.

์ด๋ž€ ์ „์Ÿ์˜ ๊ทธ๋ฆผ์ž, ์‹ค๋ฆฌ์ฝ˜๋ฐธ๋ฆฌ์˜ ์ƒˆ๋กœ์šด ํ™œ๋กœ

์ด๋ž€๊ณผ์˜ ์ „์Ÿ์ด 3์ฃผ์ฐจ์— ์ ‘์–ด๋“ค๋ฉด์„œ, ๋ฏธ๊ตญ์€ ๊ตฐ์‚ฌ์ , ๊ฒฝ์ œ์ , ๊ธฐ์ˆ ์  ์ธก๋ฉด์—์„œ ์ค‘๋Œ€ํ•œ ๊ฒฐ์ •์„ ๋‚ด๋ฆฌ๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค. ํŠธ๋Ÿผํ”„ ๋Œ€ํ†ต๋ น์€ ์ด๋ž€์— ๋ฏธ๊ตฐ ๋ณ‘๋ ฅ์„ ํŒŒ๋ณ‘ํ•  ๊ฐ€๋Šฅ์„ฑ์„ โ€œ๋ฐฐ์ œํ•˜์ง€ ์•Š๋Š”๋‹คโ€๊ณ  ๋ฐํžˆ๋ฉฐ, ํ•„์š”ํ•˜๋‹ค๋ฉด ๊ตฐ์‚ฌ๋ ฅ ๋ฐฐ์น˜๋ฅผ โ€œ๋‘๋ ค์›Œํ•˜์ง€ ์•Š๋Š”๋‹คโ€๊ณ  ๊ฐ•์กฐํ–ˆ์Šต๋‹ˆ๋‹ค. ์ด๋Š” ์ค‘๋™ ์ง€์—ญ์˜ ๊ธด์žฅ๊ฐ์„ ํ•œ์ธต ๊ณ ์กฐ์‹œํ‚ค๋Š” ๋ฐœ์–ธ์œผ๋กœ ํ•ด์„๋ฉ๋‹ˆ๋‹ค.

์—๋„ˆ์ง€ ์œ„๊ธฐ ํ•ด๊ฒฐ์„ ์œ„ํ•œ ๋ฐฑ์•…๊ด€์˜ ๊ณ ์‹ฌ๋„ ๊นŠ์–ด์ง€๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค. ๋ฏธ ์žฌ๋ฌด์žฅ๊ด€ ์Šค์ฝง ๋ฒ ์‚ฐํŠธ(Scott Bessant)๋Š” ํ–ฅํ›„ ์ด๋ž€์‚ฐ ์›์œ ์— ๋Œ€ํ•œ ์ œ์žฌ๋ฅผ ํ•ด์ œํ•  ๊ณ„ํš์ด๋ผ๊ณ  ๋ฐํ˜”์Šต๋‹ˆ๋‹ค. ์ด๋Š” ์•ฝ 1์–ต 4์ฒœ๋งŒ ๋ฐฐ๋Ÿด์— ๋‹ฌํ•˜๋Š” ์›์œ ๊ฐ€ ์‹œ์žฅ์— ํ’€๋ฆด ์ˆ˜ ์žˆ์Œ์„ ์˜๋ฏธํ•˜๋ฉฐ, ์ˆ˜๋…„๊ฐ„ ์ง€์†๋œ ๋ฏธ๊ตญ์˜ ๋Œ€์ด๋ž€ ์ •์ฑ…์„ ๋’ค์ง‘๋Š” ๊ฒฐ์ •์ž…๋‹ˆ๋‹ค. ๋ฐฑ์•…๊ด€์€ ์œ ๊ฐ€ ์•ˆ์ •์„ ์œ„ํ•ด ํ•„์‚ฌ์ ์ธ ๋…ธ๋ ฅ์„ ๊ธฐ์šธ์ด๊ณ  ์žˆ์ง€๋งŒ, ์ด๋ž€๊ณผ์˜ ์ „์Ÿ ์ค‘์—๋„ ์ด๋ž€์ด ๋” ๋งŽ์€ ์›์œ ๋ฅผ ํŒ๋งคํ•˜๋„๋ก ์žฅ๋ คํ•˜๋Š” ์—ญ์„ค์ ์ธ ์ƒํ™ฉ์„ ์ดˆ๋ž˜ํ•  ์ˆ˜ ์žˆ๋‹ค๋Š” ์ง€์ ์ด ๋‚˜์˜ต๋‹ˆ๋‹ค.

ํ•œํŽธ, ๋ฏธ๊ตญ ๊ตญ๋ฐฉ๋ถ€๋Š” ์ „์Ÿ ์ง€์†์„ ์œ„ํ•ด 2์ฒœ์–ต ๋‹ฌ๋Ÿฌ์— ๋‹ฌํ•˜๋Š” ๋ง‰๋Œ€ํ•œ ์ถ”๊ฐ€ ์˜ˆ์‚ฐ์„ ์˜ํšŒ์— ์š”์ฒญํ–ˆ์Šต๋‹ˆ๋‹ค. ์ด๋Š” ์—ฐ๊ฐ„ ์ „์ฒด ๊ตญ๋ฐฉ ์˜ˆ์‚ฐ์˜ ๊ฑฐ์˜ 4๋ถ„์˜ 1์— ํ•ด๋‹นํ•˜๋Š” ๊ทœ๋ชจ๋กœ, ๊ตญ๋ฐฉ์žฅ๊ด€ ํ”ผํŠธ ํ—‰์Šค(Pete Hgse)๋Š” ์ด ๊ธˆ์•ก์ด ๋ณ€๊ฒฝ๋  ์ˆ˜ ์žˆ์Œ์„ ์‹œ์‚ฌํ–ˆ์Šต๋‹ˆ๋‹ค. ์ด๋ฏธ ๋ถ„์—ด๋œ ์—ฌ๋ก  ์†์—์„œ ์ด ๋ง‰๋Œ€ํ•œ ์˜ˆ์‚ฐ ์š”์ฒญ์€ ์˜ํšŒ ๋‚ด ์˜จ๊ฑด ๊ณตํ™”๋‹น ์˜์›๋“ค์˜ ์šฐ๋ ค๋ฅผ ์‚ฌ๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค.

์ด๋Ÿฌํ•œ ๊ฒฉ๋™ ์†์—์„œ ํ•œํŽธ์œผ๋กœ๋Š” ์‹ค๋ฆฌ์ฝ˜๋ฐธ๋ฆฌ์˜ ๊ธฐ์ˆ  ๊ธฐ์—…๋“ค์ด ์ƒˆ๋กœ์šด ๊ธฐํšŒ๋ฅผ ํฌ์ฐฉํ•˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค. ๊ณผ๊ฑฐ ์‹ค๋ฆฌ์ฝ˜๋ฐธ๋ฆฌ์—์„œ ๊ตญ๋ฐฉ ๊ธฐ์ˆ ์€ โ€˜์•…ํ•ด์ง€์ง€ ๋ง๋ผ(Do No Evil)โ€˜๋Š” ๊ตฌ๊ธ€์˜ ๋ชจํ† ์ฒ˜๋Ÿผ ๊ธฐํ”ผ ๋Œ€์ƒ์ด์—ˆ์Šต๋‹ˆ๋‹ค. ๊ตฌ๊ธ€ ์ง์›๋“ค์€ ๊ตญ๋ฐฉ๋ถ€์™€์˜ ํ˜‘๋ ฅ์— ๋ฐ˜๋Œ€ํ•˜๋Š” ์‹œ์œ„๋ฅผ ๋ฒŒ์ด๊ธฐ๋„ ํ–ˆ์Šต๋‹ˆ๋‹ค. ๊ทธ๋Ÿฌ๋‚˜ ๋ช‡ ๋…„ ์ „๋ถ€ํ„ฐ ์ด๋Ÿฌํ•œ ๋ถ„์œ„๊ธฐ๋Š” ๊ธ‰๋ณ€ํ–ˆ์Šต๋‹ˆ๋‹ค. ๊ตฌ๊ธ€, ๋ฉ”ํƒ€(Meta), OpenAI ๊ฐ™์€ ๊ฑฐ๋Œ€ ๊ธฐ์—…๋ถ€ํ„ฐ ์†Œ๊ทœ๋ชจ ์Šคํƒ€ํŠธ์—…์— ์ด๋ฅด๊ธฐ๊นŒ์ง€ ์ˆ˜๋งŽ์€ ๊ธฐ์ˆ  ๊ธฐ์—…๋“ค์ด ๊ตญ๋ฐฉ ๊ธฐ์ˆ  ๋ถ„์•ผ๋กœ ๋ฐฉํ–ฅ์„ ์ „ํ™˜ํ•˜๊ธฐ ์‹œ์ž‘ํ–ˆ์Šต๋‹ˆ๋‹ค.

์ด๋“ค ๊ธฐ์—…์€ ์ˆ˜์–ต์—์„œ ์ˆ˜์‹ญ์–ต ๋‹ฌ๋Ÿฌ์— ์ด๋ฅด๋Š” ๊ณ„์•ฝ์„ ํ†ตํ•ด ๋ฌด๊ธฐ ์‹œ์Šคํ…œ๋ถ€ํ„ฐ ๋ฏธ ์ •๋ถ€๋ฅผ ์œ„ํ•œ ์†Œํ”„ํŠธ์›จ์–ด๊นŒ์ง€ ๋‹ค์–‘ํ•œ ๊ธฐ์ˆ ์„ ๊ฐœ๋ฐœํ•˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค. ์˜ˆ๋ฅผ ๋“ค์–ด, ๋ฐ์ดํ„ฐ ๋ถ„์„ ๊ธฐ์—… ํŒ”๋ž€ํ‹ฐ์–ด(Palantir)๋Š” ๋ฏธ ์ •๋ถ€๊ฐ€ ๊ณต์Šต ๋ชฉํ‘œ๋ฅผ ์„ ์ •ํ•˜๋Š” ๋ฐ ๋„์›€์„ ์ฃผ๋Š” โ€˜ํ”„๋กœ์ ํŠธ ๋ฉ”์ด๋ธ(Project Maven)โ€™ ์‹œ์Šคํ…œ์„ ๊ตฌ์ถ•ํ–ˆ์Šต๋‹ˆ๋‹ค. ๊ตฌ๊ธ€๊ณผ OpenAI๋Š” ์ธ๊ณต์ง€๋Šฅ(AI) ๊ธฐ์ˆ ์„ ํ™œ์šฉํ•ด ํ˜„์žฅ ์ง€ํœ˜๊ด€๋“ค์„ ์ง€์›ํ•˜๊ณ  ์žˆ์œผ๋ฉฐ, ๊ตฌ๊ธ€ ์ „ ์ตœ๊ณ ๊ฒฝ์˜์ž ์—๋ฆญ ์Šˆ๋ฏธํŠธ(Eric Schmidt)๊ฐ€ ์ฃผ๋„ํ•˜๋Š” ํ”„๋กœ์ ํŠธ๋Š” ์ด๋ž€์—์„œ ๋ฏธ๊ตฐ ์ž์‚ฐ์ด ์ด๋ž€ ๋“œ๋ก ์— ๋ฐฉ์–ดํ•˜๋Š” ๋ฐ ์‚ฌ์šฉ๋˜๋Š” ๋Œ€๋“œ๋ก  ์‹œ์Šคํ…œ์„ ๊ฐœ๋ฐœํ•˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค. ํŠธ๋Ÿผํ”„ ํ–‰์ •๋ถ€๊ฐ€ 1์กฐ ๋‹ฌ๋Ÿฌ ์ด์ƒ์˜ ๊ตญ๋ฐฉ๋น„๋ฅผ ํ• ๋‹นํ–ˆ์œผ๋ฉฐ, ์ด ์ค‘ ์ƒ๋‹น ๋ถ€๋ถ„์ด ์‹ค๋ฆฌ์ฝ˜๋ฐธ๋ฆฌ๊ฐ€ ๊ฐœ๋ฐœํ•˜๊ณ ์ž ํ•˜๋Š” ์ƒˆ๋กœ์šด ๊ตญ๋ฐฉ ๊ธฐ์ˆ ์— ํˆฌ์ž…๋  ๊ฒƒ์œผ๋กœ ์˜ˆ์ƒ๋˜๋ฉด์„œ, ๊ธฐ์ˆ  ๊ธฐ์—…๋“ค์˜ โ€˜์ „์Ÿ ํŠน์ˆ˜โ€™๋Š” ๋”์šฑ ์ปค์งˆ ์ „๋ง์ž…๋‹ˆ๋‹ค.

ํŠธ๋Ÿผํ”„ ๊ฐ€๋ฌธ์˜ ์ด์ƒ‰ ํˆฌ์ž์ฒ˜, ํŠธ๋ž€์‹ค๋ฐ”๋‹ˆ์•„์˜ ๋ฏธ์Šคํ„ฐ๋ฆฌ

๋‰ด์š•ํƒ€์ž„์Šค์˜ ์ƒˆ๋กœ์šด ์กฐ์‚ฌ์— ๋”ฐ๋ฅด๋ฉด, ํŠธ๋Ÿผํ”„ ์ „ ๋Œ€ํ†ต๋ น์˜ ๊ฐ€์กฑ ์‚ฌ์—…์ฒด์ธ ํŠธ๋Ÿผํ”„ ์กฐ์ง์ด ๋ฃจ๋งˆ๋‹ˆ์•„์—์„œ ๋ฏธ๋ฐœํ‘œ ๋ถ€๋™์‚ฐ ๊ฑฐ๋ž˜๋ฅผ ์ถ”์ง„ ์ค‘์ธ ๊ฒƒ์œผ๋กœ ๋“œ๋Ÿฌ๋‚ฌ์Šต๋‹ˆ๋‹ค. ๋ฐœ๋ฆฌ๋‚˜ ๋ชฐ๋””๋ธŒ์ฒ˜๋Ÿผ ํ™”๋ คํ•˜๊ณ  ์œ ๋ช…ํ•œ ๊ด€๊ด‘์ง€๊ฐ€ ์•„๋‹Œ, ํŠธ๋ž€์‹ค๋ฐ”๋‹ˆ์•„๋ผ๋Š” ์˜์™ธ์˜ ์žฅ์†Œ์ž…๋‹ˆ๋‹ค.

ํ˜„์žฅ์„ ์ง์ ‘ ๋ฐฉ๋ฌธํ•œ ๋‰ด์š•ํƒ€์ž„์Šค ํƒ์‚ฌ ๋ณด๋„ ๊ธฐ์ž ๋ ˆ๋ฒ ์นด ๋ฃจ์ด์ฆˆ(Rebecca Ruiz)๋Š” ๊ทธ๊ณณ์˜ ํ’๊ฒฝ์„ ์ƒ์ƒํ•˜๊ฒŒ ๋ฌ˜์‚ฌํ–ˆ์Šต๋‹ˆ๋‹ค. ๋ถ€์ง€๋Š” ๋ฐ˜์ฏค ๋ฒ„๋ ค์ง„ ์ƒํƒœ์˜€๊ณ , ๊ตฐ์‚ฌ ๊ธฐ์ง€ ์˜†์— ์œ„์น˜ํ–ˆ์œผ๋ฉฐ, ์ด ์ง€์—ญ ์ตœ๋Œ€ ๊ทœ๋ชจ์˜ ๊ฑฐ๋Œ€ํ•œ ์“ฐ๋ ˆ๊ธฐ ๋งค๋ฆฝ์ง€ ๋ช‡ ๊ณณ๊ณผ ์ธ์ ‘ํ•ด ์žˆ์—ˆ์Šต๋‹ˆ๋‹ค. ์ฝ”๋ฅผ ์ฐŒ๋ฅด๋Š” ์•…์ทจ๊ฐ€ ์ง„๋™ํ–ˆ๊ณ , ๋“ค๊ฐœ ๋ฌด๋ฆฌ๊ฐ€ ๋ฐฐํšŒํ•˜๋Š” ๋ชจ์Šต๋„ ๋ชฉ๊ฒฉ๋˜์—ˆ์Šต๋‹ˆ๋‹ค. ์ด๋Ÿฌํ•œ ํ™˜๊ฒฝ์€ ๋Ÿญ์…”๋ฆฌ ์•„ํŒŒํŠธ ๋‹จ์ง€์™€ ๊ณจํ”„ ์ฝ”์Šค๋ฅผ ์ƒ์ƒํ•˜๊ธฐ์—๋Š” ๋งค์šฐ ์ด๋ก€์ ์ธ ์žฅ์†Œ์˜€์Šต๋‹ˆ๋‹ค.

ํŠธ๋Ÿผํ”„ ์กฐ์ง์ด ์™œ ์ด ํŠน์ • ์žฅ์†Œ๋ฅผ ํ”„๋กœ์ ํŠธ ๋ถ€์ง€๋กœ ์„ ํƒํ–ˆ๋Š”์ง€๋Š” ๋ช…ํ™•ํ•˜์ง€ ์•Š์Šต๋‹ˆ๋‹ค. ๊ทธ๋Ÿฌ๋‚˜ ํ•œ ๊ฐ€์ง€ ๋‹จ์„œ๋Š” ํŠธ๋Ÿผํ”„ ์ „ ๋Œ€ํ†ต๋ น์ด ๋ฃจ๋งˆ๋‹ˆ์•„์—์„œ ๋†’์€ ์ธ๊ธฐ๋ฅผ ๋ˆ„๋ฆฌ๊ณ  ์žˆ๋‹ค๋Š” ์ ์ž…๋‹ˆ๋‹ค. ์ง€๋‚œํ•ด ๊ฐค๋Ÿฝ(Gallup) ์—ฌ๋ก ์กฐ์‚ฌ์— ๋”ฐ๋ฅด๋ฉด, ๋ฃจ๋งˆ๋‹ˆ์•„ ๊ตญ๋ฏผ์˜ ์ ˆ๋ฐ˜ ์ด์ƒ์ด ํŠธ๋Ÿผํ”„ ๋Œ€ํ†ต๋ น์˜ ์ง๋ฌด ์ˆ˜ํ–‰์— ๋Œ€ํ•ด ๊ธ์ •์ ์œผ๋กœ ํ‰๊ฐ€ํ–ˆ์œผ๋ฉฐ, ์ด๋Š” ์œ ๋Ÿฝ์—ฐํ•ฉ(EU) ๊ตญ๊ฐ€๋“ค ์‚ฌ์ด์—์„œ ์ด๋ก€์ ์ธ ๊ฒฐ๊ณผ์ž…๋‹ˆ๋‹ค.

๋ ˆ๋ฒ ์นด ๋ฃจ์ด์ฆˆ ๊ธฐ์ž๋Š” ์ด๋Ÿฌํ•œ ์›€์ง์ž„์ด ํŠธ๋Ÿผํ”„ ์ „ ๋Œ€ํ†ต๋ น์˜ ๋‘ ๋ฒˆ์งธ ์ž„๊ธฐ์—์„œ ๋‚˜ํƒ€๋‚œ ํŒจํ„ด์˜ ์—ฐ์žฅ์„ ์ƒ์— ์žˆ๋‹ค๊ณ  ๋ถ„์„ํ•ฉ๋‹ˆ๋‹ค. ์ฆ‰, ํŠธ๋Ÿผํ”„ ์กฐ์ง์€ ๊ทธ๊ฐ€ ์ธ๊ธฐ ์žˆ๋Š” ์ง€์—ญ์„ ์ฐพ์•„ ์ „ ์„ธ๊ณ„๋ฅผ ์—ฌํ–‰ํ•˜๋ฉฐ ์‚ฌ์—… ๊ธฐํšŒ๋ฅผ ๋ชจ์ƒ‰ํ•˜๊ณ  ์žˆ๋‹ค๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค. ๋ฐœ๋ฆฌ์™€ ๋ชฐ๋””๋ธŒ ๊ฐ™์€ ์ „ํ†ต์ ์ธ ๊ณ ๊ธ‰ ํœด์–‘์ง€์™€๋Š” ๊ฑฐ๋ฆฌ๊ฐ€ ๋จผ ํŠธ๋ž€์‹ค๋ฐ”๋‹ˆ์•„์—์„œ ํŠธ๋Ÿผํ”„ ์กฐ์ง์ด ์–ด๋–ค ๊ทธ๋ฆผ์„ ๊ทธ๋ฆฌ๊ณ  ์žˆ๋Š”์ง€์— ๋Œ€ํ•œ ์˜๋ฌธ์€ ์—ฌ์ „ํžˆ ๋‚จ์•„์žˆ์Šต๋‹ˆ๋‹ค.

๊ฒฉ๋™์˜ ์‹œ๋Œ€, ๋ณ€ํ™”์˜ ๋ฌผ๊ฒฐ

์ด๋ž€๊ณผ์˜ ์ „์Ÿ๊ณผ ๊ทธ๋กœ ์ธํ•œ ์‹ค๋ฆฌ์ฝ˜๋ฐธ๋ฆฌ์˜ ๋ถ€์ƒ, ๊ทธ๋ฆฌ๊ณ  ํŠธ๋Ÿผํ”„ ๊ฐ€๋ฌธ์˜ ์˜ˆ์ƒ์น˜ ๋ชปํ•œ ํˆฌ์ž์ฒ˜๋Š” ์˜ค๋Š˜๋‚  ์„ธ๊ณ„๊ฐ€ ์–ผ๋งˆ๋‚˜ ๋น ๋ฅด๊ฒŒ ๋ณ€ํ™”ํ•˜๊ณ  ์˜ˆ์ธก ๋ถˆ๊ฐ€๋Šฅํ•œ ๋ฐฉํ–ฅ์œผ๋กœ ๋‚˜์•„๊ฐ€๊ณ  ์žˆ๋Š”์ง€๋ฅผ ๋ณด์—ฌ์ค๋‹ˆ๋‹ค. ์ „์Ÿ์ด ๊ธฐ์ˆ  ๊ธฐ์—…๋“ค์—๊ฒŒ ์ƒˆ๋กœ์šด ์‹œ์žฅ์„ ์—ด์–ด์ฃผ๊ณ , ์ •์น˜์  ์ธ๊ธฐ๊ฐ€ ๋น„์ฆˆ๋‹ˆ์Šค ๊ธฐํšŒ๋กœ ์ „ํ™˜๋˜๋Š” ์ด ์‹œ๋Œ€์˜ ํ๋ฆ„์€ ์•ž์œผ๋กœ๋„ ์„ธ๊ณ„ ์ •์น˜๊ฒฝ์ œ์— ์ค‘๋Œ€ํ•œ ์˜ํ–ฅ์„ ๋ฏธ์น  ๊ฒƒ์ž…๋‹ˆ๋‹ค. ๋‰ด์š•ํƒ€์ž„์Šค ์›น์‚ฌ์ดํŠธ(nytimes.com)์—์„œ ํŠธ๋Ÿผํ”„ ์กฐ์ง์˜ ํŠธ๋ž€์‹ค๋ฐ”๋‹ˆ์•„ ํ”„๋กœ์ ํŠธ์— ๋Œ€ํ•œ ์ „์ฒด ์กฐ์‚ฌ ๋‚ด์šฉ๊ณผ ํ˜„์žฅ ์‚ฌ์ง„์„ ํ™•์ธํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.