English
Based on โ3 experts explain everything you need to know about lonelinessโ from Big Think Watch the original video
Unmasking Loneliness: Why Being Alone Isnโt Always Bad, and How to Cultivate Deeper Connections
From a young age, many of us learn a subtle but powerful stigma: if youโre alone, something must be wrong. This ingrained belief shapes our perceptions and the stories we tell ourselves about feeling lonely, often leading to negative thought patterns and behaviors. But what if this widespread understanding of loneliness is incomplete, even misleading? What if loneliness isnโt a flaw, but a signal? And what if, contrary to popular belief, spending time alone can actually be a profoundly positive experience?
Experts across psychology, neuroscience, and social science are challenging conventional wisdom, revealing that loneliness is a far more complex and impactful phenomenon than we often realize. Itโs not merely a fleeting emotion or a sign of social inadequacy; itโs a critical piece of information about our needs and a potent force that can profoundly affect our physical and mental well-being.
The Stigma and Self-Perception: How Our Minds Shape Our Loneliness
The learned stigma surrounding solitude is a powerful driver of how we experience loneliness. When we feel lonely, our minds often gravitate towards rumination and catastrophizing in social situations. We get caught up in negative thought patterns and beliefs that donโt just influence our behaviors, but also actively impact our brains and bodies.
However, experts argue that this internal narrative needs a radical shift. โLoneliness is not a reflection on who we are,โ one expert explains. โItโs a reflection of what we need. Itโs information. Itโs data.โ This reframing is crucial. Loneliness, at its core, is an emotional signal, much like hunger or thirst, indicating a need for connection. When viewed this way, it transforms from a source of shame into an opportunity for self-discovery and growth.
A Silent Health Crisis: The Profound Physical Toll of Loneliness
Beyond the psychological distress, the physical consequences of chronic loneliness are startling and severe. Studies suggest that โbeing without a close friend, being lonely, is as bad for your health as smoking 15 cigarettes a day.โ This is not hyperbole; the physiological impact is well-documented.
When loneliness triggers negative thought patterns and emotions, it initiates a stress response in our bodies. This response is associated with heightened levels of cortisol, the bodyโs primary stress hormone, and increased inflammation. Over time, this chronic stress and inflammation can weaken our immune systems, making us more susceptible to various diseases.
Neuroscience research provides compelling evidence of this connection between social connection and physical experience. In one study, participants underwent mild electric shocks while looking at photos of either their romantic partner or a complete stranger. Researchers found that those looking at a photo of their romantic partners reported feeling less fearful, and brain activity in regions associated with pain was significantly lower. This highlights how our perception of pain and threat is profoundly influenced by whether we feel connected to another person.
This research underscores a vital point: health is not solely physical and mental; it is intrinsically social. Understanding this broader definition of health can fundamentally change our behaviors. Cultivating rich friendships, embedding ourselves in our communities, and fostering a strong sense of support are not just pleasant additions to life; they are foundational elements that can โcompletely transform the way that you experience the world.โ
The โFriendship Recessionโ: A Modern Challenge to Connection
Despite the clear importance of social bonds, modern society appears to be facing what scholar Daniel Cox of the American Enterprise Institute terms a โfriendship recession.โ This describes a concerning rise in the number of people who lack a sufficient number of close friends, leaving them with fewer individuals to turn to in times of crisis.
The statistics are sobering: today, 15% of young men report not having a close friend, a dramatic increase from just 3% in the 1990s. While the pandemic impacted everyone, it disproportionately affected women, with more than half reporting they lost touch with at least some of their friends.
Several societal shifts contribute to this decline. Weโve witnessed a decrease in participation in many traditional institutions that historically fostered connection, including family structures (people marrying later or not at all), religious organizations, and even aspects of the labor market. As these traditional anchors weaken, the need for robust social relationships and connections outside these institutions becomes even more critical. This is where friends play an โhugely importantโ role.
Without these crucial connections, individuals risk not only economic or social isolation but also profound sadness. And as experts confirm, โbeing sad, it turns out, is also bad in terms of your physical as well as emotional health.โ The potential future, a โdystopian version,โ could see a world of โatomized individuals without friends, isolated, perhaps in ill health.โ This stark warning emphasizes why recognizing friendship as โincredibly important for human flourishingโ is not just a feel-good sentiment, but a societal imperative.
Breaking the Cycle: Reframing Solitude and Seeking Connection
The pervasive negative narrative around being alone is a significant hurdle. Research indicates that the media describes experiences of being alone approximately ten times more negatively than positively. This constant messaging reinforces the stigma and makes it harder for individuals to find peace or joy in solitude.
However, thereโs a powerful opportunity for intervention: shifting our beliefs about being alone. Studies tracking individuals over time reveal a fascinating insight: if youโre the kind of person who believes being alone is good for you, you actually feel good when you spend time alone. This suggests that our mindset can profoundly alter our experience of solitude.
Instead of viewing alone time as a reflection of something wrong with us or our circumstances, we can reframe it as an opportunity. It can be a chance โto be creative, to be alone with our thoughts in a pleasant way.โ This mental shift can transform a potentially negative experience into one of rejuvenation and self-discovery.
Yet, actively seeking new connections also requires courage and vulnerability. One of the necessary steps to making a friend is โadmitting that you want to make a friend. To being open to that.โ As we get older, there can be a sense of shame associated with not having enough friends, making it even harder to express this need. โActually saying โI need a friendโ is maybe one of the hardest sentences that any human being can utter,โ an expert notes. But itโs a crucial first step toward building the connections we inherently need. The more we openly discuss these challenges, the more they penetrate the broader culture, normalizing vulnerability and fostering supportive environments.
Cultivating Optimal Social Health: Itโs Personal
Ultimately, the path to overcoming loneliness and cultivating deeper connections is highly personal. Experts emphasize that โconnection with yourself is as important as connection with other people.โ Understanding your own needs, preferences, and social energy levels is paramount.
โBeing optimally socially healthy for you might be different than someone else,โ based on how much socializing you enjoy and what kinds of connection feel most fulfilling. This means thereโs no universal prescription for social well-being. Instead, loneliness can be seen as โjust an opportunity to go further, to explore your social health, to be a scientist in the laboratory that is your social life.โ
By adopting this curious, self-aware approach, we can broaden our โperception boxโ and actively work to feel more connected, overcoming the isolating experience of loneliness. Itโs about tailoring our social lives to fit our unique needs, rather than conforming to external expectations or succumbing to societal stigmas.
In a world increasingly grappling with the profound impacts of loneliness, the message from experts is clear: this isnโt a personal failing, but a complex challenge with deep roots in our biology, psychology, and societal structures. By understanding its true nature, reframing our perceptions of solitude, embracing vulnerability, and actively nurturing our social health, we can navigate the modern landscape of connection and truly flourish.
Based on โAI Automated Everything. Why Is There More Work?โ from Every Watch the original video
The Automation Paradox: Why AI Creates More Work for Humans
Artificial intelligence promises a future where machines handle the heavy lifting, freeing humanity from repetitive tasks and perhaps even complex thought. Yet, for many, the reality feels starkly different: AI is automating everything, but the workload for humans seems to be increasing, not decreasing. This isnโt just a feeling; itโs a profound paradox unfolding in companies at the cutting edge of AI adoption.
While figures like Dario Amodei of Anthropic warn that AI could eliminate over 50% of entry-level white-collar jobs, and Ken Griffin, head of the colossal hedge fund Citadel, expresses awe at agents performing โextraordinarily high-skilled jobs,โ the lived experience within companies deeply integrated with AI tells a different story. The publicโs widespread fear of job displacement stands in stark contrast to the surprising reality: AI, in practice, is amplifying the demand for human expertise.
This counterintuitive truth is at the heart of the experience at Every, a company that has been deeply immersed in AI and agent technology since before ChatGPTโs public release. For Every, a 25-person team, embracing AI isnโt about replacing humans; itโs about fundamentally reshaping work, leading to a surge in human hiring even as AI agents become ubiquitous.
AI at the Coalface: How โEveryโ Works with Agents
Every operates as a microcosm, pushing the boundaries of whatโs possible with AI. Their internal experience reveals two primary modes of working with AI agents, both of which invariably require significant human involvement.
The first, and more predictable, mode is delegation to specialized AI bots. At Every, itโs common to โ@mentionโ a bot in Slack to handle specific tasks. They have:
- Claudie: An internal AI that manages their entire consulting business, from proposals to client interactions.
- Andy: An editorial assistant, helping with content creation.
- Victor: Another agent, used for various tasks, reflecting the growing ecosystem of specialized AI tools.
These agents perform a vast array of functions, from marketing and brand research to A/B testing YouTube thumbnails, drafting consulting proposals, and even creating initial client presentation decks. They provide immense leverage, taking on tasks that previously required dedicated human effort.
The second, and arguably more transformative, mode is agent orchestration software. Tools like Codex and Co-work are evolving into the operating system for knowledge work. Here, an agent lives on the userโs computer, accessed through a graphical interface, with full access to files, applications, and a built-in browser. The speaker at Every describes conducting almost all their work within such a system: managing inboxes, writing, coding, and even analyzing the companyโs profit and loss statements.
Whatโs crucial in both scenarios is the โhuman in the loop.โ Despite the agentsโ capabilities, they are not autonomous. Claudie, for instance, thrives because Nitesh, a senior AI engineer, constantly monitors its performance, identifying shortcomings and helping it improve. Itโs not Nitesh doing all the work, but rather Nitesh enabling Claudie to provide significant leverage. Similarly, the agent orchestration systems are most powerful when thereโs a constant, interactive collaboration between the human and the AI. The idea that an agent can simply be โset up and forgottenโ is a widespread misconception; the reality is far more nuanced and human-centric.
Why Automation Breeds More Work, Not Less
The paradox deepens when we understand the ripple effects of making โyesterdayโs competenceโ cheap and accessible.
Yesterdayโs Competence, Todayโs Commodity
AI models are trained on vast datasets, which are essentially the โvisible residue of competence.โ This means that abilities once rare and expensiveโlike writing a complex pull request for a codebaseโare now democratized. Suddenly, anyone can, with the help of an AI, generate code, create images, or draft sophisticated text. This immediate accessibility causes adoption to skyrocket. Everyone wants to leverage this new, cheap competence. Marketers are generating YouTube thumbnails, operations staff are submitting pull requests, and customer service teams are drafting complex documents.
The Floodgates Open: The Rise of โSlopโ
While this newfound capability is amazing, it often leads to a glut of generic output, affectionately (or perhaps exasperatedly) termed โslop.โ This isnโt necessarily terrible work; itโs often a decent start, more than an individual could have produced before. However, if users rely solely on default AI outputs or lazy prompts, the result is predictable: a mass of similar-looking content, code, or designs. As the speaker notes, โslopโ isnโt about specific errors like overuse of m-dashes or purple accents; itโs the subtle yet pervasive sameness that humans begin to detect when a tool produces similar outputs across diverse contexts. It smells โrat-likeโ because it lacks the unique human touch.
The Expertโs New Mandate: Demand for Difference
When the market is flooded with generic โslop,โ the demand for difference โfor unique, high-quality, and truly valuable workโskyrockets. And who delivers difference? Experts.
Experts, rather than being replaced, find their roles evolving and expanding. They have a few key responses to this new landscape:
- Cleaning the Slop: Many experts initially feel annoyed, as their plate fills with the task of refining AI-generated drafts that are โa good first passโ but far from production-ready. A software engineer, for example, might receive numerous AI-generated pull requests that capture the problem but lack the nuance, efficiency, or architectural integrity required for deployment.
- Building Systems for Leverage: Crucially, experts begin to build systems and processes to absorb and leverage this new wave of AI-generated work. This could be establishing clear guidelines for AI use, creating โvibe-codedโ templates, or defining social contracts for how AI outputs are integrated. These systems transform the raw โslopโ into something truly amazing and useful.
- Standing on the Shoulders of AI: Perhaps most powerfully, experts use this newly cheapened competence as a foundation. They stand on this elevated floor to achieve things that were previously impossible for a single individual. Every, for example, now has several people each running an entire software product by themselvesโa feat unimaginable before AI agents.
In skilled knowledge work, the equation becomes clear: every agent requires a human to be truly effective. Widespread access to yesterdayโs competence dramatically increases the volume of output. This increased volume, in turn, necessitates experts to refine, systematize, and elevate it beyond generic โslop.โ
The Zenoโs Paradox of AI and Expertise
The dynamic between human expertise and AI can be likened to Zenoโs paradox, where Achilles can never quite catch the tortoise. AI (Achilles) is constantly โhoovering upโ and automating past human competence, nipping at the heels of current expertise. Yet, humans (the tortoise) are always moving forward, creating new forms of valuable work just beyond AIโs grasp. The speaker firmly believes AI will not โcatchโ or eliminate human work because truly valuable output inherently requires human involvement. Even with the advent of AGIโdefined here as an open-ended, always-on, economically viable AI that constantly finds new goalsโhuman input remains indispensable.
This isnโt to say jobs wonโt change; indeed, the speakerโs own role has been completely transformed. The specific implications will vary across different sectors of the economy. However, the core insight from companies actively pushing the boundaries of AI is clear: the widespread fear of job elimination overlooks the critical human element that makes AI outputs valuable.
Riding the Models: A Path Forward
For individuals, the message is empowering and straightforward: learn how to use AI. โRide the models,โ as the speaker advises. As AI models improve, they package more and more โcheap human competence,โ offering unprecedented power to curious and intelligent individuals. For those who embrace these tools, AI becomes a โsuper toolโ to accomplish more interesting, complex, and valuable work. The demand for such individuals, capable of leveraging AI to create genuine difference, is poised to surge.
The discourse around AI and work is complex, but practical experience offers a compelling counter-narrative to the prevailing anxieties. By understanding how AI truly integrates into the workflowโnot as a replacement, but as an amplifier and a demand generator for human expertiseโwe can move towards a more informed and optimistic future of work.
Based on โWhy the U.S. Just Indicted Cubaโs Former Presidentโ from New York Times Podcasts Watch the original video
Beyond the Charges: Unpacking Americaโs High-Stakes Gambit in Cuba
On a seemingly ordinary Wednesday, the United States Department of Justice announced an indictment that sent shockwaves through the Cuban-American community and reverberated across the geopolitical landscape. Raul Castro, the 94-year-old former leader and Communist General of Cuba, was charged with murder, among other grave offenses. These charges, stemming from a decades-old incident where three Americans were killed by the Cuban government, represent more than just a pursuit of justice; they are the latest, and perhaps most dramatic, escalation in the Trump administrationโs relentless campaign to force political change on the communist island.
The story behind this unprecedented indictment is a complex tapestry woven with historical grievances, political symbolism, and a calculated foreign policy strategy. It begins not in the halls of power, but in a highly charged public ceremony in Miami, a city synonymous with Cuban exile.
The Freedom Tower: A Stage for Justice and Symbolism
The Department of Justice press conference, as described by New York Times reporter Frenchie Robas, was anything but typical. Instead of a sterile courthouse, it was held at the iconic Freedom Tower in Miami. This choice of venue was deeply significant. In the 1960s and early 70s, the Freedom Tower served as the Cuban Refugee Center โ an โEllis Islandโ for thousands fleeing Fidel Castroโs revolution. Its cool, old-world architecture now stood as a potent symbol of hope and exile for an entire community.
The timing was equally deliberate: May 20th, Cuban Independence Day. โSymbolism squared,โ Robas noted, for an event attended by hundreds of politicians, mayors, federal prosecutors, and a โwhoโs whoโ of the Cuban exile community. It was, unequivocally, a place to be.
The atmosphere crackled with anticipation. Maline Pumarga, a prominent Cuban-American and president of Miami Dade College (which now owns the tower), opened the proceedings. Her presence underscored that this was not merely a judicial announcement, but a profound community event. โTo our brothers and sisters in Cuba, we hear you. Freedom will come,โ she declared, echoing the long-held aspirations of the exiles.
When Acting Attorney General Todd Blanch was introduced, the room erupted in a standing ovation before he even spoke. The crowd knew what was coming. โToday, we are announcing an indictment charging Raul Castro and several others with conspiracy to kill US nationals,โ Blanch stated, his words barely audible over another wave of applause and cheers. The charges were specific: conspiracy to kill US nationals, two counts of destruction of aircraft, and four counts of murder. For many in that room, it was a day they had waited 30 years to see.
Brothers to the Rescue: A Tragic Chapter in US-Cuba Relations
To understand the gravity of these charges, one must revisit the incident at their core: the February 24, 1996, shootdown of two civilian planes. This tragedy did not occur in a vacuum but was born out of the desperate humanitarian crisis of the early 1990s.
During this period, thousands of Cubans, desperate to escape the islandโs economic hardship, were risking their lives at sea. They fashioned makeshift rafts from inner tubes, old cars, and whatever debris they could find, attempting the perilous journey across the Straits of Florida. This became known as the โCuban rafter crisis,โ overwhelming the U.S. Coast Guard.
In response, a humanitarian organization called Brothers to the Rescue emerged. Composed of Cuban-American pilots, they raised money to buy small planes and flew over the treacherous waters between Florida and Cuba, spotting rafters and calling for Coast Guard assistance. They saved hundreds, if not thousands, of lives, conducting over 2,000 flights.
However, the mission of Brothers to the Rescue evolved. After the Clinton administration and the Cuban government reached an accord to stem the flow of rafters, the organizationโs original purpose diminished. They began to engage in more provocative actions, dropping leaflets and religious medallions over Cuba โ protest actions that the Cuban government deemed illegal incursions into its airspace. Havana repeatedly warned the Clinton administration to stop these flights, and even within the U.S., the organization was cautioned against continuing. But its founder, Jose Basulto, firmly believed it was his right as a โfree Cubanโ to continue.
On that fateful day in 1996, Basulto and two other planes flew towards Cuba. They crossed the 24th parallel, international airspace, but an โactivated zoneโ according to the Cubans. Basulto, following protocol, announced himself to Cuban air traffic control. His defiant response to their warnings โ โThis is my right as a free Cubanโ โ sealed the fate of two planes. The Cuban military shot them out of the sky, killing four people, including three U.S. citizens.
The shock and grief in South Florida were immense. Frenchie Robas vividly recalled attending a memorial service at the Orange Bowl, where the names of the victims were called out, each met with a collective, tearful โpresente.โ It was a moment etched in the memory of a community.
Raul Castroโs link to this tragedy is direct. As Cubaโs Defense Minister at the time, he was widely understood to have given the order for the shootdown. โBig pilots in Cuba are not going to be shooting planes out of the sky without very clear chain of command orders,โ Robas explained, leaving little doubt about his responsibility.
Why Now? The Trump Administrationโs Vice Grip
The obvious question arises: if Castroโs culpability was so clear, why did it take decades for the U.S. to bring charges? Former federal prosecutors indicated that such an indictment would have been highly destabilizing to diplomatic relations and the region, even during periods when formal ties were minimal.
The answer, then, lies in the current political climate and the Trump administrationโs aggressive posture towards Cuba. This indictment is seen as the โlatest step in that escalating campaign,โ according to Robas. The administration has systematically tightened its grip, cutting off oil supplies, leading to dire consequences on the island. Havana now experiences electricity outages for 22 hours a day. People recount harrowing stories of waking in the middle of the night to charge phones and cook meals, walking hours to work, and losing livelihoods dependent on power and internet.
The blame for this suffering is sharply divided along generational lines. Older Cubans, influenced by decades of government rhetoric, often blame the United States. Younger Cubans, however, are increasingly pointing fingers at their own governmentโs mismanagement. The Cuban government, for its part, dismisses the U.S. pressure as an โillegal genocide,โ refusing to acknowledge its own errors and blaming the Trump administration entirely. They are a regime accustomed to weathering severe hardship, and many believe the Trump administration underestimates their resolve to cling to power at any cost.
The Endgame: Extraction or Concession?
The indictment raises a critical question: is this a prelude to military action, echoing the U.S. indictment and attempted extraction of Venezuelan leader Nicolรกs Maduro? Julian Barnes, who covers the U.S. intelligence community, notes that President Trump โwants the optionโ for a military extraction of Raul Castro. However, the practicalities are complex. Special operations forces are largely engaged elsewhere, and Cuba presents a different scenario than Venezuela. Castro is no longer the official head of state; his removal, while politically influential, would not necessarily trigger an immediate change in formal leadership. Experts view it more as a โthreat, as pressure, as a signalโ rather than an imminent military operation.
So, if not military extraction, what does the U.S. truly want from Cuba? The administrationโs public comments have been vague, but recent diplomatic moves have provided clearer answers. CIA Director John Ratcliffeโs trip to Cuba, the first by a Trump cabinet official, delivered a direct message to Castroโs grandson and high-ranking Cuban intelligence officials.
The primary, concrete demand: Cuba โcanโt be a platform for Americaโs adversaries.โ This is a clear reference to the Chinese and Russian intelligence posts on the island, which are used to intercept U.S. communications, including those from military commands. โRadcliffe basically told the Cubans, โLook, if you want to do a deal with us, you got to kick our adversaries out. Theyโre spying on us from your country. They have to go,โโ Barnes revealed. This is an achievable step, a path for Cuba to reduce intelligence collection or even close the bases.
Beyond this, the economic and political demands are broader, ranging from dismantling the military-industrial conglomerate that controls much of the Cuban economy to promoting more pragmatic politicians and taking incremental steps towards free elections at the local level. These โnarrower goals,โ while not outright regime change, would represent meaningful progress.
Trumpโs Search for a Win
For many Cuban-Americans, anything short of ending communist rule will feel inadequate. However, the Trump administration, having faced challenges with expansive foreign policy goals (like in Iran, where a push for regime change proved messy and inconclusive), may be seeking a more achievable โwin.โ
โWhen he has drawn the goals narrowly, he has seen more success,โ Barnes observed. While Iran remains a difficult problem, Cuba, a long-standing โthorn in Americaโs side,โ presents an opportunity. Trump, known for his desire to cut through bureaucracy and solve long-standing problems, sees this as a chance to achieve what generations of American leaders have not.
The pressure campaign, some in the administration believe, is working better than expected. In a landscape of complex international challenges, Cuba could provide a much-needed foreign policy victory. โIf we know anything about Donald Trump, heโs always looking for a win,โ Barnes concluded.
The indictment of Raul Castro, therefore, is more than a historical reckoning. Itโs a calculated move on a grand chessboard, blending the pursuit of justice for a tragic past with a modern, high-stakes geopolitical strategy. Whether it leads to a fundamental shift in Cuba or merely another chapter in the islandโs long struggle against external pressure remains to be seen, but the stakes for both the Cuban people and the Trump administration are undeniably high.
ํ๊ตญ์ด
โ3 experts explain everything you need to know about lonelinessโ โ Big Think ๊ธฐ๋ฐ ๊ธฐ์ฌ ์๋ณธ ์์ ๋ณด๊ธฐ
์ธ๋ก์์ ๋น์ ์๊ฒ ๋ฌด์์ ๋งํ๋๊ฐ? ์ ๋ฌธ๊ฐ๋ค์ด ์ ํ๋ ์ธ๋ก์์ ๋ํ ๋ชจ๋ ๊ฒ
โํผ์ ์์ผ๋ฉด ๋ญ๊ฐ ์๋ชป๋ ๊ฒโ์ด๋ผ๋ ์ฌํ์ ๋์ธ. ์ฐ๋ฆฌ๋ ์ด๋ฆด ๋๋ถํฐ ์ธ๋ก์์ ๋ถ์ ์ ์ธ ๊ฐ์ ์ผ๋ก ํ์ตํฉ๋๋ค. ํ์ง๋ง ์ธ๋ก์์ ๋จ์ํ ํผํด์ผ ํ ๊ฐ์ ์ด ์๋๋ผ, ์ฐ๋ฆฌ์๊ฒ ์ค์ํ ๋ฉ์์ง๋ฅผ ์ ๋ฌํ๋ ์ ๋ณด์ด์ ๋ฐ์ดํฐ์ ๋๋ค. ๋น ์ฑํฌ(Big Think) ์ฑ๋์ ์์์์ ์ธ ๋ช ์ ์ ๋ฌธ๊ฐ๋ ์ธ๋ก์์ ๋ํ ์คํด๋ฅผ ํ๊ณ , ๊ทธ ๋ณธ์ง๊ณผ ์ฐ๋ฆฌ ์ถ์ ๋ฏธ์น๋ ์ํฅ, ๊ทธ๋ฆฌ๊ณ ๊ฑด๊ฐํ๊ฒ ์ธ๋ก์์ ๋ค๋ฃจ๋ ๋ฐฉ๋ฒ์ ์ฌ์ธต์ ์ผ๋ก ์ค๋ช ํฉ๋๋ค.
์ธ๋ก์์ ๋ํ ์คํด์ ์ฌํ์ ๋์ธ
์ฐ๋ฆฌ๋ ์ข ์ข ํผ์ ์๋ ๊ฒ์ ๋ํ ๋ถ์ ์ ์ธ ์ธ์์ ๋ด๋ฉดํํฉ๋๋ค. โํผ์ ์์ผ๋ฉด ๋ญ๊ฐ ์๋ชป๋ ๊ฒโ์ด๋ผ๋ ์ฌํ์ ๋์ธ์ ์ฐ๋ฆฌ ์ค์ค๋ก ์ธ๋ก์์ ๊ฒฝํํ ๋ ํ์ฑํ๋ ์ด์ผ๊ธฐ์ ์ธ์์ ๋ฐ๊พธ์ด ๋์ต๋๋ค. ์ด๋ฌํ ์๊ธฐ ์์ฌ(self-narrative)๋ ์ฐ๋ฆฌ๊ฐ ์ธ์์์ ํ๋ํ๋ ๋ฐฉ์์ ์ํฅ์ ๋ฏธ์น ๋ฟ๋ง ์๋๋ผ, ์ฐ๋ฆฌ์ ๋์ ์ ์ฒด์๋ ์ํฅ์ ์ค๋๋ค.
์ธ๋ก์์ ๋๋ ๋ ์ฐ๋ฆฌ๋ ์ฌํ์ ์ํฉ์์ ์ฝ๊ฒ ๋ฐ์ถํ๊ณ (ruminate), ์ต์ ์ ์ํฉ์ ๊ฐ์ ํ๋ฉฐ(catastrophize), ๋ถ์ ์ ์ธ ์ฌ๊ณ ํจํด๊ณผ ์ ๋ ์ ๊ฐํ๊ธฐ ์ฝ์ต๋๋ค. ์ด๋ฌํ ์๊ฐ์ ์ฐ๋ฆฌ์ ํ๋์ ์ง๋ฐฐํ๊ณ , ๋์๊ฐ ์ฐ๋ฆฌ์ ๋์ ์ ์ฒด ๊ธฐ๋ฅ์๋ ๋ถ์ ์ ์ธ ์ํฅ์ ๋ฏธ์นฉ๋๋ค. ์ธ๋ก์์ ๋น์ ์ ์ธ๊ฒฉ์ ๋ํ ๋ฐ์์ด ์๋๋ผ, ๋น์ ์ด ๋ฌด์์ ํ์๋ก ํ๋์ง์ ๋ํ ๋ฐ์์ด๋ฉฐ, ์ค์ํ ์ ๋ณด์์ ๊นจ๋ซ๋ ๊ฒ์ด ์ค์ํฉ๋๋ค.
๋ชธ๊ณผ ๋ง์์ ๋ณ๋ค๊ฒ ํ๋ ์ธ๋ก์์ ๊ทธ๋ฆผ์
์ธ๋ก์์ด ์ฐ๋ฆฌ์ ๊ฑด๊ฐ์ ๋ฏธ์น๋ ์ํฅ์ ์์ ์ด์์ผ๋ก ์ฌ๊ฐํฉ๋๋ค. ํ ์ฐ๊ตฌ์ ๋ฐ๋ฅด๋ฉด, ์นํ ์น๊ตฌ๊ฐ ์๋ ์ธ๋ก์์ ํ๋ฃจ 15๊ฐ๋น์ ๋ด๋ฐฐ๋ฅผ ํผ์ฐ๋ ๊ฒ๋งํผ ๊ฑด๊ฐ์ ํด๋ก์ธ ์ ์๋ค๊ณ ํฉ๋๋ค. ์ด์ฒ๋ผ ์ธ๋ก์์ ๋จ์ํ ๊ฐ์ ์ ๋ถํธํจ์ ๋์ด, ์ ์ฒด์ ๊ฑด๊ฐ๊น์ง ์ํํ๋ ์์ธ์ผ๋ก ์์ฉํฉ๋๋ค.
์ธ๋ก์์ ์ฐ๋ฆฌ ๋ชธ์ ์คํธ๋ ์ค ๋ฐ์์ ์ ๋ฐํฉ๋๋ค. ์ด๋ ์ฝ๋ฅดํฐ์(cortisol) ์์น๋ฅผ ๋์ด๊ณ ์ผ์ฆ(inflammation)์ ์ฆ๊ฐ์ํค๋ฉฐ, ๋ฉด์ญ ์ฒด๊ณ๋ฅผ ์ฝํ์์ผ ์ง๋ณ์ ๋ ์ทจ์ฝํ๊ฒ ๋ง๋ญ๋๋ค. ์ ๊ฒฝ๊ณผํ ์ฐ๊ตฌ์์๋ ์ด๋ฌํ ์ฐ๊ด์ฑ์ ํ์ธํ ์ ์์ต๋๋ค.
ํ ์ฐ๊ตฌ์์๋ ์ฐธ๊ฐ์๋ค์ด ๋ก๋งจํฑ ํํธ๋๋ ์์ ํ ๋ฏ์ ์ฌ๋์ ์ฌ์ง์ ๋ณด๋ฉด์ ์ฝํ ์ ๊ธฐ ์ถฉ๊ฒฉ์ ๋ฐ์์ต๋๋ค. ์ฐ๊ตฌ ๊ฒฐ๊ณผ, ๋ก๋งจํฑ ํํธ๋์ ์ฌ์ง์ ๋ณธ ์ฌ๋๋ค์ ๋๋ ค์์ ๋ ๋๋ผ๊ณ , ํต์ฆ๊ณผ ๊ด๋ จ๋ ๋ ํ๋์ด ๋ ๋ฎ์ ๊ฒ์ผ๋ก ๋ํ๋ฌ์ต๋๋ค. ์ด๋ ํ์ธ๊ณผ์ ์ฐ๊ฒฐ์ฑ์ ๋๋ผ๋ ์ ๋์ ๋ฐ๋ผ ํต์ฆ์ ๋ํ ์ธ์์ด ๋ฌ๋ผ์ง ์ ์์์ ์์ฌํฉ๋๋ค. ์ฆ, ๊ฑด๊ฐ์ ๋จ์ํ ์ ์ฒด์ , ์ ์ ์ ์ธ ๊ฒ์ ๋์ด ์ฌํ์ ์ธ ์ธก๋ฉด๊น์ง ํฌ๊ดํ๋ค๋ ๊ฒ์ ์๋ฏธํฉ๋๋ค. ํ๋ถํ ์ฐ์ ์ ๊ฐ์ง๊ณ ์ง์ญ์ฌํ์ ์์๊ฐ์ ๋๋ผ๋ฉฐ ๊ฐ๋ ฅํ ์ง์ง ์ฒด๊ณ๋ฅผ ๊ฐ๋ ๊ฒ์ ์ธ์์ ๊ฒฝํํ๋ ๋ฐฉ์์ ์์ ํ ๋ณํ์ํฌ ์ ์์ต๋๋ค.
โ์ฐ์ ๋ถํฉโ ์๋์ ๋๋
์ค๋๋ ์ฐ๋ฆฌ๋ โ์ฐ์ ๋ถํฉ(friendship recession)โ์ด๋ผ๋ ์ฉ์ด๊ฐ ๋ฑ์ฅํ ์ ๋๋ก ์น๋ฐํ ๊ด๊ณ๊ฐ ์ค์ด๋๋ ์๋๋ฅผ ์ด๊ณ ์์ต๋๋ค. ๋ฏธ๊ตญ๊ธฐ์ ์ฐ๊ตฌ์(American Enterprise Institute)์ ํ์ ๋๋์ผ ์ฝ์ค(Daniel Cox)๋ ์ด ์ฉ์ด๋ฅผ ์ฌ์ฉํด ์๊ธฐ ์ํฉ์์ ๊ธฐ๋ ์ฌ๋์ด ์ ๊ฑฐ๋, ์ผ์ ์์ ์นํ ์น๊ตฌ๊ฐ ๋ถ์กฑํ ์ฌ๋๋ค์ ์ฆ๊ฐ ํ์์ ์ค๋ช ํฉ๋๋ค.
์ถฉ๊ฒฉ์ ์ธ ํต๊ณ๋ ์ด ํ์์ ์ฌ๊ฐ์ฑ์ ๋ณด์ฌ์ค๋๋ค. 1990๋ ๋์๋ ์ ์ ๋จ์ฑ ์ค 3%๋ง์ด ์นํ ์น๊ตฌ๊ฐ ์๋ค๊ณ ๋ตํ์ง๋ง, ์ค๋๋ ์๋ ๊ทธ ์์น๊ฐ 15%์ ๋ฌํฉ๋๋ค. ํฅ๋ฏธ๋กญ๊ฒ๋ ํฌ๋ฐ๋ฏน ์ดํ์๋ ์ฌ์ฑ๋ค์ด ๊ฐ์ฅ ํฐ ์ํฅ์ ๋ฐ์, ์ ๋ฐ ์ด์์ ์ฌ์ฑ๋ค์ด ์ต์ํ ๋ช๋ช ์น๊ตฌ๋ค๊ณผ ์ฐ๋ฝ์ด ๋๊ฒผ๋ค๊ณ ๋ตํ์ต๋๋ค.
์ด๋ ๊ฐ์กฑ, ๊ฒฐํผ, ์ข ๊ต, ๋ ธ๋ ์์ฅ ๋ฑ ์ ํต์ ์ธ ์ฌํ ์ ๋๋ค์ด ์ฝํ๋๋ฉด์ ๋ฐ์ํ๋ ํ์์ ๋๋ค. ์ฌ๋๋ค์ด ๊ฒฐํผ์ ๋ฆ๊ฒ ํ๊ฑฐ๋ ์์ ํ์ง ์๊ณ , ์ข ๊ต ํ๋์ด ์ค์ด๋๋ ๋ฑ ์ฌํ ๊ตฌ์กฐ๊ฐ ๋ณํํ๋ฉด์, ์ด๋ฌํ ์ ๋ ๋ฐ์์ ์ฌํ์ ๊ด๊ณ์ ์ฐ๊ฒฐ์ ์ฐพ์ผ๋ ค๋ ํ์์ฑ์ด ๋์ฑ ์ปค์ก์ต๋๋ค. ๋ฐ๋ก ์ด ์ง์ ์์ ์น๊ตฌ์ ์ญํ ์ด ์์ฒญ๋๊ฒ ์ค์ํด์ง๋๋ค. ์น๊ตฌ๊ฐ ์๋ค๋ ๊ฒ์ ๋จ์ํ ๊ฒฝ์ ์ , ์ฌํ์ ์ผ๋ก ๊ณ ๋ฆฝ๋ ์ ์๋ค๋ ๊ฒ์ ๋์ด, ์ฌํ์ ์ ๋ฐํ๋ฉฐ ์ด๋ ์ ์ฒด์ , ์ ์์ ๊ฑด๊ฐ ๋ชจ๋์ ํด๋กญ๋ค๋ ๊ฒ์ด ๋ฐํ์ก์ต๋๋ค.
์ด๋ฌํ ์ถ์ธ๊ฐ ๊ณ์๋๋ค๋ฉด, ์น๊ตฌ ์์ด ๊ณ ๋ฆฝ๋๊ณ ๊ฑด๊ฐ๊น์ง ๋๋น ์ง๋ โํํธํ๋ ๊ฐ์ธ๋ค(atomized individuals)โ์ ์ธ๊ณ๋ผ๋ ๋์คํ ํผ์์ ๋ฏธ๋๊ฐ ๋๋ํ ์๋ ์์ต๋๋ค. ๋ฐ๋ผ์ ์ฐ๋ฆฌ๋ ์ธ๊ฐ์ ๋ฒ์์ ์ํด ์ฐ์ ์ด ์ผ๋ง๋ ์ค์ํ์ง ์ธ์ํด์ผ ํฉ๋๋ค.
์น๊ตฌ๋ฅผ ์ฌ๊ท๋ ๋ฐ ํ์ํ ๊ฐ์ฅ ์ค์ํ ๋จ๊ณ ์ค ํ๋๋ ์น๊ตฌ๋ฅผ ์ฌ๊ท๊ณ ์ถ๋ค๋ ๊ฒ์ ์ธ์ ํ๊ณ , ๊ธฐ๊บผ์ด ์ทจ์ฝํด์ง๋ ๊ฒ์ ๋๋ค. ๋์ด๊ฐ ๋ค์๋ก ์น๊ตฌ๊ฐ ๋ถ์กฑํ๋ค๋ ๊ฒ์ ๋ํ ์์น์ฌ์ ๋๋ผ๋ ๊ฒฝ์ฐ๊ฐ ๋ง์๋ฐ, โ๋์๊ฒ ์น๊ตฌ๊ฐ ํ์ํ๋คโ๊ณ ๋งํ๋ ๊ฒ์ ์ธ๊ฐ์ด ๋ด๋ฑ์ ์ ์๋ ๊ฐ์ฅ ์ด๋ ค์ด ๋ฌธ์ฅ ์ค ํ๋์ผ ์ ์์ต๋๋ค.
์ธ๋ก์์ ์ฑ์ฅ์ ๊ธฐํ๋ก ์ฌํด์ํ๊ธฐ
์ง๋ ๋ช ๋ ๊ฐ ์ธ๋ก์์ ๋ถ์ ์ ์ธ ๊ฒฐ๊ณผ์ ๋ํ ๊ด์ฌ์ด ์ฆ๊ฐํ์ต๋๋ค. ํ์ง๋ง ๋ฏธ๋์ด๋ ์ธ๋ก์์ ๊ธ์ ์ ์ผ๋ก ๋ฌ์ฌํ๋ ๊ฒฝ์ฐ๋ณด๋ค ๋ถ์ ์ ์ผ๋ก ๋ฌ์ฌํ๋ ๊ฒฝ์ฐ๊ฐ ์ฝ 10๋ฐฐ ๋ ๋ง์ต๋๋ค. ์ด๋ฌํ ํธํฅ๋ ์๊ฐ์ ์ธ๋ก์์ ๋ํ ์ฐ๋ฆฌ์ ์ธ์์ ๋์ฑ ์๊ณก์ํต๋๋ค.
ํ์ง๋ง ์ฐ๊ตฌ์ ๋ฐ๋ฅด๋ฉด, ํผ์ ์๋ ์๊ฐ์ ๊ธ์ ์ ์ผ๋ก ์๊ฐํ๋ ์ฌ๋๋ค์ ์ค์ ๋ก ํผ์ ์์ ๋ ๋ ์ข์ ๊ธฐ๋ถ์ ๋๋๋ค๊ณ ํฉ๋๋ค. ์ด ์ฐ๊ตฌ ๊ฒฐ๊ณผ๋ ์ฐ๋ฆฌ๊ฐ ์ธ๋ก์์ ๋ํ ๋ฏฟ์์ ๋ฐ๊ฟ ์ ์๋ ๊ธฐํ๋ฅผ ์ ๊ณตํฉ๋๋ค. ์ธ๋ก์์ โ๋์๊ฒ ๋ญ๊ฐ ์๋ชป๋ ๊ฒโ์ผ๋ก ์ฌ๊ธฐ๊ธฐ๋ณด๋ค๋, ์ฐฝ์๋ ฅ์ ๋ฐํํ๊ณ ์์ ์ ์๊ฐ๊ณผ ์ฆ๊ฒ๊ฒ ์๊ฐ์ ๋ณด๋ผ ์ ์๋ ๊ธฐํ๋ก ์ฌ๊ตฌ์ฑํ ์ ์์ต๋๋ค. ์ธ๋ก์์ ๋ํด ๋ ๋ง์ด ์ด์ผ๊ธฐํ๊ณ ๋ ผ์ํ ์๋ก, ์ด๋ ๋ ๋๋ฆฌ ๋ฌธํ ์์ ์ค๋ฉฐ๋ค์ด ์ฐ๋ฆฌ์ ์ธ์์ ๋ณํ์ํฌ ๊ฒ์ ๋๋ค.
๋๋ง์ ๊ฑด๊ฐํ ์ฌํ์ํ ์ฐพ๊ธฐ
๋ค๋ฅธ ์ฌ๋๋ค๊ณผ์ ์ฐ๊ฒฐ๋งํผ์ด๋ ์์ ๊ณผ์ ์ฐ๊ฒฐ๋ ์ค์ํฉ๋๋ค. ๋น์ ์๊ฒ ์ต์ ์ ์ฌํ์ ๊ฑด๊ฐ์ ๋ค๋ฅธ ์ฌ๋๊ณผ๋ ๋ค๋ฅผ ์ ์์ต๋๋ค. ์ผ๋ง๋ ๋ง์ ์ฌํ ํ๋์ ์ฆ๊ธฐ๋์ง, ์ด๋ค ์ข ๋ฅ์ ์ฐ๊ฒฐ์ด ๋น์ ์๊ฒ ๋ง์กฑ๊ฐ์ ์ฃผ๋์ง์ ๋ฐ๋ผ ๋ฌ๋ผ์ง๊ธฐ ๋๋ฌธ์ ๋๋ค.
์ธ๋ก์์ ๋จ์ง ํ ๊ฑธ์ ๋ ๋์๊ฐ ์์ ์ ์ฌํ์ ๊ฑด๊ฐ์ ํ์ํ ์ ์๋ ๊ธฐํ์ ๋๋ค. ๋น์ ์ ์ฌํ์ํ์ด๋ผ๋ ์คํ์ค์์ ๊ณผํ์๊ฐ ๋์ด๋ณด์ธ์. ์ฐ๋ฆฌ๋ ๋น์ ์ด ์ด๊ณ ์๋ โ์ธ์์ ์์(Perception Box)โ๋ฅผ ๋ํ๊ณ , ๋ ์ฐ๊ฒฐ๋์ด ์๋ค๊ณ ๋๋ผ๋ฉฐ, ์ธ๋ก์์ด๋ผ๋ ๊ณ ๋ฆฝ๋ ๊ฒฝํ์ ๊ทน๋ณตํ๋๋ก ๋์ธ ์ ์์ต๋๋ค.
๊ฒฐ๋ก : ์ธ๋ก์์ ์ง์ํ๊ณ ์ฐ๊ฒฐ์ ๋ชจ์ํ๋ค
์ธ๋ก์์ ๋จ์ํ ๋ถํธํ ๊ฐ์ ์ ๋์ด, ์ฐ๋ฆฌ์ ์ ์ฒด์ , ์ ์ ์ ๊ฑด๊ฐ์ ์ฌ๋ํ ์ํฅ์ ๋ฏธ์น๋ฉฐ ์ฌํ ๊ตฌ์กฐ์ ๋ณํ์ ๋ง๋ฌผ๋ ค ๋์ฑ ์ฌํ๋๊ณ ์๋ ํ๋ ์ฌํ์ ์ค์ํ ๋ฌธ์ ์ ๋๋ค. ๊ทธ๋ฌ๋ ์ ๋ฌธ๊ฐ๋ ์ธ๋ก์์ ๋ถ์ ์ ์ธ ๋์ธ์ผ๋ก๋ง ๋ณผ ๊ฒ์ด ์๋๋ผ, ์ฐ๋ฆฌ์๊ฒ ๋ฌด์์ด ํ์ํ์ง ์๋ ค์ฃผ๋ ๊ท์คํ ์ ํธ๋ก ์ดํดํด์ผ ํ๋ค๊ณ ๊ฐ์กฐํฉ๋๋ค.
โ์ฐ์ ๋ถํฉโ ์๋์ ์น๋ฐํ ๊ด๊ณ๋ฅผ ์ฌ๊ฑดํ๊ณ , ์ธ๋ก์์ ์ฑ์ฅ์ ๊ธฐํ๋ก ์ฌํด์ํ๋ฉฐ, ์์ ๊ณผ ํ์ธ๊ณผ์ ๊ฑด๊ฐํ ์ฐ๊ฒฐ์ ๋ชจ์ํ๋ ๊ฒ์ ์ฐ๋ฆฌ ๋ชจ๋์ ๊ณผ์ ์ ๋๋ค. ์ธ๋ก์์ ์ง์ํ๊ณ , ๊ทธ ์๋ฏธ๋ฅผ ํ์ํ๋ฉฐ, ์ฉ๊ธฐ ์๊ฒ ์ฐ๊ฒฐ์ ์ถ๊ตฌํ ๋ ์ฐ๋ฆฌ๋ ๋์ฑ ํ์๋กญ๊ณ ๊ฑด๊ฐํ ์ถ์ ๋ง๋ค์ด๊ฐ ์ ์์ ๊ฒ์ ๋๋ค.
โAI Automated Everything. Why Is There More Work?โ โ Every ๊ธฐ๋ฐ ๊ธฐ์ฌ ์๋ณธ ์์ ๋ณด๊ธฐ
AI, ๋ชจ๋ ๊ฒ์ ์๋ํํ๋๋ฐ ์ ์ผ์ ๋ ๋ง์์ง๊น? โ โEveryโ ์ฑ๋์ ์ญ์ค์ ํต์ฐฐ
AI ์๋์ ์ญ์ค: ์๋ํ๊ฐ ์ผ์๋ฆฌ๋ฅผ ๋๋ฆฐ๋ค?
์ธ๊ณต์ง๋ฅ(AI)์ด ์ฐ๋ฆฌ ์ถ์ ๋ชจ๋ ์์ญ์ ์๋ํํ๊ณ ์๋ค๋ ์์์ด ์ฐ์ผ ๋ค๋ ค์ต๋๋ค. ๋ง์ ์ฌ๋์ด AI๊ฐ ์ผ์๋ฆฌ๋ฅผ ๋์ฒดํ๊ณ , ํนํ ์ง์ ์ฅ๋ฒฝ์ด ๋ฎ์ ํ์ดํธ์นผ๋ผ ์ง์ ๊ตฐ์ ์๋ฉธ์ํฌ ๊ฒ์ด๋ผ๋ ๋ง์ฐํ ๋๋ ค์์ ๊ฐ์ง๊ณ ์์ต๋๋ค. ์ธ๊ณ์ ์ธ ํค์งํ๋ ์ํ๋ธ(Citadel)์ ์ผ ๊ทธ๋ฆฌํ(Ken Griffin) ๊ฐ์ ๊ฑฐ๋ฌผ์กฐ์ฐจ โ์ด๊ฒ์ ์ค๊ฐ ๊ด๋ฆฌ์ง์ด ์๋๋ผ, ๋งค์ฐ ๋์ ์๋ จ๋๋ฅผ ์๊ตฌํ๋ ์ผ๋ค์ด ์์ด์ ํธํ AI(Agentic AI)์ ์ํด ์๋ํ๋๋ ๋ชจ์ตโ์ด๋ผ๊ณ ๊ฒฝ๊ณ ํ ์ ๋์ ๋๋ค. ๊ทธ๋ฌ๋ AI์ ์์ด์ ํธ ์์คํ ์ ์ต์ ์ ์์ ํ์ฉํ๋ ์ด๋ค์ ์ค์ ๊ฒฝํ์ ์ด๋ฌํ ํต๋ ๊ณผ ์ฌ๋ญ ๋ค๋ฆ ๋๋ค.
์ฝํ ์ธ ๊ธฐ์ โEveryโ์ CEO์ธ ํ๋ฆฝ ๋งํด(Phillip Martin)์ ์์ ์ ์ฑ๋์์ ์ด ์ญ์ค์ ์ธ ํ์์ ์ฌ์ธต์ ์ผ๋ก ๋ถ์ํฉ๋๋ค. ๊ทธ์ ์ฃผ์ฅ์ AI๊ฐ ์๋ จ๋ ์ธ๊ฐ์ ์ ๋ฌธ์ฑ์ ๋์ค์๊ฒ ์ ๋ ดํ๊ฒ ์ ๊ณตํ์ง๋ง, ๋๋๊ฒ๋ ์ด๋ ์ ๋ฌธ๊ฐ์ ๋ํ ์์๋ฅผ ์คํ๋ ค ์ฆ๊ฐ์ํจ๋ค๋ ๊ฒ์ ๋๋ค. โEveryโ๋ 25๋ช ์ ์ง์์ผ๋ก ์ด๋ฃจ์ด์ง ์์ ํ์ฌ์ง๋ง, AI ๊ธฐ์ ์ ์ ๋ฌด์ ์ต๋ํ ์ ๋ชฉ์ํค๋ฉฐ ๋ฏธ๋์ ์ ๋ฌด ํ๊ฒฝ์ ๋ฏธ๋ฆฌ ์ฟ๋ณผ ์ ์๋ โ๋ง์ดํฌ๋ก์ฝ์ฆ(microcosm)โ ์ญํ ์ ํ๊ณ ์์ต๋๋ค. ๊ทธ๋ ๋ค๋ฉด ์๋ํ์ ์๋์ ์ ์ธ๊ฐ์ ์ผ์ ๋ ๋ง์์ง๋ ๊ฑธ๊น์?
โEveryโ์ AI ํ์ฉ๋ฒ: 25์ธ ๊ธฐ์ ์ ๋ง์ดํฌ๋ก์ฝ์ฆ
โEveryโ๋ 2022๋ ๋ถํฐ AI์ ์์ด์ ํธ๊ฐ ์ ๋ฌด ๋ฐฉ์์ ์ด๋ป๊ฒ ๋ณํ์ํฌ์ง์ ๋ํด ๊พธ์คํ ํ๊ตฌํด์์ต๋๋ค. ํนํ ์ฑGPT(ChatGPT)๊ฐ ๋ฑ์ฅํ๊ธฐ ์ ์ธ 3๋ ์ ๋ถํฐ ํ๋ฆฝ ๋งํด์ AI๊ฐ โ๋ฐฐ๋ถ ๊ฒฝ์ (allocation economy)โ๋ฅผ ์ด๋ ๊ฒ์ด๋ฉฐ, ์ธ๊ฐ ๊ด๋ฆฌ์๊ฐ ์ ๋ฌด๋ฅผ ์์ํ๊ณ , ์ธ๋ถํํ๊ณ , ๋ฏธ์ธ ๊ด๋ฆฌํ๋ ๋ฐฉ์์ฒ๋ผ AI์ ํ์ ํ๊ฒ ๋ ๊ฒ์ด๋ผ๊ณ ์์ธกํ์ต๋๋ค. ํ์ฌ โEveryโ๋ ํฌ๊ฒ ๋ ๊ฐ์ง ๋ฐฉ์์ผ๋ก AI ์์ด์ ํธ๋ฅผ ํ์ฉํ๊ณ ์๋๋ฐ, ์ด ์ค ํ๋๋ ์์ธก ๊ฐ๋ฅํ๊ณ ๋ค๋ฅธ ํ๋๋ ํจ์ฌ ์๋กญ๊ณ ์ค์ํฉ๋๋ค.
์์๋ ํ์ฉ๋ฒ: ์์ด์ ํธ์๊ฒ ์ ๋ฌด ์์
๊ฐ์ฅ ์ผ๋ฐ์ ์ธ ํํ๋ ์ฌ๋(Slack)๊ณผ ๊ฐ์ ํ์ ๋๊ตฌ์์ AI ๋ด์๊ฒ ์ง์ ์ ๋ฌด๋ฅผ ์์ํ๋ ๋ฐฉ์์ ๋๋ค. โEveryโ๋ โํด๋ก๋(Claudie)โ, โ์ค๋(Andy)โ, โ๋น ํฐ(Victor)โ ๋ฑ ์์ฒด ๊ฐ๋ฐํ๊ฑฐ๋ ์ธ๋ถ ์๋ฃจ์ ์ ํ์ฉํ๋ ๋ค์ํ AI ์์ด์ ํธ๋ค์ ์ฌ์ฉํฉ๋๋ค.
- ํด๋ก๋(Claudie): โEveryโ์ ์ปจ์คํ ์ฌ์ ์ ๋ฐ์ ์ด์ํฉ๋๋ค. ๋ง์ผํ ๋ฐ ๋ธ๋๋ ๋ฆฌ์์น๋ถํฐ ๊ณ ๊ฐ์ ์ํ ์ปจ์คํ ์ ์์ ์์ฑ, ์ฌ์ง์ด๋ ๊ณ ๊ฐ์๊ฒ ์ ๋ฌํ ๋ฐํฌ(deck)์ ์ด์์ ๋ง๋๋ ์ผ๊น์ง ๋ค์ํ๊ฒ ์ฒ๋ฆฌํฉ๋๋ค.
- ์ค๋(Andy): ํธ์ง ๋ณด์กฐ ์์ด์ ํธ ์ญํ ์ ์ํํฉ๋๋ค.
- ๋น ํฐ(Victor): ์ด ์ธ์๋ ๋ค์ํ ์ ๋ฌด์ ํ์ฉ๋๋ ์ธ๋ถ ์์ด์ ํธ ์๋ฃจ์ ์ ๋๋ค.
์ด๋ฌํ ์์ด์ ํธ๋ค์ ์ ํ๋ธ ์ธ๋ค์ผ A/B ํ ์คํธ ๊ธฐ๋ก๊ณผ ๊ฐ์ ์ธ๋ถ์ ์ธ ์์ ๋ถํฐ ๊ณ ๊ฐ๊ณผ์ ์ปค๋ฎค๋์ผ์ด์ ์ด์ ์์ฑ๊น์ง, ํญ๋์ ์ ๋ฌด๋ฅผ ์ํํ๋ฉฐ ์ง์๋ค์๊ฒ ์์ฒญ๋ ๋ ๋ฒ๋ฆฌ์ง(leverage)๋ฅผ ์ ๊ณตํฉ๋๋ค.
์๋กญ๊ณ ์ค์ํ ํ์ฉ๋ฒ: ์์ด์ ํธ ์ค์ผ์คํธ๋ ์ด์ ์์คํ
๋์ฑ ํฅ๋ฏธ๋กญ๊ณ ์ค์ํ ๋ณํ๋ ์์ด์ ํธ ์ค์ผ์คํธ๋ ์ด์ ์ํํธ์จ์ด(Agent Orchestration Software)์ ๋ฑ์ฅ์ ๋๋ค. ์ฝ๋ฑ์ค(Codex), ์ฝ์ํฌ(Co-work), ํด๋ก๋ ์ฝ๋(Claude Code)์ ๊ฐ์ ๋๊ตฌ๋ค์ด ์ด์ ๋ ์ฌ์ค์ ์ ๋ฌด์ ์ด์ ์ฒด์ (Operating System) ์ญํ ์ ํ๊ณ ์์ต๋๋ค. ์ด๋ ๋จ์ํ ์ฑ๋ด๊ณผ์ ๋ํ๋ฅผ ๋์ด, ์ฌ์ฉ์์ ์ปดํจํฐ์ ์์ฃผํ๋ฉฐ ์์คํ ์ ๋ฐ์ ์ ๊ทผํ๊ณ ๋ธ๋ผ์ฐ์ ๊น์ง ๋ด์ฅํ ์์ด์ ํธ๋ฅผ ์๋ฏธํฉ๋๋ค.
ํ๋ฆฝ ๋งํด์ ์์ ์ ๋ชจ๋ ์ ๋ฌด๊ฐ ๊ฑฐ์ ์ฝ๋ฑ์ค๋ ์ฝ์ํฌ ์์์ ์ด๋ฃจ์ด์ง๋ค๊ณ ์ค๋ช ํฉ๋๋ค. ์ด๋ฉ์ผ ์์ฑ, ์ฝ๋ฉ, ์ฌ์ง์ด ์ฌ์ ์ ๋ถ๊ธฐ๋ณ ์์ต๊ณ์ฐ์(P&L) ๋ถ์๊น์ง ๋ชจ๋ ์ด ์์ด์ ํธ ์์คํ ์์์ ์ฒ๋ฆฌํฉ๋๋ค. ์ฌ์ฉ์๋ ์ฌ๋ฌ ์์ด์ ํธ๊ฐ ๋์์ ์๋ํ๋ ๊ฒ์ ์ง์ผ๋ณด๋ฉฐ, ์์ด์ ํธ์ ๊ธด๋ฐํ๊ฒ ์ํธ์์ฉํฉ๋๋ค. ์ด๋ ์ง์ ํ AI-์ธ๊ฐ ํ์ ์์คํ ์ ํํ๋ฅผ ๋ ๋ฉฐ, ์ ๋ฌด์ ํจ์จ์ฑ์ ๊ทน๋ํํฉ๋๋ค.
ํต์ฌ ํต์ฐฐ: ์ธ๊ฐ์ ๊ฐ์ ์ ํ์์ ์ด๋ค
์ด ๋ ๊ฐ์ง ์์ด์ ํธ ํ์ฉ ๋ฐฉ์ ๋ชจ๋์ ๊ณตํต์ ์ผ๋ก ํด๋น๋๋ ์ค์ํ ์ฌ์ค์ ์ธ๊ฐ์ ๊ฐ์ ์์ด๋ ์ ๋๋ก ์๋ํ์ง ์๋๋ค๋ ๊ฒ์ ๋๋ค. ์๋ฅผ ๋ค์ด, ์ปจ์คํ ์ฌ์ ์ ์ด์ํ๋ ํด๋ก๋๋ โ๋์ดํ ์(Nitesh)โ๋ผ๋ ์ ์ AI ์์ง๋์ด์ ์ง์์ ์ธ ๋์ ์์ด๋ ์ ๊ธฐ๋ฅ์ ํ๊ธฐ ์ด๋ ต์ต๋๋ค. ๋์ดํ ์๋ ํด๋ก๋๊ฐ ์ ๋๋ก ์๋ํ์ง ์๋ ๋ถ๋ถ์ ์ฐพ์๋ด๊ณ , ๋์์์ด ๊ฐ์ ํ๋ฉฐ ํด๋ก๋์ ์ฑ๋ฅ์ ์ต์ ํํ๋ ์ญํ ์ ํฉ๋๋ค.
์ด๋ ์์ด์ ํธ ์์คํ ์ด ํ ๋ฒ ์ค์ ๋๋ฉด ์์์ ๋ชจ๋ ๊ฒ์ ์ฒ๋ฆฌํ ๊ฒ์ด๋ผ๋ ์ผ๋ฐ์ ์ธ ์คํด์๋ ๊ฑฐ๋ฆฌ๊ฐ ๋ฉ๋๋ค. AI๊ฐ ์ธ๊ฐ์ ๋ ๋ฒ๋ฆฌ์ง๋ฅผ ์์ฒญ๋๊ฒ ์ฆํญ์ํค๋ ๊ฒ์ ๋ง์ง๋ง, ์ธ๊ฐ์ด ์ค๋ฅ๋ฅผ ๊ต์ ํ๊ณ ์ฌ๋ฐ๋ฅธ ๋ฐฉํฅ์ผ๋ก ์ด๋์ด์ฃผ๋ โ์ธ๊ฐ์ ์๊ธธโ์ด ์๋ค๋ฉด ๊ทธ ํจ๊ณผ๋ ๊ธ๊ฒฉํ ๋จ์ด์ง๋๋ค. ํ๋ฆฝ ๋งํด์ โ์ธ๊ฐ๊ณผ์ ํ์ ์ด์ผ๋ง๋ก ์์ด์ ํธ ์์คํ ์ ๊ฐ๋ ฅํ๊ฒ ๋ง๋๋ ํต์ฌโ์ด๋ผ๊ณ ๊ฐ์กฐํฉ๋๋ค.
AI ์๋ํ๊ฐ ์ธ๊ฐ์ ์ผ์ ๋๋ฆฌ๋ 5๋จ๊ณ ๋ ผ๋ฆฌ
๊ทธ๋ ๋ค๋ฉด AI๊ฐ โ์ด์ ์ ์ธ๊ฐ ์ญ๋โ์ ์ ๋ ดํ๊ฒ ๋ง๋ค์์์๋ ๋ถ๊ตฌํ๊ณ , ์ ์ธ๊ฐ์ ์ผ์ด ๋ ๋ง์์ง๊ณ ์ ๋ฌธ๊ฐ์ ๋ํ ์์๊ฐ ์ฆ๊ฐํ๋ ๊ฑธ๊น์? ํ๋ฆฝ ๋งํด์ ๋ค์์ 5๋จ๊ณ ๋ ผ๋ฆฌ๋ฅผ ์ ์ํฉ๋๋ค.
1๋จ๊ณ: AI, ๊ณผ๊ฑฐ์ ์ญ๋์ ์ ๋ ดํ๊ฒ ๋ง๋ ๋ค
AI, ํนํ ๋๊ท๋ชจ ์ธ์ด ๋ชจ๋ธ(LLM)์ ๋ฐฉ๋ํ ๋ฐ์ดํฐ๋ฅผ ๊ธฐ๋ฐ์ผ๋ก ํ๋ จ๋ฉ๋๋ค. ์ด ๋ฐ์ดํฐ๋ ๊ณง ์ธ๊ฐ ์ญ๋์ โ๊ฐ์์ ์ธ ์์ฌ๋ฌผ(visible residue of competence)โ์ ๋๋ค. ์๋ฅผ ๋ค์ด, ๊ณผ๊ฑฐ์๋ ์ฝ๋๋ฒ ์ด์ค์ ์ฝ๋๋ฅผ ์ ์ถํ๊ธฐ ์ํ โํ ๋ฆฌํ์คํธ(Pull Request)โ๋ฅผ ์์ฑํ๋ ๊ฒ์ด ํฌ๊ทํ๊ณ ๊ฐ๋น์ผ ์ ๋ฌธ ๊ธฐ์ ์ด์์ต๋๋ค. ํ์ง๋ง ์ด์ ํด๋ก๋ ์ฝ๋๋ ์ฝ๋ฑ์ค์ ๊ฐ์ ๋๊ตฌ๋ฅผ ์ฌ์ฉํ๋ฉด ๋๊ตฌ๋ ์ฝ๊ฒ ํ ๋ฆฌํ์คํธ๋ฅผ ์์ฑํ ์ ์์ต๋๋ค. AI๋ ์ด๋ ๊ฒ ๊ณผ๊ฑฐ์๋ ์ ๋ฌธ๊ฐ์ ์์ญ์ด์๋ ํน์ ์ญ๋์ ๋์ค์๊ฒ ์ ๋ ดํ๊ณ ์ฝ๊ฒ ์ ๊ณตํฉ๋๋ค.
2๋จ๊ณ: ์ ๋ ดํด์ง ์ญ๋, ํญ๋ฐ์ ์ธ ํ์ฐ์ ์ด๋๋ค
์ด๋ค ์ญ๋์ด๋ ์ ๋ ดํ๊ณ ์ฝ๊ฒ ์ ๊ทผ ๊ฐ๋ฅํด์ง๋ฉด, ๊ทธ ํ์ฉ๋๋ ํญ๋ฐ์ ์ผ๋ก ์ฆ๊ฐํฉ๋๋ค. ๋๊ตฌ๋ ํ ๋ฆฌํ์คํธ๋ฅผ ์์ฑํ๊ณ , ์ด๋ฏธ์ง๋ฅผ ๋ง๋ค๊ณ , ์๋ง์ ํธ์์ ์ฌ๋ฆด ์ ์๊ฒ ๋ฉ๋๋ค. โEveryโ ๋ด๋ถ์์๋ ์ด๋ฌํ ํ์์ด ๋น๋ฒํ๊ฒ ๋ชฉ๊ฒฉ๋ฉ๋๋ค. ์ด์ํ ์ง์์ด๋ ๊ณ ๊ฐ ์๋น์ค ๋ด๋น์๊ฐ ํ ๋ฆฌํ์คํธ๋ฅผ ์ ์ถํ๊ณ , ๋ง์ผํฐ๊ฐ ์ ํ๋ธ ์ธ๋ค์ผ์ ์ง์ ๋ง๋ญ๋๋ค. โ์ ๋ ดํ ์ญ๋โ์ ๋ํ ์์๋ ๋งค์ฐ ๋๊ธฐ ๋๋ฌธ์, ์ด๋ ๋ชจ๋ ๊ณณ์ผ๋ก ๋น ๋ฅด๊ฒ ํ์ฐ๋ฉ๋๋ค.
3๋จ๊ณ: โ์ฌ๋กญ(Slop)โ์ ํ์์ ์๋ก์ด ์์ ์ ๋ฑ์ฅ
๋ฌธ์ ๋ ์ด๋ฌํ โ์ ๋ ดํ ์ญ๋โ์ ํ์ฉํ์ฌ ๋ง์ ์ฌ๋์ด ์ผ์ ์์ํ์ง๋ง, ๋๋ถ๋ถ์ ๊ฒฝ์ฐ โ๊ธฐ๋ณธ๊ฐ(default)โ ์ค์ ์ ๊ทธ๋๋ก ์ฌ์ฉํ๊ฑฐ๋ ๊ฐ์ฅ ์ฌ์ด ํ๋กฌํํธ(prompt)๋ง์ ํ์ฉํ๋ค๋ ์ ์ ๋๋ค. ๊ทธ ๊ฒฐ๊ณผ๋ฌผ์ โ์์ ํ ์๋ง์ง์ฐฝโ๋ถํฐ โ๊ด์ฐฎ์ ์์์ โ๊น์ง ๋ค์ํ์ง๋ง, ์ ๋ฌธ์ฑ์ด ๋ถ์กฑํ ์ํ์์ ๋๊ตฌ๋ง ์ฌ์ฉํ๋ ๊ฒฝ์ฐ ๋๊ฐ โ์ฌ๋กญ(Slop)โ์ด๋ผ๊ณ ๋ถ๋ฆฌ๋, ๊ฐ์น ์๋ ๊ฒฐ๊ณผ๋ฌผ์ ์์ฐํ๊ฒ ๋ฉ๋๋ค.
ํ๋ฆฝ ๋งํด์ โ์ฌ๋กญโ์ด ๋จ์ํ ํน์ ๋ฌธ์ฅ ๊ตฌ์กฐ๋ ์๊ฐ์ ํน์ง์ ์๋ฏธํ๋ ๊ฒ์ด ์๋๋ผ๊ณ ์ค๋ช ํฉ๋๋ค. ์ด๋ AI ๋๊ตฌ๊ฐ ๋ค์ํ ์ํฉ์์ ๊ธฐ๋ณธ์ ์ผ๋ก ์ ์ฌํ ์ ํ์ ๊ฒฐ๊ณผ๋ฌผ์ ์์ฐํ ๋, ์ธ๊ฐ์ด โ์ด๋๊ฐ ๋น์ทํ๊ณ ๋ถ์กฑํ๋คโ๊ณ ๋๋ผ๋ ๊ฐ๊ฐ, ์ฆ โ์ฅ ๋์๋ฅผ ๋งก๋(smell a rat)โ ๊ฒ๊ณผ ๊ฐ์ ๋ณธ๋ฅ์ ์ธ ์ธ์์ ๋๋ค. ๋ชจ๋๊ฐ ๋์ผํ ๋๊ตฌ๋ฅผ ์ฌ์ฉํ์ฌ ๋น์ทํ ๊ฒฐ๊ณผ๋ฌผ์ ๋ง๋ค์ด๋ด๊ธฐ ์์ํ๋ฉด, ์์ฅ์ ์ด โ์ฌ๋กญโ์ผ๋ก ๋์ณ๋๊ฒ ๋ฉ๋๋ค.
4๋จ๊ณ: โ์ฐจ๋ณ์ฑโ์ ๋ํ ์์ ์ฆ๊ฐ, ์ ๋ฌธ๊ฐ์ ์ญํ ๋ถ์
์์ฅ์ด โ์ฌ๋กญโ์ผ๋ก ๊ฐ๋ ์ฐจ๊ฒ ๋๋ฉด, ์์ฐ์ค๋ฝ๊ฒ โ์ฐจ๋ณ์ฑ(difference)โ์ ๋ํ ์์๊ฐ ๊ธ์ฆํฉ๋๋ค. ๋ชจ๋ ๊ฒ์ด ๋น์ทํด ๋ณด์ผ ๋, ๋ ์ฐฝ์ ์ด๊ณ ๊ณ ํ์ง์ด๋ฉฐ ์ค์ ๋ก ๊ฐ์น ์๋ ๊ฒ์ ๋ง๋ค์ด๋ผ ์ ์๋ ์กด์ฌ๊ฐ ํ์ํด์ง๋๋ค. ๋ฐ๋ก ์ด๋ ์ ๋ฌธ๊ฐ์ ์ญํ ์ด ์ค์ํด์ง๋๋ค.
์๋ฅผ ๋ค์ด, ๋๊ตฌ๋ ํ ๋ฆฌํ์คํธ๋ฅผ ์์ฑํ ์ ์๋ ์ํฉ์์ ์ ๋ฌธ ๊ฐ๋ฐ์๋ โ๋ฌธ์ ๋ ์๊ฒ ์ง๋ง, ์ด๊ฑด ์ด์์ผ ๋ฟโ์ด๋ผ๊ณ ๋งํ๋ฉฐ, ์ด ์ด์์ ์ค์ ํ๋ก๋์ ์ ๋ณํฉํ๊ณ ๋ฐฐํฌํ ์ ์๋ ์์ค์ ์ฝ๋๋ก ํ๋ฐ๊ฟ์ํต๋๋ค. ๋ง์ ์ ๋ฌธ๊ฐ๊ฐ โAI ๋๋ฌธ์ ์ฌ๋กญ์ ์ ๋ฆฌํด์ผ ํ ์ผ์ด ํจ์ฌ ๋ง์์ก๋คโ๊ณ ๋ถํํ๋ ์ด์ ๋ ์ฌ๊ธฐ์ ์์ต๋๋ค. AI๊ฐ ์๋ก์ด ์์ ์ ์์ ๋๋ฆฌ๋ฉด์, ์ด๋ฅผ ๊ฐ์น ์๋ ๊ฒฐ๊ณผ๋ฌผ๋ก ๋ณํํ๋ ์ ๋ฌธ๊ฐ์ ๋ถ๋ด๊ณผ ์ค์์ฑ์ด ๋์์ ์ปค์ง๋ ๊ฒ์ ๋๋ค.
5๋จ๊ณ: ์ ๋ฌธ๊ฐ์ ๋์: ์์คํ ๊ตฌ์ถ๊ณผ ์ญ๋ ํ์ฅ
์ด๋ฌํ ์๋ก์ด ํ๊ฒฝ์์ ์ ๋ฌธ๊ฐ๋ค์ ๋ ๊ฐ์ง ๋ฐฉ์์ผ๋ก ๋์ํฉ๋๋ค.
- ์๋ก์ด ์์ ์ ํก์ํ๊ณ ํ์ฉํ๋ ์์คํ ๊ตฌ์ถ: โEveryโ์ ๊ฒฝ์ฐ, ํ๋ฆฝ ๋งํด์ด ์์ฑํ ์ด์ ํ ๋ฆฌํ์คํธ๋ ๋น์ฅ ์ ์ฉํ์ง ์์ ์ ์์ต๋๋ค. ํ์ง๋ง ์ด์ ํ์ฌ๋ ํด๋ผ์ฐ๋ MD ํ์ผ(Cloud MD file)์ด๋ ์ ์ฅ์ ๊ท์น(repo rules), ๋๋ ํ ๋ฆฌํ์คํธ ์ ์ถ ์๊ธฐ์ ๋ฐฉ๋ฒ์ ๋ํ ์ฌํ์ ๊ณ์ฝ๊ณผ ๊ฐ์ ์์คํ ์ ๊ฐ์ถ๊ณ ์์ต๋๋ค. ์ด๋ฅผ ํตํด โ์ฌ๋กญโ์ผ๋ก ์์๋ ์ ์๋ ์๋ก์ด ์์ ๋ค์ ํก์ํ๊ณ , ์ด๋ฅผ ๋๋ผ์ด ๊ฒฐ๊ณผ๋ฌผ๋ก ์ ํํ๋ ํ๋ก์ธ์ค๋ฅผ ๋ง๋ญ๋๋ค.
- AI๋ฅผ โ๋ฐํโ ์ผ์ ์ด์ ์๋ ๋ถ๊ฐ๋ฅํ๋ ์ผ ์ํ: ์ ๋ฌธ๊ฐ๋ค์ AI๊ฐ ์ ๊ณตํ๋ โ์ด์ ์ ์ญ๋โ์ ์์ ๋ค์ ์ ๋ฌด์ ํตํฉํ์ฌ, ์ด๋ฅผ ์๋ก์ด ๊ฐ๋ฅ์ฑ์ ๋ฐํ์ผ๋ก ์ผ์ต๋๋ค. AI์ ๋์์ผ๋ก ์ด์ ์๋ ํผ์์ ๊ฐ๋นํ๊ธฐ ์ด๋ ค์ ๋ ๋ฐฉ๋ํ ์์ ์ ์ํํ๊ณ , ํจ์ฌ ๋ ๋์ ์์ค์ ๊ฒฐ๊ณผ๋ฌผ์ ์ฐฝ์ถํ ์ ์๊ฒ ๋ฉ๋๋ค. โEveryโ์์๋ ๋ช๋ช ์ง์์ด AI์ ๋์์ ๋ฐ์ ํ ์ฌ๋์ด ์ ์ฒด ์ํํธ์จ์ด ์ ํ ํ๋๋ฅผ ์ด์ํ๋ ๊ฒ์ด ๊ฐ๋ฅํด์ก๋๋ฐ, ์ด๋ ๊ณผ๊ฑฐ์๋ ์์ํ๊ธฐ ์ด๋ ค์ด ์ผ์ด์์ต๋๋ค.
๊ฒฐ๋ก ์ ์ผ๋ก, AI๊ฐ ์์ฐ๋๊ณผ ์ ๊ทผ์ฑ์ ๋์ด๋ฉด์, ์ด๋ฅผ ๊ด๋ฆฌํ๊ณ , ๋ค๋ฌ๊ณ , ์ง์ ์ผ๋ก ๊ฐ์น ์๋ ๊ฒ์ผ๋ก ๋ง๋๋ ์ธ๊ฐ ์ ๋ฌธ๊ฐ์ ์ญํ ์ ๋์ฑ ์ค์ํด์ง๋๋ค.
์ ๋ ผ์ ์ญ์ค๊ณผ AI ์๋์ ์ธ๊ฐ
ํ๋ฆฝ ๋งํด์ ์ด๋ฌํ ํ์์ ๊ณ ๋ ๊ทธ๋ฆฌ์ค์ โ์ ๋ ผ์ ์ญ์ค(Zenoโs Paradox)โ์ ๋น์ ํฉ๋๋ค. ์ํฌ๋ ์ค๊ฐ ๊ฑฐ๋ถ์ด๋ฅผ ์ซ์๊ฐ์ง๋ง, ์ํฌ๋ ์ค๊ฐ ๊ฑฐ๋ถ์ด์์ ๊ฑฐ๋ฆฌ์ ์ ๋ฐ์ ๊ฐ ๋๋ง๋ค ๊ฑฐ๋ถ์ด๋ ์กฐ๊ธ์ฉ ์์ง์ฌ ์ํฌ๋ ์ค๋ ์์ํ ๊ฑฐ๋ถ์ด๋ฅผ ์ก์ ์ ์๋ค๋ ์ญ์ค์ ๋๋ค. AI๊ฐ ์์ฒญ๋ ์๋๋ก ์ธ๊ฐ์ ์ญ๋์ โํก์(hoovering up)โํ๋ฉฐ ์ฐ๋ฆฌ ๋ค๋ฅผ ๋ฐ์ง ์ซ์์ค์ง๋ง, ์ธ๊ฐ์ ๋์์์ด ์๋ก์ด ๊ฐ์น๋ฅผ ์ฐฝ์ถํ๋ฉฐ ์์ ๋๊ฐ๋ค๋ ๊ฒ์ ๋๋ค.
๊ทธ๋ AI๊ฐ ๊ถ๊ทน์ ์ผ๋ก ์ธ๊ฐ์ ์ผ์ ์ ๊ฑฐํ ๊ฒ์ด๋ผ๋ ์๊ฐ์ ๋์ํ์ง ์์ต๋๋ค. ํ๋ฆฝ ๋งํด์ ์ง์ ํ ์ธ๊ณต ์ผ๋ฐ ์ง๋ฅ(AGI)์ โ๊ฒฐ์ฝ ๋์ง ์๊ณ ํญ์ ์๋ํ๋ฉฐ, ๊ฒฝ์ ์ ์ผ๋ก ํ๋นํ๊ณ , ๋์์์ด ์๋ก์ด ๋ชฉํ๋ฅผ ์ฐพ์๋ด๊ณ ์๋ก์ด ์ผ์ ์ํํ๋ ์์คํ โ์ด๋ผ๊ณ ์ ์ํฉ๋๋ค. ํ์ง๋ง ์๋ฌด๋ฆฌ ๋ฐ์ด๋ AI๋ผ๋, ์ธ๊ฐ์ ๊ฐ์ ์์ด๋ ๊ทธ ์ด๋ค ์์ ๋ ์ง์ ํ ๊ฐ์น๋ฅผ ์ฐฝ์ถํ๊ธฐ ์ด๋ ต๋ค๋ ๊ฒ์ด ๊ทธ์ ํต์ฌ ์ฃผ์ฅ์ ๋๋ค.
๋ฌผ๋ก ์ง์ ์ ๋ณํ๋ ํ์ฐ์ ์ ๋๋ค. ํ๋ฆฝ ๋งํด ์์ ์ ์ง์ ๋ AI ๋๋ถ์ ์์ ํ ๋ฌ๋ผ์ก์ต๋๋ค. โEveryโ์์ ๊ด์ฐฐ๋๋ ํ์๋ค์ด ๋ชจ๋ ์ฐ์ ๊ณผ ์ง์ ์ ๋์ผํ๊ฒ ์ ์ฉ๋์ง๋ ์์ ์ ์์ต๋๋ค. ํ์ง๋ง ํต์ฌ์ AI๊ฐ ์ธ๊ฐ์ ์ผ์ ์์ ํ ์ ๊ฑฐํ๋ ๊ฒ์ด ์๋๋ผ, ์ผ์ ๋ณธ์ง๊ณผ ๋ฐฉ์์ ๋ณํ์ํค๋ฉฐ ์๋ก์ด ๊ฐ์น๋ฅผ ์ฐฝ์ถํ ๊ธฐํ๋ฅผ ์ ๊ณตํ๋ค๋ ๊ฒ์ ๋๋ค.
๋ฏธ๋๋ฅผ ์ํ ์กฐ์ธ: โ๋ชจ๋ธ์ ํ๋ผ(Ride the Models)โ
๊ทธ๋ ๋ค๋ฉด ์ด๋ฌํ AI ์๋์ ์ฐ๋ฆฌ๋ ๋ฌด์์ ํด์ผ ํ ๊น์? ํ๋ฆฝ ๋งํด์ ์กฐ์ธ์ ๋ช ํํฉ๋๋ค. โ๊ทธ๋ฅ ๋ชจ๋ธ์ ํ๋ผ(Just ride the models).โ
AI ๋ชจ๋ธ์ด ๋ฐ์ ํ ์๋ก, ๋ ๋ง์ โ์ ๋ ดํ ์ธ๊ฐ ์ญ๋โ์ด ํจํค์ง ํํ๋ก ์ ๊ณต๋๋ฉฐ, ์ฐ๋ฆฌ๋ ์ด๋ฅผ ์ํ๋ ๋๋ก ํ์ฉํ ์ ์๊ฒ ๋ฉ๋๋ค. ๋๋ํ๊ณ ํธ๊ธฐ์ฌ ๋ง์ ์ฌ๋๋ค์๊ฒ AI๋ ์์ ์ ์ ์ฌ๋ ฅ์ ์ต๋ํ ๋ฐํํ ์ ์๋ ๊ฐ๋ ฅํ โ์ํผ ๋๊ตฌ(super tool)โ๊ฐ ๋ ๊ฒ์ ๋๋ค. ๊ทธ๋ฆฌ๊ณ ์ด๋ฌํ ์ญ๋์ ๊ฐ์ง ์ฌ๋๋ค์ ๋ํ ์์๋ ์์ผ๋ก ๋์ฑ ์ปค์ง ๊ฒ์ ๋๋ค.
AI๋ ๋๋ ค์์ ๋์์ด ์๋๋ผ, ์๋ก์ด ๊ฐ๋ฅ์ฑ์ ์ด์ด์ฃผ๋ ํ๋ ฅ์์ ๋๋ค. ๋ณํํ๋ ์ ๋ฌด ํ๊ฒฝ ์์์ AI ๋๊ตฌ๋ฅผ ์ ๊ทน์ ์ผ๋ก ํ์ตํ๊ณ ํ์ฉํ๋ ๊ฒ์ด์ผ๋ง๋ก, ๋ฏธ๋ ์๋์ ๊ฐ์ฅ ์ค์ํ ์ญ๋์ด ๋ ๊ฒ์ ๋๋ค.
โWhy the U.S. Just Indicted Cubaโs Former Presidentโ โ New York Times Podcasts ๊ธฐ๋ฐ ๊ธฐ์ฌ ์๋ณธ ์์ ๋ณด๊ธฐ
์นด์คํธ๋ก ๊ธฐ์, ๊ทธ ๋ค์ ์จ๊ฒจ์ง ๋ฏธ๊ตญ์ ์ฟ ๋ฐ ์ ๋ต
2020๋ 5์, ๋ฏธ๊ตญ์ด ์ฟ ๋ฐ์ ์ ๊ตญ๊ฐํ์ํ ์์ฅ ๋ผ์ธ ์นด์คํธ๋ก(Raul Castro)๋ฅผ ์ด์ธ ํ์๋ก ๊ธฐ์ํ๋ฉฐ ๊ตญ์ ์ฌํ์ ํฐ ํ์ฅ์ ์ผ์ผ์ผฐ์ต๋๋ค. ๋ด์ํ์์ฆ์ ํ์บ์คํธ โ๋ ๋ฐ์ผ๋ฆฌ(The Daily)โ๋ ์ด๋ก์ ์ธ ๊ธฐ์์ ๋ฐฐ๊ฒฝ๊ณผ ํธ๋ผํ ํ์ ๋ถ๊ฐ ์ฟ ๋ฐ์ ๋ํด ๊ฐ์ถ๊ณ ์๋ ๊ถ๊ทน์ ์ธ ๋ชฉํ๋ฅผ ์ฌ์ธต ๋ถ์ํ์ต๋๋ค. ์ด๋ ๋จ์ํ ๋ฒ์ ์กฐ์น๋ฅผ ๋์ด, ์์ญ ๋ ๊ฐ ์ด์ด์ง ๋ฏธ๊ตญ-์ฟ ๋ฐ ๊ด๊ณ์ ์ ํ์ ์ด ๋ ์ ์๋ ์ค๋ํ ์ ์น์ ๋ฉ์์ง์์ต๋๋ค.
๋ง์ด์ ๋ฏธ์ ์์ง์ ๊ธฐ์: ์์ ์ ํ์์ ์ธ๋ ค ํผ์ง ๊ณ ๋ฐ
2020๋ 5์ 20์ผ ์์์ผ, ๋ฏธ๊ตญ ๋ฒ๋ฌด๋ถ(Department of Justice)๋ ์ฟ ๋ฐ์ ์ ๊ตญ๊ฐํ์ํ ์์ฅ์ด์ ๊ณต์ฐ๋น ์๊ธฐ์ฅ์ด์๋ ๋ผ์ธ ์นด์คํธ๋ก๋ฅผ ์ด์ธ ํ์๋ก ๊ธฐ์ํ์ต๋๋ค. ์ด๋ฒ ๊ธฐ์๋ ์์ญ ๋ ์ , ์ฟ ๋ฐ ์ ๋ถ์ ์ํด ๋ฏธ๊ตญ์ธ ์ธ ๋ช ์ด ์ฌ๋งํ ์ฌ๊ฑด๊ณผ ๊ด๋ จ์ด ์์ต๋๋ค. ์ด๋ ์ฟ ๋ฐ์ ์ ์น์ ๋ณํ๋ฅผ ๊ฐ์ ํ๋ ค๋ ํธ๋ผํ ํ์ ๋ถ์ ์บ ํ์ธ ์ค ๊ฐ์ฅ ์ต๊ทผ์ด์ ๊ฐ์ฅ ๊ฐ๋ ฅํ ์กฐ์น๋ก ํ๊ฐ๋ฉ๋๋ค.
๋ด์ํ์์ฆ ๊ธฐ์ ํ๋ ์น ๋ก๋ฐ์ค(Francis Robas)๋ ๋ง์ด์ ๋ฏธ์์ ์ด๋ฆฐ ๋ฒ๋ฌด๋ถ ๊ธฐ์ํ๊ฒฌ ํ์ฅ์ ์ทจ์ฌํ๋ฉฐ ๊ทธ ์ด๋ก์ ์ธ ๋ถ์๊ธฐ๋ฅผ ์ ํ์ต๋๋ค. ์ผ๋ฐ์ ์ผ๋ก ๋ฒ์์ด๋ ๊ฒ์ฐฐ์ฒญ์์ ์ด๋ฆฌ๋ ์ด๋ฐ ๊ธฐ์ํ๊ฒฌ๊ณผ๋ ๋ฌ๋ฆฌ, ์ด๋ฒ ํ์ฌ๋ ๋ง์ด์ ๋ฏธ์ ์ ์ ๊น์ ๊ฑด๋ฌผ์ธ โํ๋ฆฌ๋ค ํ์(Freedom Tower)โ์์ ์งํ๋์์ต๋๋ค. ํ๋ฆฌ๋ค ํ์๋ 1960๋ ๋์ 70๋ ๋ ์ด ์ฟ ๋ฐ ๋๋ฏผ ์ผํฐ๋ก ์ฌ์ฉ๋์๋ ๊ณณ์ผ๋ก, ํ๋ช ์ดํ ์ฟ ๋ฐ๋ฅผ ๋ ๋ ๋ง์ด์ ๋ฏธ๋ก ์จ ์๋ง์ ์ฌ๋๋ค์๊ฒ โ์ฟ ๋ฐ ๋ง๋ช ์๋ค์ ์๋ฆฌ์ค ์์ผ๋๋(Ellis Island)โ์ ๊ฐ์ ์์ง์ ์ธ ์ฅ์์์ต๋๋ค.
๋์ฑ์ด ๊ธฐ์ํ๊ฒฌ์ด ์ด๋ฆฐ 5์ 20์ผ์ โ์ฟ ๋ฐ ๋ ๋ฆฝ ๊ธฐ๋ ์ผโ์ด์์ต๋๋ค. ์ด์ฒ๋ผ ์์ง์ ์ธ ์ฅ์์ ๋ ์ง์ ์งํ๋ ํ์ฌ๋ ๋จ์ํ ๋ฒ์ ์ ์ฐจ๋ฅผ ๋์ด์ ์ ์น์ ๋ฉ์์ง๊ฐ ๋ด๊ฒจ ์์์ ์์ฌํ์ต๋๋ค. ์๋ฐฑ ๋ช ์ ์ ์น์ธ, ์์ฅ, ์ ์ง ์ฐ๋ฐฉ ๊ฒ์ฌ ๋ฑ ๋ง์ด์ ๋ฏธ ์ฟ ๋ฐ ๋ง๋ช ๊ณต๋์ฒด์ ์ฃผ์ ์ธ์ฌ๋ค์ด ์ฐธ์ํ์ฌ ์นด์คํธ๋ก ๊ธฐ์์ ๋ํ ๋จ๊ฑฐ์ด ๊ด์ฌ์ ๋๋ฌ๋์ต๋๋ค.
๋ง์ด์ ๋ฏธ ๋ฐ์ด๋ ์นผ๋ฆฌ์ง(Miami Dade College) ์ด์ฅ์ธ ๋ง๋ฆฐ ํธ๋ง๋ฅด๊ฐ(Maline Pumarga)๋ โ์ฟ ๋ฐ์ ํ์ ์๋งค๋ค์๊ฒ, ์ฐ๋ฆฌ๋ ๋น์ ๋ค์ ๋ชฉ์๋ฆฌ๋ฅผ ๋ฃ์ต๋๋ค. ์์ ๊ฐ ์ฌ ๊ฒ์ ๋๋ค. ์ฟ ๋ฐ๋ ์์ ๋ก์ธ ๊ฒ์ ๋๋ค!โ๋ผ๊ณ ์ธ์น๋ฉฐ ํ์ฌ์ ํฌ๋ฌธ์ ์ด์์ต๋๋ค. ์ด์ด์ ํ ๋ ๋ธ๋์น(Todd Blanch) ๋ฒ๋ฌด๋ถ ์ฅ๊ด ๋ํ์ด ์๊ฐ๋์, ์ฐธ์์๋ค์ ๊ทธ๊ฐ ์ ์ ์ด๊ธฐ๋ ์ ์ ๊ธฐ๋ฆฝ ๋ฐ์์ ํํธ๋ฅผ ๋ณด๋์ต๋๋ค. ๊ทธ๋ค์ ๋ธ๋์น ์ฅ๊ด ๋ํ์ด ์ด๋ค ๋ฐํ๋ฅผ ํ ์ง ์ด๋ฏธ ์๊ณ ์์๊ธฐ ๋๋ฌธ์ ๋๋ค. ๋ธ๋์น ์ฅ๊ด ๋ํ์ โ์ค๋ ์ฐ๋ฆฌ๋ ๋ผ์ธ ์นด์คํธ๋ก์ ๋ค๋ฅธ ์ฌ๋ฌ ๋ช ์ ๋ฏธ๊ตญ ๊ตญ๋ฏผ ์ดํด ์๋ชจ ํ์๋ก ๊ธฐ์ํ๋ค๊ณ ๋ฐํํฉ๋๋คโ๋ผ๊ณ ๋งํ์๋ง์ ๋ค์ ํ๋ฒ ํฐ ๋ฐ์๋ฅผ ๋ฐ์์ต๋๋ค. ์นด์คํธ๋ก๋ ๋ฏธ๊ตญ ๊ตญ๋ฏผ ์ดํด ์๋ชจ, ํญ๊ณต๊ธฐ ํ๊ดด 2๊ฑด, ๊ทธ๋ฆฌ๊ณ ์ด์ธ 4๊ฑด์ ํ์๋ฅผ ๋ฐ๊ฒ ๋์์ต๋๋ค.
๋น๊ทน์ ๋ฐฐ๊ฒฝ: โ๊ตฌ์กฐ ํ์ ๋จโ ์ฌ๊ฑด์ ์ ๋ง
์ด๋ฒ ๊ธฐ์์ ํต์ฌ์ 1996๋ ์ ๋ฐ์ํ ๋น๊ทน์ ์ธ ์ฌ๊ฑด์ ๋๋ค. ๊ทธ๋ฌ๋ ์ด ์ฌ๊ฑด์ ์ดํดํ๊ธฐ ์ํด์๋ 1990๋ ๋ ์ด๋ก ๊ฑฐ์ฌ๋ฌ ์ฌ๋ผ๊ฐ์ผ ํฉ๋๋ค. ๋น์ ์ฟ ๋ฐ์ธ๋ค์ ์๋ง์ ์ฌ๋๋ค์ด ํ๋ธ๋ ๋ก์ ์ฐจ๋ฅผ ๊ฐ์กฐํ ๋ณดํธ, ๋๋ชฉ ๋ฑ์ ์ด์ฉํด ๋ฐ๋ค๋ฅผ ๊ฑด๋ ๋ฏธ๊ตญ์ผ๋ก ํฅํ๋ โ์ฟ ๋ฐ ๋๋ชฉ ๋๋ฏผ ์ฌํ(Cuban rafter crisis)โ๋ฅผ ๊ฒช๊ณ ์์์ต๋๋ค. ๋งค์ผ ์์ฒ ๋ช ์ ๋๋ฏผ์ด ํค์จ์คํธ(Key West)์ ๋์ฐฉํ๋ ๋ฑ ์ํฉ์ ๋งค์ฐ ์ฌ๊ฐํ์ต๋๋ค.
์ด๋ฌํ ์๊ธฐ ์์์ โ๊ตฌ์กฐ ํ์ ๋จ(Brothers to the Rescue)โ์ด๋ผ๋ ๋น์๋ฆฌ ๋จ์ฒด๊ฐ ํ์ํ์ต๋๋ค. ์ฟ ๋ฐ๊ณ ๋ฏธ๊ตญ์ธ ์กฐ์ข ์ฌ๋ค๋ก ๊ตฌ์ฑ๋ ์ด๋ค์ ๊ธฐ๊ธ์ ๋ชจ์ ์ํ ๋นํ๊ธฐ๋ฅผ ๊ตฌ์ ํ๊ณ , ํ๋ก๋ฆฌ๋ค ํดํ ์๊ณต์ ๋นํํ๋ฉฐ ๋๋ฏผ๋ค์ ์ฐพ์ ํด์ ๊ฒฝ๋น๋์ ์ฐ๋ฝํ์ฌ ์๋ฐฑ ๋ช ์ ์๋ช ์ ๊ตฌํ์ต๋๋ค. ์ด๋ค์ ํ๋์ ํด์ ๊ฒฝ๋น๋๊ฐ ๊ฐ๋นํ๊ธฐ ํ๋ค ์ ๋๋ก ๋ง์ ๋๋ฏผ๋ค์ด ๋ฐ์ํ๋ ์๊ธฐ์ ๋งค์ฐ ์ค์ํ ์ญํ ์ ํ์ต๋๋ค.
๊ทธ๋ฌ๋ 1994๋ ์ฌ๋ฆ, ํด๋ฆฐํด ํ์ ๋ถ์ ์ฟ ๋ฐ ์ ๋ถ๊ฐ ๋๋ฏผ ๋ฌธ์ ์ ๋ํ ํ์ ์ ๋งบ์ผ๋ฉด์ ์ํฉ์ ๋ณํ์ต๋๋ค. ๋ฏธ๊ตญ์ ์ฟ ๋ฐ๋ฅผ ๋ ๋๋ ์ฌ๋๋ค์ ๋ง๊ณ , ๋ฐ๋ค์์ ๋ถ์กํ ์ฌ๋๋ค์ ์ฟ ๋ฐ๋ก ๋๋ ค๋ณด๋ด๊ธฐ๋ก ํฉ์ํ์ต๋๋ค. ์ด์ ๋ฐ๋ผ โ๊ตฌ์กฐ ํ์ ๋จโ์ ๋ณธ๋์ ์๋ฌด๋ฅผ ์๊ฒ ๋์๊ณ , ์ ์ฐจ ์ฟ ๋ฐ ์ ๋ถ์ ๋ํ ์์์ฑ ํ๋์ผ๋ก ์ ํํ์ต๋๋ค. ์ด๋ค์ ์ฟ ๋ฐ ์๊ณต์ ์ ๋จ์ง๋ฅผ ๋ฟ๋ฆฌ๊ฑฐ๋ ์ข ๊ต ๋ฉ๋ฌ์ ํฌํํ๋ ๋ฑ ๋๋ฐ์ ์ธ ํ๋์ ๊ฐํํ์ต๋๋ค.
์ฟ ๋ฐ ์ ๋ถ๋ ์ด๋ฅผ ์๊ตญ ์๊ณต ์นจ๋ฒ์ผ๋ก ๋ณด๊ณ ๊ฒฉ๋ถํ์ผ๋ฉฐ, ๋ฏธ๊ตญ ์ ๋ถ์ ์ด๋ค์ ํ๋์ ์ค๋จ์ํฌ ๊ฒ์ ๊ฑฐ๋ญ ์์ฒญํ์ต๋๋ค. ๋ฏธ๊ตญ ์ ๋ถ์ ์ฌ๋ฌ ๊ด๊ณ์๋ค๋ โ๊ตฌ์กฐ ํ์ ๋จโ์๊ฒ ๊ฒฝ๊ณ ํ์ง๋ง, ๋จ์ฒด์ ์ค๋ฆฝ์ ํธ์ธ ๋ฐ์ ํ (Jose Basulto)๋ ์ด๋ฅผ ์ฟ ๋ฐ์ธ์ด์ ์ฟ ๋ฐ๊ณ ๋ฏธ๊ตญ์ธ์ผ๋ก์์ ๊ถ๋ฆฌ๋ผ๊ณ ์ฃผ์ฅํ๋ฉฐ ๋นํ์ ๊ฐํํ์ต๋๋ค.
๊ฒฐ๊ตญ 1996๋ 2์ 24์ผ, ๋ฐ์ ํ ์ ๋ ๋์ ๋ค๋ฅธ ๋นํ๊ธฐ๊ฐ ๋ง์ด์ ๋ฏธ ์ธ๊ณฝ์ ๊ณตํญ์ ์ด๋ฅํ์ฌ ์ฟ ๋ฐ ์๊ณต์ผ๋ก ํฅํ์ต๋๋ค. ์ด๋ค์ ๊ตญ์ ์๊ณต์ธ โ24๋์ (24th parallel)โ์ ๋์ด ๋นํํ๊ณ , ์ฟ ๋ฐ ํญ๊ณต ๊ตํต ๊ด์ ํ๊ณผ ๊ต์ ํ๋ ๊ณผ์ ์์ ๋ฐ์ ํ ๋ โ์์ ๋ก์ด ์ฟ ๋ฐ์ธ์ผ๋ก์์ ๋ด ๊ถ๋ฆฌโ๋ผ๊ณ ์ฃผ์ฅํ์ต๋๋ค. ์ด์ ์ฟ ๋ฐ ๊ตฐ์ ๋ ๋์ ๋นํ๊ธฐ๋ฅผ ๊ฒฉ์ถํ์ฌ ๋ฏธ๊ตญ ์๋ฏผ ์ธ ๋ช ์ ํฌํจํ ๋ค ๋ช ์ ๋ชฉ์จ์ ์์๊ฐ์ต๋๋ค. ์ด ์ฌ๊ฑด์ ๋น์ ๋ฏธ๊ตญ ํ๋ก๋ฆฌ๋ค ๋จ๋ถ ์ ์ญ์ ์์ฒญ๋ ์ถฉ๊ฒฉ์ ์๊ฒจ์ฃผ์์ต๋๋ค.
๋น์ ๊ตญ๋ฐฉ๋ถ ์ฅ๊ด์ด์๋ ๋ผ์ธ ์นด์คํธ๋ก๊ฐ ์ด ๊ฒฉ์ถ ๋ช ๋ น์ ๋ด๋ ธ๋ค๋ ๊ฒ์ ๊ฑฐ์ ์์ฌ์ ์ฌ์ง๊ฐ ์์ต๋๋ค. ์ฟ ๋ฐ ์กฐ์ข ์ฌ๋ค์ด ์๋ถ์ ๋ช ํํ ๋ช ๋ น ์์ด ๋นํ๊ธฐ๋ฅผ ๊ฒฉ์ถํ ๋ฆฌ ์์๊ธฐ ๋๋ฌธ์ ๋๋ค. ํ์ง๋ง ๋ฏธ๊ตญ์ ๋น์ ๊ทธ๋ฅผ ๊ธฐ์ํ์ง ์์์ต๋๋ค. ์ ์ง ์ฐ๋ฐฉ ๊ฒ์ฌ๋ค์ ์ธ๊ต ๊ด๊ณ์ ๋ฏธ์น ์์ฒญ๋ ๋ถ์์ ์ฑ ๋๋ฌธ์ด์๋ค๊ณ ์ค๋ช ํฉ๋๋ค.
ํธ๋ผํ ํ์ ๋ถ์ ์๋ฐ ์บ ํ์ธ: ๊ณ ์ฌ ์๊ธฐ์ ์ฟ ๋ฐ
์ด์ 94์ธ์ ๊ณ ๋ น์ธ ๋ผ์ธ ์นด์คํธ๋ก๋ฅผ ๊ธฐ์ํ ๊ฒ์ ํธ๋ผํ ํ์ ๋ถ๊ฐ ์ฟ ๋ฐ ์ ๊ถ์ ๋ํ ์๋ฐ์ ๊ฐํํ๋ ์บ ํ์ธ์ ์ต์ ๋จ๊ณ๋ก ํด์๋ฉ๋๋ค. ํธ๋ผํ ํ์ ๋ถ๋ ์ฟ ๋ฐ์ ๋ํ ์์ ๊ณต๊ธ์ ์ฐจ๋จํ๋ ๋ฑ ์ ๋ก ์๋ ์ ์ฌ๋ฅผ ๊ฐํ์ผ๋ฉฐ, ์ด๋ก ์ธํด ์๋ฐ๋(Havana) ๊ฐ์ ์ฃผ์ ๋์์์๋ ํ๋ฃจ 22์๊ฐ์ฉ ์ ๊ธฐ๊ฐ ๋๊ธฐ๋ ๋ฑ ์ฌ๊ฐํ ์ํฉ์ด ๋ฐ์ํ์ต๋๋ค.
ํ๋ ์น ๋ก๋ฐ์ค ๊ธฐ์๋ ์ฟ ๋ฐ ํ์ง์ธ๋ค์ ์ฐธ๋ดํ ์ํฉ์ ์ ํ์ต๋๋ค. ์ ๊ธฐ๊ฐ ๋๊ธฐ๋ ๊ฒ๋ณด๋ค ๋ค์ ๋ค์ด์ค๋ ๊ฒ์ด ๋ ๋ฌธ์ ๋ผ๋ ์์ด๋ฌ๋ํ ์ํฉ์ ๋๋ค. ์์ ๊ฐ๊น์ด ์ ๊ธฐ๊ฐ ๋ค์ด์ค๋ฉด ์ ์์ ๊นจ์ด๋ ํด๋ํฐ๊ณผ ๋ ธํธ๋ถ์ ์ถฉ์ ํ๊ณ , ๋ค์ ๋ ์์ฌ๋ฅผ ์ํด ๋ฐฅ๊ณผ ์ฝฉ์ ์๋ฆฌํด์ผ ํฉ๋๋ค. ์ฌ๋๋ค์ ์ง์ฅ๊น์ง ๋ช ์๊ฐ์ ๊ฑธ์ด๊ฐ์ผ ํ๊ณ , ์ ๊ธฐ๊ฐ ์์ด ์ผ์ ํ ์ ์๋ ๊ฒฝ์ฐ๋ ๋น์ผ๋น์ฌํฉ๋๋ค.
์ด๋ฌํ ์ํฉ์ ๋ํ ์ฑ ์ ์์ฌ๋ ์ธ๋๋ณ๋ก ๋ค๋ฅด๊ฒ ๋ํ๋ฉ๋๋ค. ๋์ด๋ ์ธ๋๋ ๋ฏธ๊ตญ์ ๋น๋ํ๋ ๊ฒฝํฅ์ด ์์ผ๋ฉฐ, ์ฟ ๋ฐ ์ ๋ถ ๋ํ ์ด ๋ชจ๋ ์ฌํ๊ฐ ๋ฏธ๊ตญ์ ์๋ชป์ด๋ผ๊ณ ์ฃผ์ฅํฉ๋๋ค. ๋ฐ๋ฉด ์ ์ ์ธ๋๋ ์ฟ ๋ฐ ์ ๋ถ์ ์๋ชป๋ ๊ตญ์ ์ด์์ด ์ด๋ํ ๊ฒฐ๊ณผ๋ผ๊ณ ๋น๋ํ๋ ๋ชฉ์๋ฆฌ๊ฐ ๋ ํฝ๋๋ค. ์ฟ ๋ฐ ์ ๋ถ๋ ๋ฏธ๊ตญ์ ์ ์ฌ๋ฅผ โ๋ถ๋ฒ์ ํ์ด(illegal genocide)โ์ด๋ผ๊ณ ๋น๋ํ๋ฉฐ, ์ด๋ ์ ํ ์ฌํ๋ก ์ด์ด์ง ๊ฒ์ด๊ณ ๊ทธ ์ฑ ์์ ๋ฏธ๊ตญ์ ์๋ค๊ณ ์ฃผ์ฅํฉ๋๋ค. ๊ทธ๋ค์ ๊ณผ๊ฑฐ์ ์ด๋ค ์ค์๋ ์ธ์ ํ์ง ์๊ณ ๋ชจ๋ ๊ฒ์ ํธ๋ผํ ํ์ ๋ถ์ ํ์ผ๋ก ๋๋ฆฌ๊ณ ์์ต๋๋ค.
์ฟ ๋ฐ ์ ๊ถ์ ์ง๋ 67๋ ๊ฐ ์๋ง์ ์ญ๊ฒฝ์ ํค์ณ์์ผ๋ฉฐ, ์ด๋ค ๋๊ฐ๋ฅผ ์น๋ฅด๋๋ผ๋ ๊ถ๋ ฅ์ ์ ์งํ๋ ค๋ ๊ฐํ ์์ง๋ฅผ ๊ฐ์ง๊ณ ์์ต๋๋ค. ์ผ๋ถ์์๋ ํธ๋ผํ ํ์ ๋ถ๊ฐ ์ฟ ๋ฐ ์ง๋๋ถ์ ์ด๋ฌํ ์ง๋ ์ ๊ณผ์ํ๊ฐํ๊ณ ์์ ์ ์๋ค๊ณ ์ง์ ํฉ๋๋ค.
๋ฏธ๊ตญ์ ์จ๊ฒจ์ง ์ ๋ต๊ณผ ์ต์ข ๋ชฉํ: ๋ฒ ๋ค์์๋ผ์ ์ ์ฒ ์ธ๊ฐ?
๋ฏธ๊ตญ ์ ๋ณด๊ธฐ๊ด์ ์ทจ์ฌํ๋ ๋ด์ํ์์ฆ์ ์ค๋ฆฌ์ ๋ฐ์ค(Julian Barnes) ๊ธฐ์๋ ์ด๋ฒ ์๋ฐ ์บ ํ์ธ๊ณผ CIA์ ์ญํ ์ ๋ํด ์ฌ์ธต ๋ณด๋ํ์ต๋๋ค. ๊ทธ๋ ์ฟ ๋ฐ์ ๋ํ ๋ฏธ๊ตญ์ ์ ํํ ์ต์ข ๋ชฉํ๋ ๋ถ๋ถ๋ช ํ์ง๋ง, ์๋ฐ์ด ๊ณ์๋ ๊ฒ์ด๋ผ๋ ์ ์ ๋ช ํํ๋ค๊ณ ๋งํฉ๋๋ค. ์ด๋ฒ ๊ธฐ์๋ ๋ฒ ๋ค์์๋ผ์์ ๋์ฝ๋ผ์ค ๋ง๋๋ก(Nicholas Maduro) ๋ํต๋ น์ด ๊ธฐ์๋ ํ ๋ฏธ๊ตฐ์ ์ํด ์ถ์ถ๋ ์ ์์๋ ๊ฒ์ฒ๋ผ, ์ฟ ๋ฐ ์ ๋ถ์๊ฒ๋ ๋น์ทํ ์ผ์ด ์ผ์ด๋ ์ ์๋ค๋ ๋ฉ์์ง๋ฅผ ๋ณด๋ด๋ ๊ฒ์ ๋๋ค.
๋๋๋ ํธ๋ผํ ๋ํต๋ น์ ๋ผ์ธ ์นด์คํธ๋ก๋ฅผ ๊ตฐ์ฌ์ ์ผ๋ก ์ถ์ถํ๋ ์ต์ ์ ์ํ๊ณ ์๋ค๋ ๋ณด๋๊ฐ ๋์์ต๋๋ค. ํ์ง๋ง ์ค์ ๊ตฐ์ฌ ์์ ์ผ๋ก ์นด์คํธ๋ก๋ฅผ ์ถ์ถํ ๊ฐ๋ฅ์ฑ์ ํฌ๋ฐํฉ๋๋ค. ๋๋ถ๋ถ์ ํน์ ์์ ๋ถ๋๋ ์ฌ์ ํ ์ค๋์ ์ฃผ๋ํ๊ณ ์์ผ๋ฉฐ, ์ด๋์ ๋ํ ์ ์ฌ์ ์กฐ์น์ ๋๋นํ๊ณ ์์ต๋๋ค. ์ฟ ๋ฐ๋ ๋ฒ ๋ค์์๋ผ์ ์ํฉ์ด ๋งค์ฐ ๋ค๋ฆ ๋๋ค. ๋ง๋๋ก์ ์ถ์ถ์ ๊ณต์์ ์ธ ์ง๋๋ถ ๊ต์ฒด๋ฅผ ์๋ฏธํ์ง๋ง, ๋ผ์ธ ์นด์คํธ๋ก๋ ๋น๋ก ์ ์น์ ์ผ๋ก ์ํฅ๋ ฅ์ด ํฌ์ง๋ง ํ์ฌ ๊ณต์์ ์ธ ๊ตญ๊ฐ์์๋ ์๋๋๋ค. ๊ทธ๋ฅผ ์ ๊ฑฐํ๋๋ผ๋ ์ฟ ๋ฐ ์ ๋ถ์ ๊ณต์์ ์ธ ์ง๋๋ถ๋ ๊ทธ๋๋ก ์ ์ง๋ ๊ฒ์ ๋๋ค. ๋ฐ๋ผ์ ์ธ๋ถ ์ ๋ฌธ๊ฐ๋ค์ ์ด๋ฒ ๊ธฐ์๊ฐ ์ค์ ํ๋๋ณด๋ค๋ ์ํ์ด์ ์๋ฐ์ ์ ํธ๋ก ๋ณด๊ณ ์์ต๋๋ค.
๊ทธ๋ ๋ค๋ฉด ๋ฏธ๊ตญ์ ์ฟ ๋ฐ๋ก๋ถํฐ ๋ฌด์์ ์ป์ผ๋ ค๋ ๊ฒ์ผ๊น์? ๋จ์ํ ์ ๊ถ ๊ต์ฒด๋ฅผ ๋ชฉํ๋ก ํ๋ ๊ฒ์ผ๊น์? ํธ๋ผํ ํ์ ๋ถ์ ๊ณต์์ ์ธ ์ ์ฅ์ ๋ค์ ๋ชจํธํ์ง๋ง, ์ต๊ทผ ์ผ๋ถ ๋จ์๋ค์ด ๋๋ฌ๋๊ธฐ ์์ํ์ต๋๋ค. ์กด ๋ํธํด๋ฆฌํ(John Ratcliffe) CIA ๊ตญ์ฅ์ ์ฟ ๋ฐ๋ฅผ ๋ฐฉ๋ฌธํ์ฌ ๋ผ์ธ ์นด์คํธ๋ก์ ์์์ด์ ๋ด๋ฌด๋ถ ์ฅ๊ด, ์ ๋ณด๊ตญ์ฅ์๊ฒ ๋ฉ์์ง๋ฅผ ์ ๋ฌํ์ต๋๋ค. ๊ทธ๋ โ์ฟ ๋ฐ๊ฐ ๋ฏธ๊ตญ์ ์ ๋๊ตญ์ ์ํ ํ๋ซํผ์ด ๋์ด์๋ ์ ๋๋คโ๊ณ ๊ฐ์กฐํ์ต๋๋ค. ์ด๋ ์ฟ ๋ฐ์ ์ฃผ๋ํ๊ณ ์๋ ์ค๊ตญ ์ ๋ณด ์์ค๊ณผ ๋ฌ์์ ๊ตฐ์ฌ ๊ธฐ์ง๋ฅผ ์ง์นญํ๋ ๊ฒ์ด์์ต๋๋ค. ์ด ์์ค๋ค์ ๋ฏธ๊ตญ ๋ด ํต์ ์ ๋์ฒญํ๋ ๋ฐ ์ฌ์ฉ๋๋ฉฐ, ๋ฏธ๊ตญ์ ์ด ๊ธฐ์ง๋ค์ ํ์๋ฅผ ๊ฐ๋ ฅํ ์๊ตฌํ๊ณ ์์ต๋๋ค. ์ด๋ ์ฟ ๋ฐ๊ฐ ์ค์ ๋ก ์ทจํ ์ ์๋ ๊ตฌ์ฒด์ ์ธ ์กฐ์น์ ๋๋ค.
๊ฒฝ์ ์ , ์ ์น์ ์๊ตฌ๋ ๋ค์ ๋ชจํธํ์ง๋ง, ๋ฏธ๊ตญ์ด ์ํ๋ ๊ฒ์ ๋ค์๊ณผ ๊ฐ์ต๋๋ค. ์ฒซ์งธ, ์ฟ ๋ฐ ๊ฒฝ์ ์ ์๋น ๋ถ๋ถ์ ํต์ ํ๋ ๊ตฐ์ฐ๋ณตํฉ์ฒด์ ๊ถ๋ ฅ์ ์ถ์ํ๋ ๊ฒ์ ๋๋ค. ๋์งธ, ๋ณด๋ค ์ค์ฉ์ ์ด๊ณ ๋ ๊ต์กฐ์ ์ธ ์ ์น์ธ๋ค์ ์ ๋ฉด์ ๋ด์ธ์ฐ๋ ๊ฒ์ ๋๋ค. ์ ์งธ, ์์ ์ ๊ฑฐ๋ฅผ ํฅํ ๋ฐ๊ฑธ์์ ๋ด๋๋ ๊ฒ์ ๋๋ค. ์๋ฅผ ๋ค์ด, ์ง๋ฐฉ ์ ๊ฑฐ๋ฅผ ๋์ฑ ์๋ฏธ ์๊ฒ ๋ง๋๋ ๊ฒ์ ์ ๊ถ ๊ต์ฒด๋ ์๋์ง๋ง ์ค์ง์ ์ธ ์ง์ ์ด ๋ ์ ์์ต๋๋ค.
ํธ๋ผํ ํ์ ๋ถ๊ฐ ์ด๋ฌํ ์์ ์ง์ ์ด๋ผ๋ ์ฑ๊ณต์ผ๋ก ๊ฐ์ฃผํ ์ง๋ ์ค์ํ ๋ฌธ์ ์ ๋๋ค. ๋ฒ ๋ค์์๋ผ์ ๊ฒฝ์ฐ, ๋ง๋๋ก๋ฅผ ๋ฏธ๊ตญ์ผ๋ก ๋ฐ๋ ค์ ์ฌํ์ ์ธ์ฐ๋ ๊ฒ์ด ํธ๋กํผ์ ๊ฐ์์ต๋๋ค. ํ์ง๋ง ์นด์คํธ๋ก๋ฅผ ์ถ์ถํ์ง ์๊ณ ๋ ์จ๊ฑดํ ๋ณํ๋ฅผ ๋ชฉํ๋ก ํ๋ค๋ฉด, ๊ถ์์ฃผ์ ์ ๋ถ์ ํํํ๋ค๋ ๋นํ์ ๋ฐ์ ์ํ์ด ์์ต๋๋ค. ๋ฏธ๊ตญ ๋ด์๋ ์ฟ ๋ฐ์ ๊ณต์ฐ์ฃผ์ ํต์น ์ข ์์ ์ผ์ํ๋ ๊ฐ๋ ฅํ ์ฟ ๋ฐ๊ณ ๋ฏธ๊ตญ์ธ ์ ๊ถ์ ์ง๋จ์ด ์กด์ฌํ๊ธฐ ๋๋ฌธ์ ๋๋ค. ๊ทธ๋ฌ๋ ํธ๋ผํ ํ์ ๋ถ๋ ํฐ ๋ชฉํ๋ฅผ ์ถ๊ตฌํ๋ค ์คํจํ๊ธฐ๋ณด๋ค, ์์ ๋ชฉํ๋ฅผ ํตํด ์ฑ๊ณต์ ๊ฑฐ๋๋ ๊ฒ์ด ๋ ํจ๊ณผ์ ์ผ ์ ์๋ค๋ ๊ฒ์ ์๊ณ ์์ต๋๋ค. ์ด๋๊ณผ์ ๊ด๊ณ์์ ๋ณด์๋ฏ์ด, ๊ด๋ฒ์ํ ์ ๊ถ ๊ต์ฒด๋ ๋งค์ฐ ์ด๋ ต๊ณ ์์ธก ๋ถ๊ฐ๋ฅํ ๊ฒฐ๊ณผ๋ฅผ ์ด๋ํ ์ ์์ต๋๋ค.
์ ์ง๊ธ์ธ๊ฐ? ํธ๋ผํ์ โ์น๋ฆฌโ ์ฐพ๊ธฐ
์ฟ ๋ฐ๋ ์์ญ ๋ ๊ฐ ๋ฏธ๊ตญ์ ๊ณจ์นซ๊ฑฐ๋ฆฌ์์ต๋๋ค. ์ด๋ฒ ์๋ฐ ์บ ํ์ธ์ด ํธ๋ผํ ํ์ ๋ถ๋ง์ ๋ ์์ ์ธ ๋ชจํ์ ์๋๋๋ค. ์ฌ๋ฌ ์ธ๋์ ๋ฏธ๊ตญ ์ง๋์๋ค์ด ์ฟ ๋ฐ ์ ๋ถ๋ฅผ ๋ฐ๊พธ๋ ค ๋ ธ๋ ฅํด์์ผ๋ฉฐ, ์ด๋ ๋ฏธ๊ตญ ๋ด ๊ฐ๋ ฅํ ์ ๊ถ์๋ค์๊ฒ ๋งค์ฐ ์ค์ํ ๋ฌธ์ ์ ๋๋ค. ํธ๋ผํ ๋ํต๋ น์ ์์ ์ ๊ด๋ฃ์ฃผ์๋ฅผ ํํํ๊ณ ์ค๋ ๋ฌธ์ ๋ค์ ํด๊ฒฐํ ์ ์๋ ์ธ๋ฌผ๋ก ์ฌ๊น๋๋ค. ๊ทธ๋ ์ง์ ์ด๋ ์ญ์ฌ์ ๋ํ ์ธ๋ด์ฌ์ด ์ ๊ณ , ์์ง์ ํ์ผ๋ก ๋ฌธ์ ๋ฅผ ํด๊ฒฐํ ์ ์๋ค๊ณ ๋ฏฟ์ต๋๋ค.
๋ํ, ํ์ฌ ์ด๋๊ณผ์ ๊ด๊ณ๊ฐ ๋งค์ฐ ๋ณต์กํ๊ณ ์ด๋ ค์ด ์ํฉ์์, ํธ๋ผํ ํ์ ๋ถ ๋ด ์ผ๋ถ ์ธ์ฌ๋ค์ ์ฟ ๋ฐ์์์ ์ฑ๊ณต์ด ์ด๋์ ๋๊ตญ์ ํ๊ฐํ๋ ๋ฐ ๋์์ด ๋ ์ ์๋ค๊ณ ๋ณด๊ณ ์์ต๋๋ค. ์ฆ, ํธ๋ผํ ๋ํต๋ น์ ํญ์ โ์น๋ฆฌโ๋ฅผ ์ฐพ๊ณ ์์ผ๋ฉฐ, ์ฟ ๋ฐ๋ ํ์ฌ ๊ทธ์ ์๋ฐ ์บ ํ์ธ์ด ์์๋ณด๋ค ์ ์๋ํ๊ณ ์์ด ์ ์ฌ์ ์ธ โ์น๋ฆฌโ์ ๊ธฐํ๊ฐ ๋ ์ ์๋ค๋ ๊ฒ์ ๋๋ค.
๋ผ์ธ ์นด์คํธ๋ก์ ๋ํ ๊ธฐ์๋ ๋จ์ํ ๋ฒ์ ์ฌ๊ฑด์ด ์๋๋๋ค. ์ด๋ ํธ๋ผํ ํ์ ๋ถ์ ๋์ฟ ๋ฐ ์ ๋ต์ ๋ํ ๊ฐ๋ ฅํ ์ ํธํ์ด๋ฉฐ, ๋ณต์กํ ์ญ์ฌ์ ์ ์น์ ์ดํด๊ด๊ณ๊ฐ ์ฝํ ๋ฏธ๊ตญ-์ฟ ๋ฐ ๊ด๊ณ์ ๋ฏธ๋๋ฅผ ๊ฐ๋ ํ๋ ์ค์ํ ์งํ๊ฐ ๋ ๊ฒ์ ๋๋ค.