English
Based on โThis Founder is Making 1B+ Excel Workers 20x Faster | Meridian, John Lingโ from EO Watch the original video
Beyond the Grid: How John Ling is Turning One Billion Excel Users into AI Powerhouses
In the glass towers of Manhattan, the lights rarely go out. Inside, thousands of the worldโs brightest young minds are hunched over glowing monitors, performing a ritual that has remained largely unchanged for thirty years: manual data entry in Microsoft Excel. They are building Leveraged Buyout (LBO) models, three-statement financials, and complex valuations, one cell at a time.
To John Ling, the co-founder and CEO of Meridian, this isnโt just a grueling rite of passage for junior bankersโitโs a massive, untapped opportunity for a technological revolution.
โI donโt believe any person on the planet has spent a thousand hours trying to build financial models with AI,โ Ling says. โIf you think about the bankers, theyโre just like, โWeโre just going to do it by hand.โ And if you donโt know how to do it, you probably just donโt know how to do it.โ
Ling is on a mission to change that. Backed by over $15 million in funding led by Andreessen Horowitz, his startup, Meridian, is building the โAI for spreadsheets.โ His goal is bold: to take the one billion people who live in Excel and make them 20 times faster.
The First Principles of a High Performer
Lingโs path to Meridian wasnโt accidental. It was forged at Scale AI, the data-labeling juggernaut where he earned a reputation as a โtop 1% performer.โ At Scale, Ling was known for a relentless, first-principles approach to business. He didnโt just stay in his lane; he treated every problem in the company as his own, viewing โmore workโ as simply โmore opportunities to learn.โ
โI think I just really enjoyed learning new things,โ Ling explains. โThere are 50 problems, and with each problem, you learn a little bit more about something completely different.โ
This intellectual curiosity allowed him to witness the development of Large Language Models (LLMs) from the front row. He spent his days obsessing over data quality, benchmarks, and internal efficiencies. It was here that he realized the trajectory of the world was about to changeโand that most people were still looking in the wrong direction.
The โVibe Codingโ Revelation
The spark for Meridian came from a personal โahaโ moment. Ling began using Cursor, an AI-powered code editor, and watched his productivity explode. Tasks that previously took two weeks were suddenly being completed in thirty minutes. He calls this โvibe codingโโa state where the barrier between thought and execution vanishes, mediated by a โsuper-powerful calculator.โ
โI had a moment where I was just like, โWow, this thing is magical,โโ Ling says. โI wanted everyone to use it. I felt like if you donโt know how to vibe code, youโre going to be left behind.โ
However, when he traveled from the tech-heavy atmosphere of San Francisco to New York, he found a jarring disconnect. While SF was buzzing with โcrazy architecturesโ and AI unlocks, the finance world in New York was still operating on manual labor.
The reason, Ling realized, was a lack of decomposition. In coding, developers build the tools they use; they know exactly where a model fails and why. In finance, the workflows are so specialized and opaque that most people assume AI simply โcanโt do it.โ Ling disagreed. He saw that if you break a complex financial workflow into its constituent parts, AI can handle many of them exceptionally wellโif someone is willing to do the investigation.
Excel: The Worldโs Most Distributed Programming Language
At the heart of Meridianโs vision is a fundamental reclassification of what a spreadsheet actually is. Ling doesnโt see Excel as just a grid for numbers; he sees it as the most widely used programming language on Earth.
โOur goal is to say, โHey, how can we help all the people that spend a lot of time in spreadsheet software today move 20 times faster?โโ
Meridian isnโt just about automating a few formulas. Itโs about โinfusing the work done in spreadsheets with meaningful intelligence.โ By augmenting the knowledge worker in the same way AI has augmented the software developer, Meridian aims to transform the very nature of white-collar work.
The market is staggering. Knowledge work is one of the largest sectors of the global economy, and spreadsheets are its primary tool. If Meridian can successfully apply the โCursor modelโ to the financial world, they arenโt just building a toolโthey are redefining a profession.
The Philosophy of โThe Askโ and the Power of Failure
Lingโs leadership style is as unconventional as his product. He promotes an environment where experimentation is mandatory and failure is a recognized part of the process.
โYou learn by trying things youโve never tried before. And if you [mess] up, you [mess] up. Itโs okay,โ he says. At Meridian, if an employee fails at a task, the response isnโt a reprimand; itโs a surge of support from the rest of the team to find a new way forward.
This โsuspension of disbeliefโ is what Ling believes separates entrepreneurs from everyone else. Itโs the same mindset that leads him to encourage others to reach out to anyoneโeven industry titans.
โDonโt think that Satya Nadella will never respond to your email,โ Ling advises. โIf you think that way, he obviously never will. But if you reach out, you might be surprised.โ
The Future: 10,000 Hours of AI Finance
As Meridian grows, Ling remains obsessed with the idea of โintuition.โ He believes that by spending thousands of hours playing with the technology, one develops a โsixth senseโ for what AI will be able to do three, six, or twelve months down the line.
He also offers a surprising insight into the value of prompting. While many see prompting as a chore, Ling sees it as a tool for self-clarity. Much like the Y Combinator application process forces founders to articulate their business, explaining a task to an LLM forces a human to actually understand what they want to achieve.
โLearning how to be specific about your ask gives yourself a lot of clarity,โ Ling says.
For the one billion people currently โdoing it by hand,โ that clarityโcombined with Meridianโs technologyโmight just be the key to the next great leap in human productivity. John Ling isnโt just building a company; heโs trying to build a masterpiece that transforms how the world thinks, one cell at a time.
Based on โFrom Coder to Manager: Navigating the Shift to Agentic Engineering with Notion Co-Founder Simon Lastโ from No Priors: AI, Machine Learning, Tech, & Startups Watch the original video
The Agentic Office: How Notion is Reimagining Work for a World of Autonomous Agents
In late 2022, during a company offsite in Mexico, Simon Last and his Notion co-founder Ivan Zhao sat down with an early version of GPT-4. It wasnโt just a better chatbot; it was a revelation. For the founders of a company built on the philosophy of โtools for thought,โ the realization was instantaneous: the nature of โthoughtโ in the digital workspace was about to change forever.
โIt was immediately clear that the time was now,โ Last recalls. โIt was smart enough to follow complicated instructions, and the scope of its knowledge was incredibly deep. We knew it was only going to get better.โ
That moment sparked a fundamental pivot for Notion. What began as a versatile workspace for humans to organize their notes and databases is rapidly evolving into a collaborative platform where humans and AI agents work side-by-side. In a wide-ranging conversation on the No Priors podcast, Last outlined how this โagenticโ shift is rewriting the rules of engineering, product design, and personal productivity.
From Writing Assistant to General Assistant
Notionโs AI journey began with what Last calls the โshort-term visionโ: the AI Writer. Launched in early 2023, it served as a sophisticated editing partnerโrewriting paragraphs, changing tones, and summarizing meeting notes. It was a โsingle-stepโ task that required no external memory.
But the โlong-term visionโ was far more ambitious. Notion wanted to build a general assistant capable of using the platform exactly like a human does: querying databases, creating documents, and weaving together disparate pieces of information to complete complex, multi-day tasks.
The bridge between these two visions was โQ&A,โ a feature that performs a semantic index of a userโs entire workspace. Unlike traditional keyword searches, which often fail in the messy reality of corporate wikis, semantic indexing allows the AI to understand the meaning of a query.
โThe AI doesnโt really care what the tree structure of your folders is,โ Last explains. โAll it cares about is finding the snippet of text that has the context you need. We tell people now: donโt worry as much about organization. Just get the data in there.โ
The โSwitzerlandโ of Models
One of Notionโs core strategies in the AI arms race is neutrality. Rather than tethering itself to a single provider, Notion views itself as the โSwitzerland for models.โ
Whether itโs OpenAIโs latest frontier model, Anthropicโs Claude, or high-performing open-source models from China, Notionโs infrastructure is designed to be model-agnostic. This allows users to swap the โbrainโ of their assistant as technology evolves, ensuring they arenโt locked into a provider that might be leapfrogged next month.
โOur customers donโt want to be locked in,โ says Last. โWe want to be the place where you can get access to all the best models at any time, in a collaborative workspace that is designed for humans and agents to coordinate.โ
Designing for a New Kind of User: The Agent
Perhaps the most surprising technical shift at Notion is the realization that their software now has a new type of customer: the AI itself.
Historically, Notionโs internal data structures (APIs) were designed for humans writing code. They were verbose, complex, and โhorrible for an agent,โ according to Last. To solve this, the engineering team had to rethink their primitives. They developed a specific โMarkdown dialectโ that translates Notionโs complex block-based structure into a format LLMs can read and write naturally. They even integrated SQLite interfaces to help agents interact with databases more efficiently.
โWe essentially have a new customer, which is the agent,โ Last says. โWe had to take on the engineering challenge of making our APIs convenient for them. Now, the agents are naturally good at it.โ
The Death of Manual Coding
The shift to โagentic engineeringโ isnโt just something Notion sells; itโs how they work. Last, a co-founder of a multi-billion dollar tech company, admits he has undergone a radical personal transformation.
โI havenโt written code since last summer,โ Last reveals. โI donโt type code anymore. Iโve gone from being a coder to an agent manager.โ
In the modern Notion engineering workflow, humans no longer spend their days typing out syntax. Instead, they design end-to-end tasks, set up verification loops, and act as the โouter verifierโ to ensure the agentโs output is correct. This hasnโt necessarily reduced team sizes, but it has drastically widened the gap between the median engineer and the โ100xโ engineer who can effectively harness these tools.
Last shares a personal record: he once had a coding agent running for 13 days straight, autonomously working through a backlog of tasks while he monitored its progress. โMy goal these days is to have as many agents running as possible,โ he says. โEvery night before I go to bed, I make sure Iโve given them enough work so that theyโre still spinning when I wake up.โ
The Future: Custom Agents for Every Task
The latest milestone in this evolution is the launch of Custom Agents. Unlike a general personal assistant, a Custom Agent can be assigned a specific โjobโโlike a digital employee.
Last uses several himself. One is an email triage agent that has access to his work and personal accounts. Through a process of โinterviewsโ and feedback, the agent learned his preferences and now automatically archives 95% of his inbox, leaving only the essential messages. Another agent lives in a Slack channel, listening for bug reports and product feedback, then autonomously routing them to the correct engineering teamโs database.
โThe concept is very intuitive once you get past the technical barrier,โ Last notes. โItโs a very human-like interface.โ
A New Philosophy of Work
Before the AI revolution, Notionโs mission was to be the best tool for humans to perform their work. Today, that mission has been updated: Notion aims to be the best tool for humans to manage agents who do the work for them.
While the mission has shifted, the core building blocks of Notionโthe pages, the databases, the Kanban boardsโremain as relevant as ever. In a world of autonomous digital workers, humans still need a โcoordination structureโ to see what is happening.
โIf youโre working with a swarm of 100 background coding agents, you donโt want 100 chat threads,โ Last concludes. โYou want a Kanban board. Itโs the same coordination structure as before, just with a much faster set of workers.โ
Based on โPalantir CEO on Iran, AI Weapons and Americaโs Advantage | a16z American Dynamism Summitโ from a16z Watch the original video
The Architect of Deterrence: Alex Karp on AI, the โFreak Show,โ and the Moral Necessity of American Might
For two decades, Palantir was the โfreak showโ of Silicon Valley. While other startups were busy optimizing ad clicks or disrupting laundry services, Alex Karp and his team were embedded in the dark, unfashionable corners of the world: intelligence agencies, war zones, and the logistical labyrinths of the Department of Defense.
At the fourth annual a16z American Dynamism Summit, Karp, the CEO of Palantir and the โOGโ of the movement to relink technology with national interest, didnโt look like a man seeking vindication. He looked like a man who had already found it. In a wide-ranging conversation, Karp laid out a stark vision for the future: one where American military superiority is the only thing standing between Western liberalism and a new era of authoritarian dominance.
The Return of Deterrence
The conversation began against a backdrop of escalating global tension. Addressing the volatile state of the Middle East and the necessity of Western intervention, Karp was unapologetic. For years, he argued, American deterrence had been โeviscerated.โ Now, through a combination of raw courage and cutting-edge technology, that deterrence is being rebuilt in real-time.
โWe are the power that actually has the decisive vote,โ Karp stated, โand that is with military superiority.โ
He pointed to recent operations as evidence of a new technological reality. In Karpโs view, the dominance displayed by the U.S. and its allies isnโt just about having more boots on the ground; itโs about a โtrinityโ of software, hardware, and AI. This isnโt the โparasiticโ software of the pastโthe kind Karp mocks as โsupplying a steak dinnerโ to bureaucratsโbut a specialized orchestration of data that allows the West to see, decide, and act faster than any adversary.
โI literally believe weโre doing the work of a higher purpose,โ Karp said. โItโs us, or China, or Russia. I donโt know how you feel about those decisions, but I believe we are making sure we have the decisive vote.โ
The Silicon Valley Blind Spot
Karpโs sharpest barbs were reserved for his neighbors in Palo Alto. He identified a dangerous disconnect between the tech elite and the rest of the country. Silicon Valley, he argued, is obsessed with โpositive-sumโ gamesโthe idea that everyone can win through innovation. But in the realm of global power, Karp insists the game is โzero-sum.โ
He issued a dire warning to the creators of Large Language Models (LLMs) and white-collar automation: if the tech industry is perceived as a force that destroys domestic jobs while refusing to support the American warfighter, the backlash will be existential.
โIf Silicon Valley believes we are going to take away everyoneโs white-collar jobโฆ and youโre going to screw the military, if you donโt think thatโs going to lead to the nationalization of our technology, youโre [expletive],โ Karp warned.
He described a โhorseshoe effectโ where the far left and the far right eventually agree on only one thing: that the tech industry is not โpaying the billsโ for society and should be seized by the state. To avoid this, Karp argues that the tech elite must stop being โeffing spoiledโ and start empathizing with the โsoldier from Iowaโ who risks everything to protect the system that allowed them to get rich.
The Meritocracy of the Battlefield
One of the more surprising moments of the talk was Karpโs defense of the Department of Defense as a social institution. He described the military as the most meritocratic environment in Americaโa place that integrated long before the rest of society and remains the only institution revered across every demographic.
โIf you were a Black American, you got your break in America by going to the military,โ Karp noted. โItโs the only institution revered by the American peopleโฆ precisely because itโs been meritocratic.โ
For Karp, the moral mission of Palantir is simple: ensure that the American warfighter is the most likely person to come home safely, and that those trying to harm them know they wonโt. This isnโt just about lethality; itโs about the preservation of a culture that allows for dissent, individuality, and freedom.
The Neurodivergent Advantage
Karp, who is famously open about his own dyslexia and โintrovertedโ nature, linked Americaโs technological edge directly to its tolerance for โfreaks.โ He views the United States as a sanctuary for the โneurologically divergentโโthose whose minds donโt fit the standard mold but who possess the โoutlier IQโ necessary to build world-changing technology.
โYou could define our wonderful country as a place where everyone who was divergent in ideology, thought, religion, or neurologically came to have a better place where they could express their freedom,โ he said.
This philosophy extends to the culture at Palantir. Karp describes his role as an โartist,โ managing a collection of brilliant, often difficult individuals who wouldnโt thrive in a traditional corporate environment. He pushes back against โwokeโ corporate culture, which he views as a performance of being different while everyone actually thinks the same.
โI tend to gravitate towards people who are unique,โ Karp explained. โI donโt really care about their politics. I care about their ability to think and do.โ
The โMarkโ in the Room
As the summit concluded, Karp offered a piece of advice for the new generation of โAmerican Dynamismโ foundersโthose building the next generation of defense tech. His warning was against the arrogance of intelligence.
โThe single biggest mistake people make in this area is because theyโre intelligent in one area, they assume theyโre intelligent in all areas,โ Karp said. โIf you donโt know who the mark is, youโre the mark.โ
For Karp, success in the high-stakes world of national defense requires more than just a high IQ; it requires the humility to talk to a general or a soldier and realize that their expertise is just as vital as a line of code.
Palantir may have started as a โfreak show,โ but in Alex Karpโs eyes, it has become the blueprint for how America wins the 21st century: by embracing its outliers, arming its soldiers with the best software on earth, and never apologizing for holding the decisive vote.
ํ๊ตญ์ด
โThis Founder is Making 1B+ Excel Workers 20x Faster | Meridian, John Lingโ โ EO ๊ธฐ๋ฐ ๊ธฐ์ฌ ์๋ณธ ์์ ๋ณด๊ธฐ
โ์์ ๋ ธ๋์ 10์ต ๋ช ์ ์ํ ํด๋ฐฉ๊ตฐโ โ ๋ฉ๋ฆฌ๋์ ์กด ๋ง์ด ๊ทธ๋ฆฌ๋ AI ์คํ๋ ๋์ํธ์ ๋ฏธ๋
์ ์ธ๊ณ์์ ๊ฐ์ฅ ๋๋ฆฌ ์ฐ์ด๋ ํ๋ก๊ทธ๋๋ฐ ์ธ์ด๋ ๋ฌด์์ผ๊น? ํ์ด์ฌ(Python)๋, ์๋ฐ(Java)๋ ์๋๋ค. ๋ฐ๋ก ๋ง์ดํฌ๋ก์ํํธ์ **์์ (Excel)**์ด๋ค. ์ ์ธ๊ณ 10์ต ๋ช ์ด์์ ์ฌ์ฉ์๊ฐ ๋งค์ผ ์คํ๋ ๋์ํธ ์์์ ์ซ์๋ฅผ ๋ค๋ฃจ๊ณ ๋น์ฆ๋์ค ์์ฌ๊ฒฐ์ ์ ๋ด๋ฆฐ๋ค. ํ์ง๋ง ๊ธฐ์ ์ ์ต์ ์ ์ธ ์ค๋ฆฌ์ฝ๋ฐธ๋ฆฌ์์ ์ฝ๋ฉ ๋ฐฉ์์ด ํ๋ช ์ ์ผ๋ก ๋ณํ๋ ๋์, ๊ธ์ต๊ณผ ์ปจ์คํ ์ ์ค์ฌ์ง์ธ ์์คํธ๋ฆฌํธ์ ํ๊ฒฝ์ ์์ญ ๋ ์ ๊ณผ ํฌ๊ฒ ๋ค๋ฅด์ง ์๋ค.
์ค์ผ์ผ AI(Scale AI) ์ถ์ ์ ์ฐ์ ์ฐฝ์ ๊ฐ์ด์ **๋ฉ๋ฆฌ๋์(Meridian)**์ ๊ณต๋ ์ฐฝ์ ์์ธ ์กด ๋ง(John Ling) CEO๋ ์ด ์ง์ ์ ์ฃผ๋ชฉํ๋ค. ๊ทธ๋ ์ ๋ชจ๋๊ฐ ์์์ ์ผ๋ก ๊ธ์ต ๋ชจ๋ธ์ ๋ง๋๋ ๋ฐ ์๋ฐฑ ์๊ฐ์ ํ๋นํ๊ณ ์๋์ง ์๋ฌธ์ ๋์ง๋ค. ์๋๋ ์ผ ํธ๋ก์์ธ (a16z)๋ก๋ถํฐ 1,500๋ง ๋ฌ๋ฌ(์ฝ 200์ต ์) ์ด์์ ํฌ์๋ฅผ ์ ์นํ๋ฉฐ ์ฃผ๋ชฉ๋ฐ๊ณ ์๋ ๊ทธ๋ฅผ ํตํด, AI๊ฐ ๋ฐ๊ฟ ์ง์ ๋ ธ๋์ ๋ฏธ๋๋ฅผ ๋ค์ฌ๋ค๋ณด์๋ค.
1. ์ค์ผ์ผ AI์์์ ๊นจ๋ฌ์: โ๋ฐฐ์์ ์ ๋ฌด์ ์ฐ์ฅ์ ์ด๋คโ
์กด ๋ง์ ๋ฉ๋ฆฌ๋์์ ์ธ์ฐ๊ธฐ ์ , ๋ฐ์ดํฐ ๋ผ๋ฒจ๋ง ๋ฐ AI ์๋ฃจ์ ๋ถ์ผ์ ์ ๋์ฝ ๊ธฐ์ ์ธ **์ค์ผ์ผ AI(Scale AI)**์์ ํต์ฌ ์ธ์ฌ๋ก ํ์ฝํ๋ค. ๊ทธ๊ณณ์์ ๊ทธ๋ AI ๊ธฐ์ ์ด ์ธ์์ ์ด๋ป๊ฒ ๋ฐ๊ฟ์ง ๊ฐ์ฅ ๊ฐ๊น์ด์์ ์ง์ผ๋ณด๋ฉฐ โ์ 1์์น(First Principles)โ์ ๊ธฐ๋ฐํ ์ฌ๊ณ ๋ฐฉ์์ ์ตํ๋ค.
โ์ ๋ ์๋ก์ด ๊ฒ์ ๋ฐฐ์ฐ๋ ๊ณผ์ ์ ์ ๋ง ์ฆ๊น๋๋ค. ์ ๋ฌด๊ฐ ๋ง์์ง๋ค๋ ๊ฑด ๊ทธ๋งํผ ๋ ๋ง์ด ๋ฐฐ์ธ ๊ธฐํ๊ฐ ์๊ธด๋ค๋ ๋ป์ด์์ฃ .โ
๊ทธ๋ ์ค์ผ์ผ AI์์ ์ง๋ฌด์ ๊ฒฝ๊ณ์ ๊ฐํ์ง ์์๋ค. ๋ฐ์ดํฐ์ ํ์ง์ ์ ์ํ๊ณ , ๋๊ท๋ชจ ์ธ์ด ๋ชจ๋ธ(LLM)์ ๋ฒค์น๋งํฌ(์ฑ๋ฅ ์ธก์ ๊ธฐ์ค)๋ฅผ ์ค์ ํ๋ฉฐ, ๋ด๋ถ ํ๋ก์ธ์ค๋ฅผ AI๋ก ํจ์จํํ๋ ๋ฐฉ๋ฒ์ ๋์์์ด ์ฐ๊ตฌํ๋ค. ๋๋ฃ๋ค์ ๊ทธ๋ฅผ โ์์ 1%์ ์ฑ๊ณผ์โ๋ก ๊ธฐ์ตํ๋ค. ๊ทธ๋ ๋จ์ํ ์ฃผ์ด์ง ์ผ์ ํ๋ ๊ฒ์ด ์๋๋ผ, โ์ด ์ผ์ ์ ํด์ผ ํ๋๊ฐ?โ๋ฅผ ๋์์์ด ์ง๋ฌธํ๋ฉฐ ๊ธฐ์ ์ ๊ถค์ ์ ์ถ์ ํ๋ค.
2. โ๋ฐ์ด๋ธ ์ฝ๋ฉ(Vibe Coding)โ์ ์ถฉ๊ฒฉ: 2์ฃผ์ ์์ ์ 30๋ถ์ผ๋ก
์กด ๋ง์ด AI์ ๋ง๋ฒ์ ์ค๊ฐํ ์๊ฐ์ ์ฝ๋ฉ ๋๊ตฌ์ธ โ์ปค์(Cursor)โ๋ฅผ ์ฌ์ฉํ๋ฉด์๋ถํฐ์๋ค. ์์ **โ๋ฐ์ด๋ธ ์ฝ๋ฉ(Vibe Coding)โ**์ด๋ผ ๋ถ๋ฆฌ๋ ์ด ๋ฐฉ์์ ๊ฐ๋ฐ์๊ฐ ๋ณต์กํ ๋ฌธ๋ฒ์ ์ฝ๋งค์ด์ง ์๊ณ , AI์ ๋ํํ๋ฉฐ ์ง๊ด์ ์ผ๋ก ์ฝ๋ฉํ๋ ๊ฒ์ ์๋ฏธํ๋ค.
โ๊ณผ๊ฑฐ์ 2์ฃผ์ผ์ด ๊ฑธ๋ฆฌ๋ ์์ ์ด ๋จ 30๋ถ, ๊ธธ์ด์ผ ๋ฐ๋์ ๋ง์ ๋๋๋ ๊ฒ์ ๋ณด์์ต๋๋ค. ์ด๊ฑด ๋จ์ํ ๋๊ตฌ๊ฐ ์๋๋ผ โ์ํผ ๊ณ์ฐ๊ธฐโ๊ฐ ๋ฑ์ฅํ ๊ฒ๊ณผ ๊ฐ์์ฃ .โ
ํ์ง๋ง ๊ทธ๋ ๋ด์์ ๊ธ์ต๊ฐ ์น๊ตฌ๋ค์ ๋ง๋ ๋๋ง๋ค ๊ธฐ๋ฌํ ๊ดด๋ฆฌ๊ฐ์ ๋๊ผ๋ค. ์ํ๋์์ค์ฝ์ ๊ฐ๋ฐ์๋ค์ AI๋ก ์์ฐ์ฑ์ ์์ญ ๋ฐฐ์ฉ ๋์ด๊ณ ์๋๋ฐ, ๋ด์์ ์ํ๊ฐ๋ค์ ์ฌ์ ํ ์์์ ์ผ๋ก LBO(์ฐจ์ ๋งค์, Leveraged Buyout) ๋ชจ๋ธ์ ๋ง๋ค๋ฉฐ ๋ฐค์ ์ง์์ฐ๊ณ ์์๊ธฐ ๋๋ฌธ์ด๋ค.
3. โ์๋ฌด๋ AI๋ก ๊ธ์ต ๋ชจ๋ธ์ 1,000์๊ฐ ๋์ ๋ง๋ค์ด๋ณด์ง ์์๋คโ
์กด ๋ง์ ํต์ฌ์ ์ธ ํต์ฐฐ์ ์ ์ํ๋ค. ์ฝ๋ฉ ๋ถ์ผ์์๋ ๊ฐ๋ฐ์๋ค์ด ์ค์ค๋ก๋ฅผ ์ํ AI ๋๊ตฌ๋ฅผ ์ง์ ๋ง๋ค๊ธฐ ๋๋ฌธ์ ๋ฐ์ ์ด ๋น ๋ฅด์ง๋ง, ๊ธ์ต์ด๋ ์ปจ์คํ ๊ฐ์ ์ ๋ฌธ ์์ญ์์๋ ์ ๋ฌธ๊ฐ๋ค์ด AI๋ฅผ ๊น์ด ์๊ฒ ํ๊ณ ๋ค ์๊ฐ์ด ์์๋ค๋ ๊ฒ์ด๋ค.
โ๋จ์ธ์ปจ๋, ์ง๊ตฌ์์์ AI๋ง์ ์ด์ฉํด ๊ธ์ต ๋ชจ๋ธ์ ๊ตฌ์ถํ๋ ๋ฐ 1,000์๊ฐ์ ์์๋ถ์ ์ฌ๋์ ๋จ ํ ๋ช ๋ ์์ ๊ฒ๋๋ค. ์ํ๊ฐ๋ค์ ๊ทธ์ โ์ง์ ์์ผ๋ก ํ๋ ๊ฒ ๋น ๋ฅด๋คโ๊ณ ์๊ฐํ์ฃ . ํ์ง๋ง ์ ๋ฌด ํ๋ฆ์ ์ธ๋ถํ(Decomposition)ํด๋ณด๋ฉด, AI๊ฐ ๋๋ผ์ธ ์ ๋๋ก ์ ํด๋ผ ์ ์๋ ์์ญ์ด ๋ถ๋ช ํ ์กด์ฌํฉ๋๋ค.โ
๊ทธ๋ ์ฌ๋๋ค์ด AI๊ฐ ๋ด๋์ ๋ต์ด ํ๋ ธ์ ๋ โ์ด ๋ชจ๋ธ์ ๋ชป ์ฐ๊ฒ ๋คโ๋ผ๊ณ ์ฝ๊ฒ ํฌ๊ธฐํ๋ ๊ฒฝํฅ์ ์ง์ ํ๋ค. ์ ์ซ์๊ฐ ํ๋ ธ๋์ง, ์ด๋ค ๋ ผ๋ฆฌ์ ๋จ๊ณ์์ ์ค๋ฅ๊ฐ ๋ฌ๋์ง ์ง์ํ๊ฒ ํ๊ณ ๋๋ โ์กฐ์ฌ(Investigation)โ์ ๊ณผ์ ์ด ์์๋ค๋ ๊ฒ์ด๋ค. ๋ฉ๋ฆฌ๋์์ ๋ฐ๋ก ์ด ๋ฌธ์ ๋ฅผ ํด๊ฒฐํ๊ธฐ ์ํด ํ์ํ๋ค.
4. ๋ฉ๋ฆฌ๋์์ ๋น์ : ์คํ๋ ๋์ํธ์ 20๋ฐฐ ๊ฐ์ํ
๋ฉ๋ฆฌ๋์์ ๋ชฉํ๋ ๋ช ํํ๋ค. ์คํ๋ ๋์ํธ ์ฌ์ฉ์๋ฅผ ์ํ AI๋ฅผ ๊ตฌ์ถํ์ฌ ๊ทธ๋ค์ ์ ๋ฌด ์๋๋ฅผ 20๋ฐฐ ์ด์ ๋์ด๋ ๊ฒ์ด๋ค.
- ์ ๋ฌด์ ๋ถํด์ ์ฌ๊ตฌ์ฑ: ๋ณต์กํ ๊ธ์ต ๋ชจ๋ธ๋ง ๊ณผ์ ์ AI๊ฐ ์ฒ๋ฆฌ ๊ฐ๋ฅํ ๋จ์๋ก ๋๋์ด ์๋ํํ๋ค.
- ์ ๋ฌธ ์ง์์ ์ด์: ์ฝ๋ฉ ๋๊ตฌ๊ฐ ๊ฐ๋ฐ์์ ์๋๋ฅผ ์ฝ๋ฏ, ๋ฉ๋ฆฌ๋์์ ์ฌ๋ฌด ๋ถ์๊ฐ๋ ์ปจ์คํดํธ์ ๋ ผ๋ฆฌ๋ฅผ ์ดํดํ๋ AI๋ฅผ ์งํฅํ๋ค.
- ์ง์ ๋ ธ๋์ ๋ณํ: ์์ ์ ์ฌ์ฉํ๋ ์ ์ธ๊ณ ์์ต ๋ช ์ ์ง์ ๋ ธ๋์๋ค์ด ๋จ์ ๋ฐ๋ณต ์์ ์์ ๋ฒ์ด๋ ๋ ๊ฐ์น ์๋ ์์ฌ๊ฒฐ์ ์ ์ง์คํ๊ฒ ๋ง๋ ๋ค.
5. ์ฐฝ์ ๊ฐ ์ ์ : โ์ ๋ ๊ฑฐ๋ผ๋ ๋ฏฟ์์ ๊ฑฐ๋ถํ๋ผโ
์กด ๋ง์ ์๋น ์ฐฝ์ ๊ฐ๋ค์๊ฒ **โ๋ถ์ ์ ์ ์ํ๋ ๋ฅ๋ ฅ(Suspension of Disbelief)โ**์ด ํ์ํ๋ค๊ณ ๊ฐ์กฐํ๋ค. ๋จ๋ค์ด โ๊ทธ๊ฑด ๋ถ๊ฐ๋ฅํดโ, โ๋ฏธ์น ์ง์ด์ผโ๋ผ๊ณ ๋งํ ๋, โ์ ์ ๋ผ? ํ๋ฒ ํด๋ณด์โ๋ผ๊ณ ๋งํ ์ ์๋ ์ฉ๊ธฐ๋ค.
๊ทธ๋ ์ฌํฐ์ ๋๋ธ๋ผ(๋ง์ดํฌ๋ก์ํํธ CEO) ๊ฐ์ ๊ฑฐ๋ฌผ์๊ฒ๋ ์ด๋ฉ์ผ์ ๋ณด๋ด๋ ๊ฒ์ ๋๋ ค์ํ์ง ๋ง๋ผ๊ณ ์กฐ์ธํ๋ค. โ๋ต์ฅ์ด ์ ์ฌ ๊ฑฐ๋ผ๊ณ ๋ฏฟ์ผ๋ฉด ๋น์ฐํ ์ ์ค๊ฒ ์ง๋ง, ๋ณด๋ด๋ณด๋ฉด ์์ธ์ ๊ฒฐ๊ณผ๊ฐ ์๊ธธ ์๋ ์์ต๋๋ค. ์๋ํ์ง ์์ผ๋ฉด ์๋ฌด ์ผ๋ ์ผ์ด๋์ง ์์ฃ .โ
๋ํ, ์คํจ๋ฅผ ์ฉ์ธํ๋ ๋ฌธํ๋ฅผ ๊ฐ์กฐํ๋ค. โ์ฐ๋ฆฌ ํ์ฌ๋ ์ง์์ด ์ค์ค๋ก ๋ฌธ์ ๋ฅผ ํด๊ฒฐํ๋ ค๋ค ์คํจํ๋ ๊ฒ์ ์ฅ๋ คํฉ๋๋ค. ์คํจํ๋ฉด ๋๋ฃ๋ค์ด ๋ฌ๋ ค๊ฐ ๋์์ฃผ๋ฉด ๋ฉ๋๋ค. ์คํํ๊ณ ์คํจํด๋ ๊ด์ฐฎ์ ํ๊ฒฝ์ ๋ง๋ค์ด์ผ ํ์ ์ด ์ผ์ด๋ฉ๋๋ค.โ
๊ฒฐ๋ก : AI ์๋, ์ธ๊ฐ์๊ฒ ํ์ํ ๊ฒ์ โ๋ช ํํ ์ง๋ฌธโ
์กด ๋ง์ AI์ ๋ ๋ง์ ์๊ฐ์ ๋ณด๋ผ์๋ก ๊ธฐ์ ์ ๋ํ ์ง๊ด์ด ์๊ธฐ๋ฉฐ, ์ด๋ 3๊ฐ์ ํ, 6๊ฐ์ ํ์ ๋ฏธ๋๋ฅผ ์์ธกํ๋ ํ์ด ๋๋ค๊ณ ๋งํ๋ค. ํนํ ๊ทธ๋ AI์๊ฒ ์์ ์ ์ค๋ช ํ๋ ๊ณผ์ ์์ฒด๊ฐ ์ธ๊ฐ์๊ฒ ํฐ ์ ์ต์ ์ค๋ค๊ณ ๊ฐ์กฐํ๋ค.
โAI์๊ฒ ๋ฌด์์ ์ํ๋์ง ๊ตฌ์ฒด์ ์ผ๋ก ์ค๋ช ํ๋ ค๊ณ ๋ ธ๋ ฅํ๋ค ๋ณด๋ฉด, ์คํ๋ ค ์๊ธฐ ๋จธ๋ฆฟ์์ด ๋ช ํํด์ง๋๋ค. ๋ด๊ฐ ์ ๋ง๋ก ๋ฌด์์ ํ๋ ค๋์ง ์ค์ค๋ก ๊นจ๋ซ๊ฒ ๋๋ ๊ฒ์ด์ฃ .โ
๋ฉ๋ฆฌ๋์์ด ๊ทธ๋ฆฌ๋ ๋ฏธ๋๋ ๋จ์ํ ์์ ์นธ์ ์๋์ผ๋ก ์ฑ์ฐ๋ ๊ฒ์ด ์๋๋ค. ๊ทธ๊ฒ์ ์ธ๊ฐ์ ๋ณต์กํ ๋ ผ๋ฆฌ๋ฅผ AI์ ํ์ ํ์ฌ ๋ ๋น ๋ฅด๊ณ ์ ํํ๊ฒ ํ์คํํ๋, ์๋ก์ด ์ฐจ์์ ์ง์ ๋ ธ๋ ์๋๋ฅผ ์ฌ๋ ๊ฒ์ด๋ค. 10์ต ๋ช ์ ์์ ์ฌ์ฉ์๋ค์ด โ20๋ฐฐ ๋น ๋ฅธ ์๋โ๋ฅผ ๊ฐ๊ฒ ๋ ๋, ์ธ์์ ๊ณผ์ฐ ์ด๋ค ๋ชจ์ต์ผ๋ก ๋ณํ ๊น? ์กด ๋ง๊ณผ ๋ฉ๋ฆฌ๋์์ ํ๋ณด๊ฐ ๊ธฐ๋๋๋ ์ด์ ๋ค.
โFrom Coder to Manager: Navigating the Shift to Agentic Engineering with Notion Co-Founder Simon Lastโ โ No Priors: AI, Machine Learning, Tech, & Startups ๊ธฐ๋ฐ ๊ธฐ์ฌ ์๋ณธ ์์ ๋ณด๊ธฐ
[๋งค๊ฑฐ์ง ํ ํฌ] ์ฝ๋์์ ๊ด๋ฆฌ์๋ก: ๋ ธ์ ๊ณต๋ ์ฐฝ์ ์ ์ฌ์ด๋จผ ๋ผ์คํธ๊ฐ ๊ทธ๋ฆฌ๋ โ์์ด์ ํธ ์์ง๋์ด๋งโ์ ๋ฏธ๋
โ์ด์ ์ ๋ ์ฝ๋๋ฅผ ์ง์ ํ์ดํํ์ง ์์ต๋๋ค. ๋์ ์์ด์ ํธ ๊ตฐ๋จ์ ๊ด๋ฆฌํ์ฃ .โ
์ ์ธ๊ณ ์์ฒ๋ง ๋ช ์ ์์ฐ์ฑ์ ์ฑ ์์ง๋ ๋๊ตฌ, ๋ ธ์ (Notion)์ ๊ณต๋ ์ฐฝ์ ์ ์ฌ์ด๋จผ ๋ผ์คํธ(Simon Last)๊ฐ ๋์ง ์ด ํ๋ง๋๋ ์ค๋๋ ์ํํธ์จ์ด ์์ง๋์ด๋ง๊ณผ ์์ฐ์ฑ ๋๊ตฌ๊ฐ ์ง๋ฉดํ ๊ฑฐ๋ํ ํจ๋ฌ๋ค์ ๋ณํ๋ฅผ ์์งํฉ๋๋ค. ํ์บ์คํธ โNo Priorsโ์ ์ถ์ฐํ ๊ทธ๋ ๋ ธ์ ์ด ๋จ์ํ ๋ฉ๋ชจ ์ฑ์ ๋์ด ์ธ๊ฐ๊ณผ AI ์์ด์ ํธ๊ฐ ํ์ ํ๋ ํ๋ซํผ์ผ๋ก ์งํํ๋ ๊ณผ์ ๊ณผ, ๊ทธ ๊ณผ์ ์์ ๋ชฉ๊ฒฉํ ๊ธฐ์ ์ ยท์กฐ์ง์ ๋ณํ๋ฅผ ์ฌ๋ ์๊ฒ ๊ณต์ ํ์ต๋๋ค.
1. ๋ฉ์์ฝ์์์ ๊นจ๋ฌ์: โ์๊ฐ์ด ์๋คโ
๋ ธ์ ์ด AI๋ฅผ ๋์ ํ๊ฒ ๋ ๊ฒฐ์ ์ ์ธ ๊ณ๊ธฐ๋ 2022๋ ๋ง, ๋ฉ์์ฝ์์ ์ด๋ฆฐ ํ์ฌ ์ํฌ์์ผ๋ก ๊ฑฐ์ฌ๋ฌ ์ฌ๋ผ๊ฐ๋๋ค. ๋น์ ์ฌ์ด๋จผ๊ณผ ๊ณต๋ ์ฐฝ์ ์ ์ด๋ฐ ์(Ivan Zhao)๋ ์ถ์ ์ ์ด์๋ GPT-4๋ฅผ ์ฒ์ ์ ํ์ต๋๋ค.
์ฌ์ด๋จผ์ ๋น์๋ฅผ ์ด๋ ๊ฒ ํ์ํฉ๋๋ค. โ๋ณต์กํ ์ง์๋ฅผ ์๋ฒฝํ๊ฒ ๋ฐ๋ฅด๋ ์ง๋ฅ, ๊ทธ๋ฆฌ๊ณ ๋ฐฉ๋ํ๊ณ ๊น์ ์ธ๊ณ ์ง์์ ์ ์จ์ ๋๊ผ์ต๋๋ค. ๊ทธ ์๊ฐ ์ด๋ฐ๊ณผ ์ ๋ ํ์ ํ์ฃ . โ์ง๊ธ์ด ๋ฐ๋ก AI๋ฅผ ์ ์ฉํ ๋โ๋ผ๋ ๊ฒ์์.โ
๋ ธ์ ์ ์ฆ์ โํ์ด๊ฑฐ ํ(Tiger Team, ํน์ ๋ชฉํ๋ฅผ ์ํด ๊ตฌ์ฑ๋ ์์ ์ ์ ํ)โ์ ๊พธ๋ ค ๋ ๊ฐ์ง ํธ๋์ ๋น์ ์ ์ค์ ํ์ต๋๋ค.
- ๋จ๊ธฐ ๋น์ : ๋ฌธ์ ์์ฑ์ ๋๋ โAI ๋ผ์ดํฐ(AI Writer)โ. (2023๋ 2์ ์ถ์)
- ์ฅ๊ธฐ ๋น์ : ์ธ๊ฐ์ฒ๋ผ ๋ฐ์ดํฐ๋ฒ ์ด์ค๋ฅผ ์ฟผ๋ฆฌํ๊ณ , ๋ฌธ์๋ฅผ ์์ ํ๋ฉฐ, ๋ณต์กํ ์ ๋ฌด๋ฅผ ์์จ์ ์ผ๋ก ์ํํ๋ โ๋ฒ์ฉ ์์ด์ ํธ(General Assistant)โ.
2. ๊ฒ์์ ๋ฏธํ: ์ ๋ ธ์ ์ ๋ค๋ฅธ ์๋น์ค๋ณด๋ค ๊ฒ์์ ์ํ ๊น?
๋ ธ์ AI์ ํต์ฌ ๊ธฐ๋ฅ ์ค ํ๋๋ โQ&Aโ์ ๋๋ค. ์ฌ์ฉ์๊ฐ ์ง๋ฌธํ๋ฉด ์ํฌ์คํ์ด์ค ๋ด์ ๋ฐฉ๋ํ ๋ฐ์ดํฐ๋ฅผ ๊ฒ์ํด ๋ต์ ๋ด๋๋ ๋ฐฉ์์ด์ฃ . ํฅ๋ฏธ๋ก์ด ์ ์ ๋ ธ์ ์ด ์ฌ๋(Slack)์ด๋ ๊ตฌ๊ธ ๋๋ผ์ด๋ธ(Google Drive) ๊ฐ์ ์ธ๋ถ ํด๊น์ง ์ธ๋ฑ์ฑ(Indexing)ํ์ฌ ํตํฉ ๊ฒ์ ์๋น์ค๋ฅผ ์ ๊ณตํ๋ค๋ ๊ฒ์ ๋๋ค.
โ์๋ ๊ทธ ์๋น์ค๋ฅผ ๋ง๋ ํ์ฌ๋ค๋ณด๋ค ๋ ธ์ ์ด ๊ฒ์์ ๋ ์ํ๋ ๋น๊ฒฐ์ด ๋ญ๋โ๋ ์ง๋ฌธ์ ์ฌ์ด๋จผ์ **โ์ฅ์ธ์ ์ (Craft)โ**๊ณผ **โ๊ฒฝํ์ ๋ฐ๋ณต(Empirical Iteration)โ**์ ๊ผฝ์์ต๋๋ค.
โ๋ฐ์ดํฐ ์์ค๋ง๋ค ํน์ฑ์ด ๋ค๋ฆ ๋๋ค. ์ฌ๋์ ๋ฉ์์ง์ ๊ตฌ๊ธ ๋๋ผ์ด๋ธ์ ๋ฌธ์๋ ์์ ํ ๋ค๋ฅธ ํํ์ฃ . ์ฐ๋ฆฌ๋ ๋จ์ํ ๋๊ท๋ชจ ์ธ์ด ๋ชจ๋ธ(LLM)์ ์ฐ๊ฒฐํ๋ ๋ฐ ๊ทธ์น์ง ์๊ณ , ๋งค์ผ ์ง์ ์ฟผ๋ฆฌ๋ฅผ ๋์ ธ๋ณด๋ฉฐ ๊ฒ์ ํ์ง์ ํ๋ํฉ๋๋ค. โAI์ ์ค๋ ๋(AI-pilled)โ ๊ฐ๊ฐ๊ณผ ๋ํ ์ผ์ ๋ํ ์ง์ฐฉ์ด ์ฐจ์ด๋ฅผ ๋ง๋ญ๋๋ค.โ
์ด์ ๋ ธ์ ์ ์ฌ์ฉ์๋ค์๊ฒ โ์ ๋ฆฌ์ ๋๋ฌด ๊ณต๋ค์ด์ง ๋ง๋ผโ๊ณ ์กฐ์ธํฉ๋๋ค. AI๊ฐ ์๋ฒ ๋ฉ(Embedding, ๋ฐ์ดํฐ๋ฅผ ๋ฒกํฐ ํํ๋ก ๋ณํํ๋ ๊ธฐ์ )์ ํตํด ๊ตฌ์กฐ์ ์๊ด์์ด ํ์ํ ๋งฅ๋ฝ์ ์ฐพ์๋ผ ์ ์๊ธฐ ๋๋ฌธ์ ๋๋ค.
3. ์์ง๋์ด๋ง์ ์ข ๋ง? ์๋, โ1,000๋ฐฐ ์์ง๋์ดโ์ ํ์
์ฌ์ด๋จผ ๋ผ์คํธ๋ ๋ ธ์ ๋ด๋ถ์ ๊ฐ๋ฐ ๋ฌธํ๊ฐ ์ง๋ 2~3๋ ์ฌ์ด ์์ ํ ๋ฐ๋์๋ค๊ณ ๊ฐ์กฐํฉ๋๋ค. ๊ฐ์ฅ ๋๋ผ์ด ๋ณํ๋ โAI ํ๋์ค(AI Harness, AI ๋ชจ๋ธ์ ์ ํ์ ์ฐ๊ฒฐํ๋ ์์คํ ์ธํ๋ผ)โ๋ฅผ 6๊ฐ์๋ง๋ค ํต์งธ๋ก ์๋ก ์ด๋ค๋ ์ ์ ๋๋ค.
โ๊ธฐ์ ๋ฐ์ ์๋๊ฐ ๋๋ฌด ๋นจ๋ผ์ ์์ ๋ฐฉ์์ ๊ณ ์ํ๋ฉด ๋ค์ฒ์ง๋๋ค. ๋งค๋ฒ ์๋ก ์์ํ๋ ๊ฒ ์คํ๋ ค ๋ ํจ์จ์ ์ด์ฃ .โ
์ด๋ฌํ ๊ณผ๊ฐํ ์๋๊ฐ ๊ฐ๋ฅํ ์ด์ ๋ ๋ ธ์ ์์ง๋์ด๋ค์ด โ์ฝ๋ฉ ์์ด์ ํธโ๋ฅผ ์ ๊ทน์ ์ผ๋ก ํ์ฉํ๊ธฐ ๋๋ฌธ์ ๋๋ค. ์ฌ์ด๋จผ ๋ณธ์ธ๋ ์๋ ์ฌ๋ฆ ์ดํ๋ก ์ง์ ์ฝ๋๋ฅผ ํ์ดํํ์ง ์์ต๋๋ค. ๋์ ์์ด์ ํธ์๊ฒ ์ ์ฒด์ ์ธ ์ค๊ณ๋ฅผ ๋งก๊ธฐ๊ณ , ๊ฒฐ๊ณผ๋ฌผ์ ๊ฒ์ฆ(Verify)ํ๊ณ ๋ชจ๋ํฐ๋งํ๋ โ๊ด๋ฆฌ์โ ์ญํ ์ ์ํํฉ๋๋ค.
๊ทธ๋ ์ด ์ง์ ์์ **โ1,000๋ฐฐ ์์ง๋์ดโ**์ ๊ฐ๋ ์ ์ ์ํฉ๋๋ค.
- ์ค๊ฐ ์์ค์ ์์ง๋์ด: ๋๊ตฌ ํ์ฉ ์ญ๋์ ๋ฐ๋ผ ์ฑ๊ณผ๊ฐ ๊ฒฐ์ ๋จ.
- ์์ ์์ง๋์ด: ์์ด์ ํธ๋ฅผ ์กฐ์จํ์ฌ ํผ์์ ์์ญ ๋ช ๋ถ์ ๊ฒฐ๊ณผ๋ฌผ์ ๋ง๋ค์ด๋.
- ๊ฒฐ๊ณผ: ํํ์ (Minimum bar)์ ๊ทธ๋๋ก์ง๋ง, ์ํ์ (Maximum bar)์ด ๋ฌดํ๋๋ก ํ์ฅ๋จ.
4. ์์จ ์ฃผํํ๋ ์ํฌ์คํ์ด์ค: โ์ปค์คํ ์์ด์ ํธโ์ ๋ฑ์ฅ
์ต๊ทผ ๋ ธ์ ์ โ์ปค์คํ ์์ด์ ํธ(Custom Agents)โ ๊ธฐ๋ฅ์ ์ ๋ณด์์ต๋๋ค. ์ด๋ ์ฌ์ฉ์๊ฐ ํน์ ๋ชฉ์ ์ ์ํด ๋ง๋ AI ๋น์๋ก, ๋ฐฐ๊ฒฝ์์ ์์จ์ ์ผ๋ก ์๋ํฉ๋๋ค. ์ฌ์ด๋จผ์ ์์ ์ ์ค์ ํ์ฉ ์ฌ๋ก๋ฅผ ์๊ฐํ์ต๋๋ค.
- ์ด๋ฉ์ผ ์ ๋ฆฌ ์์ด์ ํธ: ๋ชจ๋ ์ ๋ฌด ๋ฐ ๊ฐ์ธ ์ด๋ฉ์ผ์ ์ฝ๊ณ , ์ค์ํ์ง ์์ ๋ฉ์ผ์ ์๋์ผ๋ก ์์นด์ด๋ธํฉ๋๋ค. ์ด๊ธฐ์๋ ์ฌ์ด๋จผ์ ํผ๋๋ฐฑ์ ๋ฐ์ผ๋ฉฐ ํ์ต(Memory)์ ์์๊ณ , ์ด์ ๋ ์๋ฒฝํ๊ฒ ๊ทธ์ ์ทจํฅ์ ํ์ ํด ์ด๋ฉ์ผ ์คํธ๋ ์ค๋ฅผ ์์ ์ฃผ์์ต๋๋ค.
- ๋ฒ๊ทธ ๋ผ์ฐํ ์์ด์ ํธ: ์ฌ๋ ์ฑ๋์ ์ฌ๋ผ์ค๋ ์๋ง์ ๊ณ ๊ฐ ํผ๋๋ฐฑ๊ณผ ๋ฒ๊ทธ ์ ๊ณ ๋ฅผ ๋ถ์ํด, ๋ด๋น ํ์ ๋ฐ์ดํฐ๋ฒ ์ด์ค์ ์๋์ผ๋ก ํ ๋นํฉ๋๋ค.
๊ทธ๋ โ์์ด์ ํธ๊ฐ ์ค์ค๋ก์ ๋ฅ๋ ฅ์ ํ์ฅํ๋ โ๋ถํธ์คํธ๋ํ(Bootstrapping)โ์ด ๋ฏธ๋์ ํต์ฌโ์ด๋ผ๊ณ ๋งํฉ๋๋ค. ์์ด์ ํธ๊ฐ ํ์ํ ๋๊ตฌ๊ฐ ์๋ค๋ฉด ์ค์ค๋ก ์ฝ๋๋ฅผ ์ง์ ํตํฉ ๊ธฐ๋ฅ์ ๋ง๋ค๊ณ ์คํํ๋ ๋จ๊ณ๊น์ง ๋์๊ฐ๋ ๊ฒ์ ๋๋ค.
5. ๊ฒฐ๋ก : ๋๊ตฌ์์ ํํธ๋๋ก, ๋ ธ์ ์ ์๋ก์ด ์ฒ ํ
๋ ธ์ ์ ๊ทผ๊ฐ์ ์ด๋ฅธ๋ฐ โ์๊ฐ์ ์ํ ๋๊ตฌ(Tools for Thought)โ ์ปค๋ฎค๋ํฐ์ ์์ต๋๋ค. AI ์๋ ์ด์ ์ ๋ ธ์ ์ด โ์ธ๊ฐ์ด ์ง์ ์ผ์ ํ๊ธฐ ์ํ ์ต๊ณ ์ ๋๊ตฌโ์๋ค๋ฉด, ์ด์ ๋ ธ์ ์ ๋ชฉํ๋ **โ์ธ๊ฐ์ด ์์ด์ ํธ๋ฅผ ๊ด๋ฆฌํ๋ฉฐ ํจ๊ป ์ผํ๋ ์ต๊ณ ์ ํ๋ซํผโ**์ด ๋๋ ๊ฒ์ ๋๋ค.
์ฌ์ด๋จผ ๋ผ์คํธ๋ ๋งํฉ๋๋ค. โ์ฐ๋ฆฌ๋ ์ฌ์ ํ ๋ฌธ์(Document)์ ๋ฐ์ดํฐ๋ฒ ์ด์ค(Database)๋ผ๋ ๊ธฐ๋ณธ ๋จ์๊ฐ ํ์ํฉ๋๋ค. ๋ค๋ง ์ด์ ๊ทธ๊ฒ๋ค์ ์ธ๊ฐ๋ฟ๋ง ์๋๋ผ AI ์์ด์ ํธ๊ฐ ์ฝ๊ณ ์ฐ๊ธฐ์๋ ๊ฐ์ฅ ํธํ ํํ์ฌ์ผ ํ์ฃ .โ
๋ ธ์ ์ ํ์ฌ ์์ด์ ํธ๊ฐ ๋ ์ ์ดํดํ ์ ์๋๋ก โ๋งํฌ๋ค์ด(Markdown) ๋ณํ ์ธ์ดโ๋ฅผ ์ค๊ณํ๊ณ , ์์ด์ ํธ ์ ์ฉ API๋ฅผ ์ต์ ํํ๋ ๋ฐ ์ง์คํ๊ณ ์์ต๋๋ค. ์ธ๊ฐ๊ณผ AI๊ฐ ํ ๊ณต๊ฐ์์ ์บ๋ฒ์ค(Canvas)๋ฅผ ๊ณต์ ํ๋ฉฐ ํ์ ํ๋ ์๋, ๋ ธ์ ์ ๊ทธ ๊ฑฐ๋ํ ๋ณํ์ ์ค์ฌ์์ โ์์ด์ ํธ ๋งค๋์ โ๊ฐ ๋ ์ฐ๋ฆฌ๋ฅผ ๊ธฐ๋ค๋ฆฌ๊ณ ์์ต๋๋ค.
์ด ๊ธฐ์ฌ๋ ํ์บ์คํธ โNo Priorsโ์ ์ฌ์ด๋จผ ๋ผ์คํธ ์ธํฐ๋ทฐ ๋ด์ฉ์ ๋ฐํ์ผ๋ก ์์ฑ๋์์ต๋๋ค.
โPalantir CEO on Iran, AI Weapons and Americaโs Advantage | a16z American Dynamism Summitโ โ a16z ๊ธฐ๋ฐ ๊ธฐ์ฌ ์๋ณธ ์์ ๋ณด๊ธฐ
โ์ ์์ ์นํจ๋ AI๊ฐ ๊ฒฐ์ ํ๋คโ ํ๋ํฐ์ด CEO ์๋ ์ค ์นดํ๊ฐ ๋์ง๋ ์๊ตฌ ๋ฌธ๋ช ์ ๊ฒฝ๊ณ ์ ํฌ๋ง
์ต๊ทผ ์ค๋ฆฌ์ฝ๋ฐธ๋ฆฌ์์ ๊ฐ์ฅ ๋จ๊ฑฐ์ด ํ๋๋ ๋จ์ฐ โ์๋ฉ๋ฆฌ์นธ ๋ค์ด๋ด๋ฏธ์ฆ(American Dynamism)โ์ ๋๋ค. ์ด๋ ๊ธฐ์ ํ์ ์ ํตํด ๋ฏธ๊ตญ์ ๊ตญ๊ฐ ๊ฒฝ์๋ ฅ๊ณผ ์๋ณด๋ฅผ ๊ฐํํ๋ ค๋ ์์ง์์ ๋ปํฉ๋๋ค. ์ง๋ 20๋ ๊ฐ ์ด ๋ถ์ผ์ โ์ด๋จ์โ๋ก ๋ถ๋ฆฌ๋ฉฐ ๋ฌต๋ฌตํ ๊ธธ์ ๊ฑธ์ด์จ ๊ธฐ์ ์ด ์์ต๋๋ค. ๋ฐ๋ก ํ๋ํฐ์ด(Palantir Technologies)์ ๋๋ค.
a16z๊ฐ ์ฃผ์ตํ โ์๋ฉ๋ฆฌ์นธ ๋ค์ด๋ด๋ฏธ์ฆ ์๋ฐโ์ ์ฐธ์ํ ํ๋ํฐ์ด์ CEO ์๋ ์ค ์นดํ(Alex Carp)๋ ํน์ ์ ๊ฑฐ์นจ์๋ ํ๋ฒ์ผ๋ก ์ค๋ ์ ์ธ, AI ๋ฌด๊ธฐ์ฒด๊ณ, ๊ทธ๋ฆฌ๊ณ ์ค๋ฆฌ์ฝ๋ฐธ๋ฆฌ๊ฐ ์ง๋ฉดํ ์๊ธฐ์ ๋ํด ์ฌ๋ ์๋ ํต์ฐฐ์ ๊ณต์ ํ์ต๋๋ค. 20๋ ๋์ โ๊ดด์ง๋ค์ ์งํฉ์โ ์ทจ๊ธ์ ๋ฐ์ผ๋ฉด์๋ ๋๋ด ๋ฏธ๊ตญ์ ์ ๋ต์ ์ฐ์๋ฅผ ์ ์ฆํด๋ธ ๊ทธ์ ๋ฉ์์ง๋ฅผ ์ ๋ฆฌํ์ต๋๋ค.
1. ์ฌ๊ฑด๋ ๋ฏธ๊ตญ์ ์ต์ ๋ ฅ: ๊ธฐ์ ์ด ์ ์ฅ์ ์ง๋ฐฐํ๋ค
์๋ ์ค ์นดํ๋ ์ต๊ทผ ์ค๋์์ ๋ฒ์ด์ง ์ผ๋ จ์ ๊ตฐ์ฌ ์์ ์ ์ธ๊ธํ๋ฉฐ ๋ํ๋ฅผ ์์ํ์ต๋๋ค. ๊ทธ๋ ํ๋ ์ฝํ๋์๋ ๋ฏธ๊ตญ์ **์ต์ ๋ ฅ(Deterrence)**์ด ๊ธฐ์ ์ ํ์ผ๋ก ๋ค์ ์ด์๋๊ณ ์๋ค๊ณ ๊ฐ์กฐํ์ต๋๋ค.
โ์ฐ๋ฆฌ๋ ์ง๊ธ ์ด๋ ๊ตญ๊ฐ๋ ๋ณด์ ํ์ง ๋ชปํ ๊ฐ๋ ฅํ ์ต์ ๋ ฅ์ ๋ฐํํ๊ณ ์์ต๋๋ค. ์ด๋ ๋จ์ํ ๊ตฐ๋์ ๊ท๋ชจ ๋ฌธ์ ๊ฐ ์๋๋๋ค. ์ ์์ ๋ณธ์ง์ด ๊ธฐ์ ๋ก ๋ณํํ๊ธฐ ๋๋ฌธ์ ๋๋ค.โ
๊ทธ๋ ์ 2์ฐจ ์ธ๊ณ๋์ ๋น์ ๋ฏธ๊ตญ์ด ์น๋ฆฌํ ์ ์์๋ ์๋๋ ฅ์ด ๊ธฐ์ ์ ์ฐ์์์์ ์๊ธฐ์์ผฐ์ต๋๋ค. ์ค๋๋ โ์คํผ๋ ์ด์ ๋ฏธ๋๋์ ํด๋จธ(Operation Midnight Hammer)โ์ ๊ฐ์ ์์ ์์ ๋ณด์ฌ์ค ์๋์ ์ธ ์ฑ๊ณผ๋ ์ํํธ์จ์ด, ํ๋์จ์ด, ๊ทธ๋ฆฌ๊ณ AI๊ฐ ๊ฒฐํฉ๋ ํ์ด๋ธ๋ฆฌ๋ ์ฒด๊ณ์ ์น๋ฆฌ๋ผ๋ ๊ฒ์ ๋๋ค. ์นดํ๋ ์๊ตฌ ์ฌํ๊ฐ ๋ฒ์น์ฃผ์์ ํ๋ฑ์ ์งํฅํด์ผ ํ์ง๋ง, ํ์ค ์ธ๊ณ์์๋ ๊ฒฐ๊ตญ **๊ตฐ์ฌ์ ์ฐ์(Military Superiority)**๋ง์ด ๊ฒฐ์ ์ ์ธ ํฌํ๊ถ์ ํ์ฌํ๋ค๊ณ ๋จ์ธํ์ต๋๋ค.
2. ์ค๋ฆฌ์ฝ๋ฐธ๋ฆฌ๋ฅผ ํฅํ ๊ฒฝ๊ณ : โ๊ตญ์ ํ์ ๋๋๊ฐ ๋ฌธ์์ ์ ์๋คโ
์นดํ๋ ์ค๋ฆฌ์ฝ๋ฐธ๋ฆฌ์ ์๋ฆฌํธ์ฃผ์์ ์ ์น์ ํธํฅ์ฑ์ ๋ํด ๋ ์ ๋นํ์ ์์ด์ง ์์์ต๋๋ค. ๊ทธ๋ ๊ธฐ์ ๊ธฐ์ ๋ค์ด ๊ตญ๊ฐ ์๋ณด์ ๋ ธ๋์์ ์ถ์ ์ธ๋ฉดํ ๊ฒฝ์ฐ, ๊ธฐ์ ์์ฒด๊ฐ **๊ตญ์ ํ(Nationalization)**๋๋ ๋น๊ทน์ ๋ง์ดํ ์ ์๋ค๊ณ ๊ฒฝ๊ณ ํ์ต๋๋ค.
- ํ์ดํธ์นผ๋ผ ์ค์ ๋ฌธ์ : AI๊ฐ ๊ณ ํ๋ ฅ ํ์ดํธ์นผ๋ผ์ ์ผ์๋ฆฌ๋ฅผ ๋บ๋ ๋ฐ๋ง ์ง์คํ๊ณ , ์ ์ ๊ตญ๊ฐ๋ฅผ ์งํค๋ ๊ตฐ์ธ๋ค์ ์ํ ๊ธฐ์ ๊ฐ๋ฐ์ ์ธ๋ฉดํ๋ค๋ฉด ๋์ค์ ๋ถ๋ ธ๋ฅผ ํผํ ์ ์์ต๋๋ค.
- ์ ์น์ ๊ณ ๋ฆฝ: ์ค๋ฆฌ์ฝ๋ฐธ๋ฆฌ๊ฐ ํฐ๋ก์จํ (Palo Alto) ์์์๋ง ์ธ๊ธฐ ์๋ ๊ธฐ์ ์ ๋งค๋ชฐ๋๋ค๋ฉด, ์์ฑํด D.C.์ ์ผ๋ฐ ๋ฏธ๊ตญ ์๋ฏผ๋ค๋ก๋ถํฐ ์ธ๋ฉด๋ฐ๊ฒ ๋ ๊ฒ์ ๋๋ค.
๊ทธ๋ โ์ค๋ฆฌ์ฝ๋ฐธ๋ฆฌ๊ฐ 160์ ์ง๋ฅ์ง์(IQ)๋ฅผ ๊ฐ์ก์์ง๋ ๋ชฐ๋ผ๋, ๋์ค์ ์ ์๋ฅผ ์ฝ์ง ๋ชปํ๋ค๋ฉด ๊ฒฐ๊ตญ ๊ธฐ์ ํต์ ๊ถ์ ๊ตญ๊ฐ์ ๋นผ์๊ธฐ๊ฒ ๋ ๊ฒโ์ด๋ผ๋ฉฐ, ๊ธฐ์ ๊ณผ ๊ตญ๊ฐ ์ด์ต์ ์ ๋ ฌ(Alignment)์ ๊ฐ๋ ฅํ ์ด๊ตฌํ์ต๋๋ค.
3. AI ๊ฒฝ์์ โ์ ๋ก์ฌ ๊ฒ์โ์ด๋ค
๋ง์ ์ค๋ฆฌ์ฝ๋ฐธ๋ฆฌ ๊ธฐ์ ๋ค์ด AI ๋ฐ์ ์ โ๋ชจ๋์๊ฒ ์ด๋ก์ด ๊ฒ์โ์ผ๋ก ํฌ์ฅํ์ง๋ง, ์นดํ์ ์๊ฐ์ ๋ํนํฉ๋๋ค. ๊ทธ๋ ๊ตญ๊ฐ ๊ฐ์ AI ๊ฒฝ์์ ์ฒ ์ ํ ์ ๋ก์ฌ(Zero-sum) ๊ฒ์์ผ๋ก ๊ท์ ํฉ๋๋ค.
โ์ด๋ก ์ ์ผ๋ก๋ ๋ชจ๋๊ฐ ํ๋ฑํ ์์คํ ์ด ์ข๊ฒ ์ง๋ง, ํ์ค์ ์ฐ๋ฆฌ ์๋๋ฉด ์ค๊ตญ, ํน์ ๋ฌ์์์ ๋๋ค. ์๊ตฌ ๋ฌธ๋ช ์ด ์ฐ์๋ฅผ ์ ํ์ง ๋ชปํ๋ฉด ์ฐ๋ฆฌ๊ฐ ๋๋ฆฌ๋ ์์ ์ ๊ถ๋ฆฌ๋ ์ฌ๋ผ์ง๋๋ค.โ
๊ทธ๋ ๊ฑฐ๋์ธ์ด๋ชจ๋ธ(LLM)์ ํฌํจํ ์ฒจ๋จ ๊ธฐ์ ์ด ๋จ์ํ ๊ฒฝ์ ์ ๋๊ตฌ๋ฅผ ๋์ด, ์๊ตฌ์ ๊ฐ์น๊ด์ ์ํธํ๋ ๋ฐฉํจ๊ฐ ๋์ด์ผ ํ๋ค๊ณ ์ฃผ์ฅํฉ๋๋ค. ๋ฏธ๊ตญ์ด ๊ธฐ์ ์ ๊ฒฉ์ฐจ๋ฅผ ๋ฒ๋ฆฌ์ง ๋ชปํ ๋ ๋ฅ์ณ์ฌ ๋ฌธํ์ , ๊ตฐ์ฌ์ ์๋ฐ์ ์์ ์ด์์ผ ๊ฒ์ด๋ผ๋ ๋ถ์์ ๋๋ค.
4. ํ๋ํฐ์ด์ ๋น๋ฐ ๋ณ๊ธฐ: โ์ ๊ฒฝ๋ค์์ฑโ๊ณผ ๊ฐ์ฑ ์๋ ์ธ์ฌ
ํ๋ํฐ์ด๊ฐ ์ด๋ป๊ฒ ๊ทธํ ๋ก ๋ ๋ณด์ ์ธ ์ ํ๋ค์ ๋ง๋ค์ด๋ผ ์ ์์์๊น์? ์นดํ๋ ๊ทธ ๋น๊ฒฐ๋ก **์ ๊ฒฝ๋ค์์ฑ(Neurodivergent)**์ ๊ผฝ์์ต๋๋ค. ๊ทธ๋ ๋๋ ์ฆ(Dyslexia)์ ์๊ณ ์๋ ์์ ์ ๊ฒฝํ์ ๊ณต์ ํ๋ฉฐ, ๋จ๋ค๊ณผ ๋ค๋ฅด๊ฒ ์๊ฐํ๋ ์ธ์ฌ๋ค์ ์ค์์ฑ์ ๊ฐ์กฐํ์ต๋๋ค.
- ๊ฐ๋ณ์ ์ธ ํ๋ ์ด๋ถ: ํ๋ํฐ์ด๋ ์ง์๋ค์๊ฒ ํ์ฌ์ ๋งค๋ด์ผ์ ๊ฐ์ํ์ง ์์ต๋๋ค. ๋์ ๊ฐ์๊ฐ ๊ฐ์ง ๊ณ ์ ํ ์ฌ๋ฅ์ ๊ทน๋ํํ ์ ์๋๋ก ๋์ต๋๋ค.
- ์ด์ง์ ์ธ ์ง๋จ: ๋ฏผ์ฃผ๋น์, ๊ณตํ๋น์, ์์ ๊ฐ, ๊ธฐ์ ์๊ฐ ํ๋ฐ ์์ฌ ํ ๋ก ํ๋ ๋ฌธํ๊ฐ ํ๋ํฐ์ด์ ํ์ ๋๋ค.
์นดํ๋ โ๋ฏธ๊ตญ์ ์ง์ ํ ๊ฐ์ ์ ์ฌ์, ์ข ๊ต, ๊ทธ๋ฆฌ๊ณ ์ ๊ฒฝํ์ ์ผ๋ก ๋ค๋ฅธ ๋ชจ๋ ์ด๋ค์ด ๋ชจ์ฌ ์์ ๋กญ๊ฒ ์์ ์ ํํํ ์ ์๋ค๋ ์ โ์ด๋ผ๋ฉฐ, ์ด๋ฌํ โ๋ค๋ฆโ์ ๊ธฐ์ ๋ก ์ฆํญ์ํค๋ ๊ฒ์ด ํ๋ํฐ์ด์ ์ฌ๋ช ์ด๋ผ๊ณ ์ค๋ช ํ์ต๋๋ค.
5. ๊ฒฐ๋ก : โ์ฐ๋ฆฌ ๊ตฐ์ธ์ด ๋ฌด์ฌํ ๋์์ค๊ฒ ํ๋ ๊ฒโ
๊ฐ์ฐ์ ๋ง์ง๋ง์์ ์นดํ๋ ํ๋ํฐ์ด๊ฐ ํ๋ ๊ฐ์ฅ ์ค์ํ ์ผ์ ๊ฒฐ๊ตญ **โ๋ฏธ๊ตญ ์ ์ฌ(War fighters)๋ค์ด ์ง์ผ๋ก ๋ฌด์ฌํ ๋์์ฌ ํ๋ฅ ์ ๋์ด๋ ๊ฒโ**์ด๋ผ๊ณ ๋งํ์ต๋๋ค. ๋์์ ์ ๋ค์๊ฒ๋ โ๋ฏธ๊ตญ์ ๊ณต๊ฒฉํ๋ฉด ๊ฒฐ์ฝ ๋ฌด์ฌํ ๋์๊ฐ ์ ์๋คโ๋ ๋ช ํํ ๋ฉ์์ง๋ฅผ ์ฃผ๋ ๊ฒ์ด ์ง์ ํ ํํ์ ๊ธธ์ด๋ผ๊ณ ์ญ์คํ์ต๋๋ค.
๊ทธ๋ ๊ธฐ์ ์ฐฝ์ ์๋ค์๊ฒ ์กฐ์ธํ์ต๋๋ค. โ๋๋ํ๋ค๊ณ ๋ชจ๋ ๋ถ์ผ์์ ์ ๋ฅํ ๊ฒ์ ์๋๋ค. ํ์ฅ์ ๊ตฐ์ธ๋ค์ ๋ง๋ ๊ณต๊ฐํ๊ณ , ๊ทธ๋ค์ด ์ฒํ ํ์ค์ ์ดํดํ๋ผ.โ
์๋ ์ค ์นดํ์ ๋ฉ์์ง๋ ๋ช ํํฉ๋๋ค. ๊ธฐ์ ์ ์ง๊ณต ์ํ์์ ์กด์ฌํ์ง ์์ต๋๋ค. ๊ธฐ์ ์ด ์์ ๋ฏผ์ฃผ์ฃผ์๋ผ๋ ๊ฐ์น์ ๊ฒฐํฉํ ๋, ๋น๋ก์ ์ธ์์ ์งํค๋ ๊ฐ๋ ฅํ ํ์ด ๋๋ค๋ ๊ฒ์ ๋๋ค. ํ๋ํฐ์ด๋ ์ง๋ 20๋ ๊ฐ ์ด ์ ๋ ์ ์ฆ๋ช ํด ์์ผ๋ฉฐ, ์ด์ ๊ทธ โ์ด๋จ์์ ์ฒ ํโ์ ๋ฏธ๊ตญ์ ์๋ก์ด ํ์ค์ด ๋๊ณ ์์ต๋๋ค.
๐ก ํต์ฌ ์ฉ์ด ์ ๋ฆฌ
- ์๋ฉ๋ฆฌ์นธ ๋ค์ด๋ด๋ฏธ์ฆ(American Dynamism): ๊ตญ๊ฐ ์๋ณด, ์๋์ง, ์ ์กฐ ๋ฑ ๊ตญ๊ฐ์ ๊ทผ๊ฐ์ด ๋๋ ์ฐ์ ์ ๊ธฐ์ ํ์ ์ ์ ๋ชฉํด ๋ฏธ๊ตญ์ ์ญ๋์ฑ์ ๋์ด๋ฆฌ๋ ค๋ ์์ง์.
- ์ต์ ๋ ฅ(Deterrence): ์๋๋ฐฉ์ด ๊ณต๊ฒฉํ์ ๋ ์ป๋ ์ด์ต๋ณด๋ค ํผํด๊ฐ ํจ์ฌ ํฌ๋ค๋ ๊ฒ์ ๋ณด์ฌ์ค์ผ๋ก์จ ๊ณต๊ฒฉ ์์ฌ๋ฅผ ๊บพ๋ ํ.
- ์ ๊ฒฝ๋ค์์ฑ(Neurodivergent): ์ํ, ๋๋ ์ฆ, ADHD ๋ฑ ๋์ ๊ธฐ๋ฅ์ ์ฐจ์ด๋ฅผ โ์ฅ์ โ๊ฐ ์๋ ์ธ๊ฐ์ โ๋ค์์ฑโ์ผ๋ก ๋ณด๋ ๊ด์ .
- ์ ๋ก์ฌ(Zero-sum): ํ์ชฝ์ ์ด๋์ด ๋ค๋ฅธ ์ชฝ์ ์์ค๋ก ์ด์ด์ ธ ์ ์ฒด ํฉ์ด 0์ด ๋๋ ์ํ. ์ฌ๊ธฐ์๋ ๊ธ๋ก๋ฒ ๊ธฐ์ ํจ๊ถ ๊ฒฝ์์ ์น์ดํจ์ ์๋ฏธํจ.