YouTube Digest

May 5, 2026

Listen to this issue

English

Based on โ€œThe One Rule That Separates Real AI Agents from Demos | Yutori, Abhishek Dasโ€ from EO Watch the original video

Beyond the Demo: The Uncompromising Rule for Real AI Agents

In the burgeoning landscape of artificial intelligence, the promise of autonomous agents capable of navigating the web on our behalf is a compelling vision. Yet, for many, this promise often crumbles upon the first interaction. Youโ€™ve likely encountered them: the โ€œhundred different agent productsโ€ claiming to โ€œdo anything on the web,โ€ only to falter on the simplest tasks. This pervasive unreliability is a significant pain point for Abhishek Das, co-founder and co-CEO of Yutori. Das is on a mission to redefine what an AI agent truly means, guided by a singular, uncompromising rule: โ€œIf itโ€™s not good enough to work on the first try, itโ€™s not good enough.โ€

The Pitfalls of Imperfect Automation

The current state of AI agents, while exciting in concept, often falls short in execution. Das explains the fundamental mathematical challenge inherent in complex workflows: โ€œIf we think of a 10-step, 20-step or 50-step workflow, even if the accuracy at each step is like 90%, the 10% error rate compounds very quickly.โ€ This means a seemingly robust 90% accuracy per step can quickly lead to an overall success rate that is โ€œquite lowโ€ for any multi-stage task. The technology, he asserts, is simply โ€œnot there yet to do long horizon workflows.โ€

What truly irks Das, however, is not just the technical limitation, but the industryโ€™s growing โ€œtolerance for non-determinism and low reliability in shipping products.โ€ He pushes back vehemently against this โ€œnormalization of slop,โ€ especially for agentic products. For Das, a product that requires multiple attempts to succeed isnโ€™t a product; itโ€™s a frustrating prototype. He firmly believes that โ€œif itโ€™s not good enough to work on the first try, itโ€™s not good enough.โ€ This โ€œfirst-try imperativeโ€ is the bedrock of Yutoriโ€™s philosophy.

A Vision Forged in Curiosity and Code

Abhishek Dasโ€™s journey to co-founding Yutori is rooted in a deep-seated curiosity and a rebellious spirit. Hailing from a family of doctors, his aversion to blood quickly steered him away from medicine. Instead, he found his calling in engineering, captivated by โ€œthe scientific process and methodโ€ โ€“ the whole life cycle of coming up with hypotheses, designing experiments to validate or invalidate them, drawing conclusions, and then formulating the next set of hypotheses. โ€œI think that is a very neat sort of process and method. I found that really inspiring,โ€ he recounts.

His time at IIT Rurki was transformative. Initially enrolled in electrical engineering, he soon realized his true passion lay elsewhere. โ€œThat was sort of the first major sort of rebellious streak in me,โ€ he recalls, describing how he decided to stop paying as much attention to electrical engineering and instead spent his time learning programming and building software. The vibrant programming club and culture at IIT Rurki, particularly a group called SDS Labs, became his crucible. There, a small group of โ€œ10-15 codersโ€ were constantly โ€œtinkering and hack building a ton of applications for the internet for the rest of the campus.โ€

Watching users interact with his creations โ€“ โ€œbuilding something from scratch and putting it out there and seeing how users interact with I think that was like a dopamine hit that kept like sort of fueling this.โ€ This early exposure to user-centric development, fueled by late nights and shared obsession with like-minded peers, ignited a lifelong ambition to build something of his own. He had considered it at the end of his undergrad and PhD, but โ€œfor various reasons didnโ€™t end up doing it. So, it was just a matter of time.โ€ He wanted to โ€œpush on that vision that I care about as opposed to working on somebody elseโ€™s vision.โ€

That vision is nothing less than a complete reimagining of our digital interactions. โ€œOver the last two or three decades, web browsers by and large have stayed the same,โ€ Das observes. โ€œThere is an opportunity now to reimagine what that experience looks and feels like.โ€ Yutori envisions a future where we interact with โ€œAI assistants that take actions and complete tasks on users behalf on the web,โ€ often proactively in the background. Das believes โ€œdigital agents will become a realityโ€ sooner than physical ones, allowing us to interact with the web at a โ€œslightly higher level of abstraction.โ€

This isnโ€™t about AI replacing humans, but about โ€œhumans and agents working together to overall improve productivity.โ€ By delegating โ€œall the mundane stuff to AI assistants, AI agents on our behalf,โ€ we can โ€œfocus on tasks and stuff thatโ€™s more meaningful, thatโ€™s more interesting to us.โ€ Crucially, itโ€™s also about accessibility. โ€œMy parents for example no longer have to learn every new website and how to operate it,โ€ he illustrates. โ€œIf they can just tell an assistant that this is what I want to do on this particular website and it does it for them reliably, then thatโ€™s awesome.โ€

Engineering Trust: Yutoriโ€™s Blueprint for Reliability

The path to reliable agents, however, is fraught with challenges. The very nature of agents โ€“ making a โ€œsequence of decisionsโ€ โ€“ means that even small errors can derail an entire workflow. Yutoriโ€™s solution isnโ€™t just about preventing errors, but about gracefully recovering from them. โ€œBeing able to recognize when it makes mistakes and backtrack from that to then go down a different branch is really, really important.โ€

This commitment translates into rigorous โ€œevals and guardrails.โ€ Every single production query that a user runs goes through a โ€œfairly comprehensive set of evalsโ€ that lets Yutori โ€œquickly identify where these agents are doing well versus not, which domains need more work and so on.โ€ The webโ€™s dynamic nature means agents will never be trained on every single website out there. โ€œIt is very natural to expect models to also make mistakes,โ€ Das admits. The key, he reiterates, is the agentโ€™s ability to โ€œrecognize and then backtrack and correct itself to do the right thingโ€ when faced with an unfamiliar scenario.

Dasโ€™s disdain for โ€œthe normalization of slop and non-determinism and poor reliabilityโ€ is a core tenet of Yutoriโ€™s development philosophy. They are not just building features; they are engineering trust.

The Craft of โ€œFeeling Seenโ€: Beyond Features

In a world where generative AI makes it โ€œvery easy to come up with first prototypesโ€ using coding LLMs, Das argues that โ€œthe true differentiator is in taste and craft โ€“ in how intuitive and well-designed the product is.โ€ This goes beyond simply fulfilling user requests.

Yutori adopts an 80/20 approach, prioritizing features based on user feedback but also relying heavily on intuition to identify unspoken needs. โ€œVery often there are ways to build product that users may not be asking for,โ€ he says, โ€œbut if you built itโ€ฆ then users feel seen and they feel like oh this is someone who is listening to us.โ€ He offers the example of two-factor authentication auto-fill on mobile phones. โ€œIt is hard to imagine like a user asking for that feature. But it saves a few seconds multiple times a day for people all across the world.โ€ These โ€œtiny paper cuts,โ€ he explains, are where thoughtful design truly shines, making users โ€œfeel seen.โ€

To cultivate this โ€œtasteโ€ and โ€œcraft,โ€ Yutori rigorously โ€œdogfoodsโ€ its own product. Weekly sessions are dedicated to testing new features internally, with only a fraction making it to external users. This constant internal refinement builds the โ€œmuscleโ€ for discerning what โ€œawesome or magical feels like.โ€ Itโ€™s โ€œa lot of reps to build that muscle,โ€ he notes.

Transparency: The Foundation of Trust

Das also champions the concept of โ€œproof of workโ€ in AI. Referencing his involvement in the highly cited Grad-CAM project โ€“ which helped visualize what deep learning models โ€œlook atโ€ to make predictions โ€“ he emphasizes the importance for models to convey โ€œnot just the final prediction or the final answer, but also the proof of work: what are the steps that went into coming up with this final prediction or the final answer.โ€

Yutori integrates this philosophy directly into its product. When an agent (or โ€œscoutโ€) monitors the web and generates a report, users can โ€œinspect the work that went in in behind the scenesโ€ โ€“ seeing which websites were visited and what information the agent actually looked at to pull out a piece of information. This transparency, he stresses, is โ€œvery, very important for trust building for users to be able to trust that yes this is a reliable product.โ€

For Das, trust is earned through meticulous attention to detail, both visible and invisible. โ€œIf we put attention to detail into parts of the product that users can see, then the user is more likely to trust the parts of the product that they cannot see.โ€ This philosophy underpins Yutoriโ€™s commitment to โ€œbuild something meaningful, to build something right, to bring a vision of the future to life and building like delightful and reliable product experiences. It doesnโ€™t just appear out of nowhere.โ€ It takes time, hard work, and an unwavering dedication to getting it right every single time.

In a world saturated with AI demos that promise much and deliver little, Abhishek Das and Yutori are charting a different course. Their โ€œfirst-try imperativeโ€ isnโ€™t just a product development rule; itโ€™s a foundational principle for an AI-powered future built on trust, reliability, and genuinely intuitive experiences. As AI agents increasingly weave into the fabric of our digital lives, Yutori aims to ensure they are not just smart, but consistently dependable โ€“ transforming mundane digital chores into seamless, invisible assistance.


Based on โ€œThe internal AI tool thatโ€™s transforming how Stripe designs products | Owen Williamsโ€ from How I AI Watch the original video

Stripeโ€™s AI Secret: How an Internal Tool is Redefining Product Design

In the fast-paced world of tech, where product design can often feel like a bottleneck, companies are constantly searching for ways to accelerate innovation while maintaining a high bar for quality. Stripe, a company renowned for its meticulous design and user experience, is no exception. However, instead of relying solely on off-the-shelf tools, Stripe has cultivated a powerful internal AI-driven prototyping platform called โ€œProtodashโ€ and โ€œProtodash Studio,โ€ spearheading a quiet revolution in how its products are conceptualized, built, and refined.

Owen Williams, a design manager at Stripe, found himself at the forefront of this transformation. As he observed designers grappling with generic AI tools, he recognized a critical gap: these tools, while powerful, often produced what he affectionately termed โ€œTailwind indigo slopโ€โ€”a bland, unbranded aesthetic that failed to capture Stripeโ€™s distinctive visual identity. This โ€œuncanny valleyโ€ effect broke immersion in design reviews, making it difficult to assess whether a prototype truly represented a Stripe product. His dream was simple yet ambitious: โ€œI want something thatโ€™s like vzero but fast,โ€ a tool deeply integrated with Stripeโ€™s unique design system.

The Pain of Prototyping in a Data-Rich World

The challenge of creating high-fidelity prototypes is particularly acute for data-heavy products like Stripeโ€™s dashboards. Imagine the sheer effort involved in prototyping a data dashboard in traditional tools like Figma: accounting for all interactions, filters, multiple states (zero data, a bunch of data, error states), internationalization (how does it look in Dutch?), or different business models (startup vs. enterprise). Owen highlights this pain point vividly: โ€œIt is nearly impossible to do that in Figma.โ€ Designers would spend countless hours manually duplicating components, faking data with โ€œ10 xxxโ€ placeholders, and creating static screenshots for every possible scenario.

This manual, labor-intensive process not only slowed down the design cycle but also limited the scope of exploration. Designers were often forced to present isolated JPEGs or simplified flows, making it difficult for stakeholders to grasp the full user journey or the nuances of dynamic data. The absence of interactive prototypes meant feedback often focused on static visuals rather than functional experience, leading to superficial discussions.

Protodash: Crafting Prototypes with Code and AI

Owen Williams, with his engineering background, approached this problem pragmatically. He envisioned a system that could lower the barrier to entry for designers while ensuring prototypes adhered to Stripeโ€™s exacting quality standards. The solution was Protodash, an internal prototyping tool that leverages a bundle of Cursor rules, React components, and Stripeโ€™s proprietary design system, โ€œSale.โ€

The first iteration of Protodash was a basic setup: a router, Sale components, and an integration with Stripeโ€™s internal MCP (Managed Component Platform) server. The core innovation, however, lay in the โ€œbundled rulesโ€ fed to LLMs like Cursor (or Claude Code, as mentioned in the transcript). These rules taught the AI how to interpret design requests, prioritize Stripeโ€™s design system, and avoid common pitfalls like generating generic Tailwind styles. For instance, if a user pasted a Figma link, the AI was instructed to check the Sale MCP server before writing any code. This ensured that the generated prototype was built from Stripeโ€™s established building blocks, not generic elements.

The result was transformative. โ€œIt gets you there like 90% of the way most of the time,โ€ Owen explains. Designers, who are โ€œhigh craft,โ€ could then refine the remaining 10%. The impact on design reviews was immediate: โ€œItโ€™s so convincing that Iโ€™m like, is this the real product or am I looking at like something fake?โ€ This level of realism, achieved rapidly, meant stakeholders could interact with prototypes that felt indistinguishable from the live product.

The Magic of Devboxes and Protodash Studio

Initially, Protodash ran locally, requiring designers to โ€œnpm run dev.โ€ However, Stripeโ€™s sophisticated โ€œdevboxโ€ infrastructure elevated the experience further. Devboxes are cloud-based development environments that can be spun up in minutes, pre-configured with all necessary tools and dependencies. Now, designers could simply navigate to a URL, and a devbox instance of Protodash would be ready to go, eliminating the need for local setup or managing a large monorepo. This โ€œno npm requiredโ€ magic was a game-changer, especially for designers who might not have the same powerful machines as engineers.

Building on this, Owen developed โ€œProtodash Studio,โ€ an extra layer that transformed Protodash into a browser-based โ€œvibe coding platform.โ€ This studio environment allows users to interact with the LLM directly within their browser, without needing to open a separate coding editor. Users can manage their prototypes, view a โ€œvibe prototyping feedโ€ for inspiration, and even remix othersโ€™ prototypes. The most compelling feature is the embedded LLM, where a user can simply type a requestโ€”like โ€œI want to add a new variant of my prototype where there chat a line chatโ€โ€”and the AI will build it entirely in the browser.

Owen admits much of Protodash Studio was built through an iterative process of โ€œyelling at Claude Code for 18 months,โ€ leveraging his engineering background to guide the AIโ€™s architecture. The internal nature of the tool meant it didnโ€™t need to be โ€œproduction gradeโ€ in the traditional sense, allowing for rapid experimentation and feature addition.

Transforming the Design Review: Demos, Not Memos

One of the most significant impacts of Protodash has been on the design review process. Owen passionately advocates for โ€œdemos not memos,โ€ a philosophy championed by other Stripe design leaders. He laments the โ€œdrowning in presentationsโ€ of the past five years, where reviews involved static JPEGs and Figma screenshots. With Protodash, designers can share a URL, allowing everyone in the review to interact directly with a clickable, dynamic prototype. โ€œCan I just say being in a design review where I can click things is my favorite,โ€ Owen enthuses, expressing his goal to โ€œnever want to see a slideshow again.โ€

Protodash Studio further enhances this by integrating a robust design review mode. Stakeholders can comment directly on the prototype at specific points, eliminating the need for fragmented Google Docs or lengthy email threads. These comments are then summarized by the AI, and crucially, they can be directly fed back to the AI for fixes. Owen describes the delight of being able to tell the AI, โ€œThe filter pattern isnโ€™t right here. Please add three more options,โ€ and then, after the AI implements the changes, responding to feedback providers with โ€œI fixed Katie Dโ€™s feedback for you. Hereโ€™s the receipts thread.โ€ This drastically reduces the โ€œbusy workโ€ associated with post-review follow-ups.

The PM Empowerment Revolution: A Surprising Twist

Perhaps the most unexpected and impactful outcome of Protodash is its adoption by Product Managers (PMs). Owen initially felt โ€œa little nervousโ€ seeing PMs design, wondering, โ€œwhatโ€™s going to happen?โ€ However, this apprehension quickly turned to excitement. PMs can now articulate their ideas with unprecedented clarity by building high-quality, interactive prototypes that look and feel like Stripe products.

This empowerment has several profound benefits:

  • Unblocking PMs: When a designer isnโ€™t immediately available, PMs can explore ideas and build initial concepts themselves, preventing roadblocks.
  • Better Communication: PMs can โ€œmanifest the thing that they want to,โ€ leading to more concrete and productive conversations with their dedicated designers. Instead of abstract discussions, they can point to a functional prototype.
  • Earlier User Research: With realistic prototypes, PMs can conduct user research much earlier in the product development cycle, gathering valuable feedback from actual users without the need for elaborate mockups or fully coded features.
  • Stronger Advocacy: PMs can use high-fidelity prototypes to advocate more effectively for their projects, demonstrating the potential of an idea rather than just describing it.

The tool changes the nature of conversations, shifting them from arguments about staffing to discussions about the work itself. As Owen notes, it helps PMs โ€œtalk the same languageโ€ as designers and engineers, especially for complex technical features like webhooks with multiple states.

Building a Bespoke Culture Tool

Owen strongly believes in the value of building internal tools that are precisely matched to a companyโ€™s culture and cadence. While external tools like Vzero are excellent, they canโ€™t be as deeply integrated or customized to Stripeโ€™s specific needs, design system, and review processes. Protodash isnโ€™t just a prototyping tool; itโ€™s a cultural enabler.

The platform is designed for contribution, with Owen actively encouraging designers to submit pull requests to evolve its features. This fosters a sense of ownership and allows the tool to adapt organically to the changing needs of the design and product teams. The ability to quickly integrate a Google Doc Product Definition (PD) and have the AI build a prototype from it, or to enable โ€œcrazy eightsโ€ mode for generative, out-of-the-box design exploration, highlights its flexibility. It even allows for temporary deviations from the core design system by granting the LLM access to other frameworks like Tailwind, enabling generative exploration beyond established components.

During a live demo, Owen illustrated this power by prompting Protodash to build a โ€œBlack Friday Cyber Monday dashboard for a pet store showing live sales on a chart at the top of the page, a ticker with the latest sales and top performing products, with realistic real-time data.โ€ Within moments, the AI began assembling the page, calling on Stripeโ€™s Sale MCP server to select appropriate components and structure the layout. This rapid, on-the-fly generation of complex, data-driven interfaces underscores the toolโ€™s transformative potential.

Owenโ€™s journey with Protodash at Stripe is a compelling case study in how internal tooling, powered by thoughtful AI integration, can fundamentally reshape product design and development. By addressing specific pain points, fostering collaboration, and empowering diverse teams, Stripe is not just building better products; itโ€™s building a better way to build.


Based on โ€œWhat Drives Political Violence in Americaโ€ from New York Times Podcasts Watch the original video

Beyond the Fringe: How Mainstream America Embraces Political Violence

In recent times, a disturbing drumbeat of political violence has echoed across the United States. From the murder of a state representative in Minnesota to an arson at a governorโ€™s residence, and multiple reported assassination attempts against former President Donald Trump, these acts signal a dangerous escalation. The question that persistently arises is: how did we arrive at this point, and how much worse could it possibly get?

To dissect this unsettling trend, the New York Times spoke with Professor Robert Pape of the University of Chicago, a leading expert on political violence who has advised every White House from 2001 to 2024. His insights paint a sobering picture of an America grappling with a new and profound threat: โ€œviolent populism.โ€

The Age of Violent Populism: A Threat from Within

Professor Pape asserts that the most critical fact about political violence in America today is not the rise of foreign adversaries or terrorist groups, but the alarming acceptance of violence by tens of millions of Americans across the political spectrum. This widespread normalization, he argues, poses a greater risk to American democracy than any external threat.

What exactly is โ€œviolent populismโ€? Itโ€™s a phenomenon where political violence is no longer confined to a radical fringe or a few militia groups, but finds significant social approval among a broad segment of the populace. This acceptance, Pape explains, has a two-fold dangerous effect:

  1. Empowering Volatile Individuals: For individuals already predisposed to anger or impulsivity due to their psychosocial reasons, the perceived social approval for violent acts can push them โ€œover the edge.โ€ They see a prospect of validation for actions that would otherwise be condemned.
  2. Silencing the Alarms: When millions embrace political violence, it makes it incredibly difficult for law enforcement agencies like the FBI to receive crucial tips. People are more likely to โ€œlook the other wayโ€ or discount what they see, eroding the informal intelligence network vital for preventing attacks.

This era, Pape clarifies, is distinct from a civil war, as there are no organized armies clashing. Yet, itโ€™s far beyond โ€œnasty politics as usual.โ€ It represents a spiraling cycle of violence fueled by broad social acceptance.

Alarming Statistics: A Nation Normalizing Assassination

Papeโ€™s research, conducted through 20 nationally representative surveys over the last five years by the University of Chicago Project on Security and Threats (CPOS), reveals shocking figures on the social acceptance of political violence.

For much of the Biden administration, surveys found that 10% of the body politic supported political violence to restore Donald Trump to the presidency. Further focus groups revealed an even more chilling detail: 55% of those who endorsed the โ€œuse of forceโ€ believed it meant assassination. This implies that at one point, roughly half of that 10%โ€”or 20 million American adultsโ€”found assassination acceptable for political ends.

Disturbingly, this percentage has since doubled. The latest surveys indicate that between 14% and 21% of Americans now support some form of political violence. The acceptance isnโ€™t limited to one side; when asked about the perceived acceptability of violence against figures like Charlie Kirk or the attempted assassination of Nancy Pelosi, 10% for each event found it acceptable. This translates to 20 million American adults on each side of the political divide.

In essence, roughly one in five Americans now supports some form of political violence, and this number is increasing on both sides of the spectrum.

Echoes of History, With a Dangerous Twist

While the current climate feels unprecedented, Pape points to other periods of โ€œviolent populismโ€ in American history. The 1920s saw the resurgence of the second Ku Klux Klan, growing from hundreds of thousands to 4-6 million members who actively supported political violence, not just in the South but across the nation. The 1960s also witnessed a string of political assassinations and collective violence, like the 1968 Chicago convention riots, with opinion polls from the era showing similar levels of public support for such acts.

However, a critical difference sets the current era apart: โ€œIt is pretty much the first time itโ€™s happened on both sides at the same time.โ€ The 1920s violence was predominantly from the right, and the 1960s largely from the left. Today, the escalation is bipartisan, making it uniquely perilous.

Quantifying this escalation, Papeโ€™s study of threats to members of Congress from 2001 to 2024, based on Department of Justice prosecutions, reveals a fivefold increase in threats since the first year of the Trump presidency in 2017. This surge affected both Democratic and Republican targets, indicating that Trumpโ€™s rise acted as a โ€œlightning rod on both sides.โ€ The dramatic spike has remained consistently high for an eight-year period, spanning both the Trump and Biden presidencies, suggesting that the underlying drivers are deeper than any single individual.

The Tectonic Shifts: Demographics and Wealth

At its core, this rise in political violence is not simply the impact of a president or social media, but the profound effect of social change. Pape identifies two major, interconnected social changes reshaping the United States:

  1. The Demographic Tipping Point: For the first time in its 250-year history, America is transitioning from a white-majority democracy to a white-minority democracy. In 1960, non-Hispanic whites constituted about 88% of Americans; today, that figure is 57%, projected to be 49% in about 20 years. Immigration is a primary driver of this shift, making it the โ€œnumber one lightning rod issue in America today.โ€

    • On the Right: The established group fears losing political power and declining economic prospects, leading to a desire to halt or reverse this change.
    • On the Left: The emerging groups welcome and often seek to accelerate this change, as they stand to benefit. This fundamental upheaval directly fuels support for political violence, as shown in Papeโ€™s surveys.
  2. The Concentration of Wealth: Beginning in the mid-1980s, an accelerating shift of wealth to the top 1% has occurred, regardless of which party holds power. The bottom 90% of Americans are effectively subsidizing this enormous gain for the wealthiest few. Both Republican and Democratic constituents, comprising the bottom 90%, perceive their economic prospects being โ€œclobbered even more.โ€ This unresolved economic disparity, ignored by both parties, further fuels the bipartisan nature of rising political violence.

These shifts create a sense of profound powerlessness and a fear of โ€œpolitical exclusionโ€ or โ€œlockoutโ€ for many. As groups feel their grievances cannot be addressed through conventional politics, and their very future seems existentially threatened, the appeal of more drastic measures grows.

The New Face of the Attacker: From Fringe to Mainstream

A persistent misunderstanding, Pape argues, is the profile of those who commit political violence. We often imagine them as the โ€œlosers of societyโ€โ€”marginalized, unemployed individuals with little to lose. However, Papeโ€™s analysis of the January 6th attackers, compared to right-wing political violence perpetrators from previous decades, reveals a starkly different picture.

The โ€œold profileโ€ typically consisted of unemployed individuals, often members of militia groupsโ€”the โ€œfringe.โ€ But among the hundreds arrested after January 6th, the profile was dramatically different:

  • Unemployment rates were only at the national average (around 7%).
  • Attackers included โ€œbusiness owners, doctors, lawyers, CEOsโ€โ€”insurrectionists in business suits living in nice suburbs.
  • Only about 10% were members of militia groups, a significant drop from previous findings.

This new profile indicates that the โ€œhalvesโ€โ€”educated, middle-class, and even upper-middle-class individualsโ€”are increasingly supporting political violence. Their motivation stems not from having nothing, but from the fear of losing what they have, of seeing their stature and influence decline rapidly. โ€œWe like our monsters to be villains that are far away from us,โ€ Pape notes, explaining why this shift in attacker profile is so difficult for society to acknowledge.

From Rhetoric to Reality: The Role of Leaders and Social Media

While structural issues are the โ€œfire,โ€ rhetoric acts as an accelerant. Pape highlights how top-down rhetoric from political leaders can push already predisposed individuals towards violence. He cites examples like Donald Trump famously suggesting โ€œpunching people out,โ€ Beto Oโ€™Rourke advocating to โ€œthrow that punch first,โ€ and Gavin Newsom speaking of โ€œpunching a bully back in the mouth.โ€ When authority figures normalize or even glorify aggressive action, others are more likely to follow suit.

Social media, while not the root cause, certainly โ€œadds gasoline to a fire.โ€ It provides an immediate manifestation of the millions who support political violence, creating echo chambers where such acts are cheered on and lionized. Donald Trump Jr.โ€™s social media post mocking the attack on Paul Pelosi, which generated โ€œenormous social media support, lots of laughter,โ€ is a prime example of how top-down rhetoric can interact with bottom-up amplification, fostering a sense of social approval for would-be attackers. Yet, Pape cautions against overstating social mediaโ€™s role, reminding us that previous eras of violent populism occurred long before the internet.

A Choice for a Better Future

Given these deep-seated drivers, what can be done? Pape outlines a two-stage approach:

1. Longer-Term Structural Solutions (Post-Midterms):

  • Addressing Demographic Change: While not aiming to stop the demographic shift, Pape suggests โ€œslowing it down so our political institutions can catch up.โ€ This, he clarifies, would mean adopting policies akin to the Obama-era immigration approach.
  • Reversing Wealth Inequality: Pape argues for reducing the wealth shift to the top 1%, rather than merely stopping or continuing it. He notes that both parties currently cater to their donors, failing to address this core economic disparity.

2. Immediate Action (Before the Midterms):

  • Leadersโ€™ Joint Condemnation: In the short term, the most impactful action is for political leaders to jointly condemn political violence. Pape proposes a โ€œradicalโ€ idea: Hakeem Jeffries (Democratic House leader) reaching out to Donald Trump for a joint video statement condemning violence from all sides.
    • Evidence of Impact: Pape points to the effectiveness of such rhetoric in the past. After the first Trump assassination attempt in summer 2024, President Biden delivered a speech condemning violence, and surveys showed a 20% drop in support for political violence. A similar effect was observed after a second attempt, when Biden was joined by Kamala Harris. While not the sole cause, this rhetoric โ€œcertainly didnโ€™t make anything worse.โ€

The overarching message, Pape emphasizes, is that despite all structural determinants and pressures, โ€œthe future of political violence is a choice.โ€ While tens of millions support violence, a significant counterweight exists: 75% of the U.S. population โ€œabhors political violence.โ€ This silent majority, Pape insists, must exert its agency. They need to actively voice their opposition to violence, sending letters and making calls to their representatives, demanding joint condemnation from leaders.

โ€œLeaders listen to their constituents,โ€ Pape concludes. If their inboxes are filled only with demands to โ€œhit the other side harder,โ€ thatโ€™s what theyโ€™ll do. But if the 75% who oppose violence step up and make their voices heard, a different future is possible. It is a choice that โ€œwe the peopleโ€ have to make, for the sake of the country and its future.


ํ•œ๊ตญ์–ด

โ€œThe One Rule That Separates Real AI Agents from Demos | Yutori, Abhishek Dasโ€ โ€” EO ๊ธฐ๋ฐ˜ ๊ธฐ์‚ฌ ์›๋ณธ ์˜์ƒ ๋ณด๊ธฐ

โ€์ฒซ ์‹œ๋„์— ์ž‘๋™ํ•˜์ง€ ์•Š์œผ๋ฉด, ์ถฉ๋ถ„์น˜ ์•Š๋‹คโ€ โ€“ ์‹ ๋ขฐํ•  ์ˆ˜ ์žˆ๋Š” AI ์—์ด์ „ํŠธ์˜ ์‹œ๋Œ€๋ฅผ ์—ด๋‹ค

ํ˜„์žฌ AI ์—์ด์ „ํŠธ์˜ ํ•œ๊ณ„: โ€œ์ฒซ ์‹œ๋„์— ์ž‘๋™ํ•˜์ง€ ์•Š์œผ๋ฉด ์ถฉ๋ถ„์น˜ ์•Š๋‹คโ€

์˜ค๋Š˜๋‚  ์ธ๊ณต์ง€๋Šฅ(AI) ๊ธฐ์ˆ ์€ ๋ˆˆ๋ถ€์‹œ๊ฒŒ ๋ฐœ์ „ํ•˜๊ณ  ์žˆ์ง€๋งŒ, ์›น ๊ธฐ๋ฐ˜ AI ์—์ด์ „ํŠธ(agent) ์ œํ’ˆ๋“ค์€ ์—ฌ์ „ํžˆ ์‹ ๋ขฐ์„ฑ์ด๋ผ๋Š” ๊ทผ๋ณธ์ ์ธ ๋ฌธ์ œ์— ์ง๋ฉดํ•ด ์žˆ์Šต๋‹ˆ๋‹ค. ์œ ํ† ๋ฆฌ(Yutori)์˜ ๊ณต๋™ ์ฐฝ์—…์ž์ด์ž ๊ณต๋™ CEO์ธ ์•„๋น„์‰‘ ๋‹ค์Šค(Abhishek Das)๋Š” ์ด๋Ÿฌํ•œ ํ˜„์‹ค์— ๋Œ€ํ•ด ๊ฐ•ํ•œ ๋น„ํŒ์˜ ๋ชฉ์†Œ๋ฆฌ๋ฅผ ๋ƒ…๋‹ˆ๋‹ค. ๊ทธ๋Š” ์ˆ˜๋งŽ์€ ์—์ด์ „ํŠธ ์ œํ’ˆ๋“ค์ด โ€œ์›น์—์„œ ๋ฌด์—‡์ด๋“  ํ•  ์ˆ˜ ์žˆ๋‹คโ€๊ณ  ์ฃผ์žฅํ•˜์ง€๋งŒ, ๋ง‰์ƒ ์‚ฌ์šฉํ•ด๋ณด๋ฉด ์ œ๋Œ€๋กœ ์ž‘๋™ํ•˜์ง€ ์•Š๋Š” ๊ฒฝ์šฐ๊ฐ€ ํƒœ๋ฐ˜์ด๋ผ๊ณ  ์ง€์ ํ•ฉ๋‹ˆ๋‹ค.

๋ฌธ์ œ์˜ ํ•ต์‹ฌ์€ ๋ณต์žกํ•œ ์ž‘์—… ํ๋ฆ„(workflow)์—์„œ ๋ฐœ์ƒํ•˜๋Š” ์˜ค๋ฅ˜์˜ ๋ˆ„์ (compounding error)์— ์žˆ์Šต๋‹ˆ๋‹ค. ๋‹ค์Šค์— ๋”ฐ๋ฅด๋ฉด, 10๋‹จ๊ณ„, 20๋‹จ๊ณ„, ํ˜น์€ 50๋‹จ๊ณ„๋กœ ์ด๋ฃจ์–ด์ง„ ์ž‘์—… ํ๋ฆ„์—์„œ ๊ฐ ๋‹จ๊ณ„์˜ ์ •ํ™•๋„๊ฐ€ 90%์— ๋‹ฌํ•˜๋”๋ผ๋„, 10%์˜ ์˜ค๋ฅ˜์œจ์€ ๋งค์šฐ ๋น ๋ฅด๊ฒŒ ๋ˆ„์ ๋˜์–ด ์ „์ฒด ์ž‘์—…์˜ ์„ฑ๊ณต๋ฅ ์„ ํ˜„์ €ํžˆ ๋‚ฎ์ถฅ๋‹ˆ๋‹ค. ์ด๋Š” โ€˜์žฅ๊ธฐ์ ์ธ ์ž‘์—… ํ๋ฆ„(long horizon workflows)โ€˜์„ ์ˆ˜ํ–‰ํ•˜๊ธฐ ์œ„ํ•œ ๊ธฐ์ˆ ์ด ์•„์ง ์ถฉ๋ถ„ํžˆ ๋ฐœ์ „ํ•˜์ง€ ๋ชปํ–ˆ์Œ์„ ์˜๋ฏธํ•ฉ๋‹ˆ๋‹ค.

๋‹ค์Šค CEO๋Š” ํŠนํžˆ โ€œ์ œํ’ˆ ์ถœ์‹œ ๊ณผ์ •์—์„œ ๋น„๊ฒฐ์ •๋ก ์ (non-determinism)์ด๊ณ  ๋‚ฎ์€ ์‹ ๋ขฐ์„ฑ(low reliability)์— ๋Œ€ํ•œ ๊ด€์šฉ์ด ์ผ๋ฐ˜ํ™”๋˜๊ณ  ์žˆ๋Š” ๋“ฏํ•˜๋‹คโ€๋Š” ์ ์— ๋ถˆ๋งŒ์„ ํ‘œํ•ฉ๋‹ˆ๋‹ค. โ€œ์ฒซ ์‹œ๋„์— ์ œ๋Œ€๋กœ ์ž‘๋™ํ•  ๋งŒํผ ์ถฉ๋ถ„ํžˆ ์ข‹์ง€ ์•Š๋‹ค๋ฉด, ๊ทธ๊ฒƒ์€ ์ถฉ๋ถ„ํžˆ ์ข‹์€ ์ œํ’ˆ์ด ์•„๋‹ˆ๋‹คโ€๋ผ๊ณ  ๊ทธ๋Š” ๋‹จํ˜ธํ•˜๊ฒŒ ๋งํ•ฉ๋‹ˆ๋‹ค. ๊ทธ๋Š” ์ด๋Ÿฌํ•œ ๋ฌด์ฑ…์ž„ํ•œ ์ œํ’ˆ ์ถœ์‹œ ๊ด€ํ–‰์ด ์—์ด์ „ํŠธ ์ œํ’ˆ ๋ถ„์•ผ์—์„œ๋Š” ๊ฒฐ์ฝ” ์šฉ๋‚ฉ๋  ์ˆ˜ ์—†๋‹ค๊ณ  ๊ฐ•์กฐํ•˜๋ฉฐ, ์‹ ๋ขฐ์„ฑ์ด์•ผ๋ง๋กœ ์ง„์ •ํ•œ ์ฐจ๋ณ„์ ์ด๋ผ๊ณ  ์—ญ์„คํ•ฉ๋‹ˆ๋‹ค.

์œ ํ† ๋ฆฌ(Ytori)์˜ ๋น„์ „: ์›น๊ณผ์˜ ์ƒํ˜ธ์ž‘์šฉ์„ ์žฌ์ •์˜ํ•˜๋‹ค

์œ ํ† ๋ฆฌ๋Š” ๋ฐ”๋กœ ์ด๋Ÿฌํ•œ ๋ฌธ์ œ์˜์‹์—์„œ ์ถœ๋ฐœํ–ˆ์Šต๋‹ˆ๋‹ค. ์œ ํ† ๋ฆฌ๋Š” ์‚ฌ์šฉ์ž ๋Œ€์‹  ์›น์—์„œ ํ–‰๋™์„ ์ทจํ•˜๊ณ  ์ž‘์—…์„ ์™„๋ฃŒํ•˜๋Š” ์‹ ๋ขฐํ•  ์ˆ˜ ์žˆ๋Š” ์—์ด์ „ํŠธ๋ฅผ ๊ตฌ์ถ•ํ•˜์—ฌ, ์‚ฌ์šฉ์ž๊ฐ€ ์ž์‹ ์—๊ฒŒ ๊ฐ€์žฅ ์˜๋ฏธ ์žˆ๋Š” ์ผ์— ์ง‘์ค‘ํ•  ์ˆ˜ ์žˆ๋„๋ก ๋•๋Š” ๊ฒƒ์„ ๋ชฉํ‘œ๋กœ ํ•ฉ๋‹ˆ๋‹ค. ์„ธ ๋ช…์˜ ๊ณต๋™ ์ฐฝ์—…์ž ๋ชจ๋‘ AI ์—ฐ๊ตฌ์ž ์ถœ์‹ ์ธ ์œ ํ† ๋ฆฌ๋Š” ๋‹จ์ˆœํ•œ ์—์ด์ „ํŠธ ํšŒ์‚ฌ๋ฅผ ๋„˜์–ด, ์›น๊ณผ์˜ ์ƒํ˜ธ์ž‘์šฉ ๋ฐฉ์‹์„ ๊ทผ๋ณธ์ ์œผ๋กœ ์žฌ์ •์˜ํ•˜๋ ค๋Š” ํฐ ๋น„์ „์„ ํ’ˆ๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค.

๋‹ค์Šค CEO๋Š” ์ง€๋‚œ 20~30๋…„๊ฐ„ ์›น ๋ธŒ๋ผ์šฐ์ €์˜ ํ˜•ํƒœ๊ฐ€ ํฌ๊ฒŒ ๋ณ€ํ•˜์ง€ ์•Š์•˜๋‹ค๋Š” ์ ์„ ์ง€์ ํ•˜๋ฉฐ, ์ด์ œ๋Š” ์›น ๊ฒฝํ—˜์„ ์žฌ๊ณ (reimagine)ํ•  ๊ธฐํšŒ๊ฐ€ ์™”๋‹ค๊ณ  ๋งํ•ฉ๋‹ˆ๋‹ค. ๋ฏธ๋ž˜์—๋Š” AI ๋น„์„œ์™€ ๋Œ€ํ™”ํ•˜๋ฉฐ ์›น์ƒ์˜ ์ž‘์—…์„ ์ฒ˜๋ฆฌํ•˜๊ณ , ๋งŽ์€ ์—์ด์ „ํŠธ๊ฐ€ ์‚ฌ์šฉ์ž ๋ชฐ๋ž˜ ๋ฐฑ๊ทธ๋ผ์šด๋“œ์—์„œ ๋Šฅ๋™์ ์œผ๋กœ(proactive manner) ์ž‘๋™ํ•  ๊ฒƒ์ด๋ผ๋Š” ์˜ˆ์ธก์ž…๋‹ˆ๋‹ค. ๊ทธ๋Š” ๋ฌผ๋ฆฌ์  ์—์ด์ „ํŠธ(physical agents)๋ณด๋‹ค ๋””์ง€ํ„ธ ์—์ด์ „ํŠธ(digital agents)๊ฐ€ ๋จผ์ € ํ˜„์‹คํ™”๋  ๊ฒƒ์ด๋ผ๊ณ  ๋‚ด๋‹ค๋ด…๋‹ˆ๋‹ค.

๋ฏธ๋ž˜์˜ ์›น ์ƒํ˜ธ์ž‘์šฉ์€ ํ›จ์”ฌ ๋†’์€ ์ถ”์ƒํ™” ์ˆ˜์ค€(higher level of abstraction)์—์„œ ์ด๋ฃจ์–ด์งˆ ๊ฒƒ์ž…๋‹ˆ๋‹ค. ์‚ฌ์šฉ์ž๊ฐ€ ์ง์ ‘ ๋””์ง€ํ„ธ ํ—ˆ๋“œ๋ ›์ผ(digital chores)์„ ์ˆ˜๋™์œผ๋กœ ์ฒ˜๋ฆฌํ•˜๋Š” ๋Œ€์‹ , AI ์—์ด์ „ํŠธ์—๊ฒŒ ์ผ์ƒ์ ์ด๊ณ  ๋ฐ˜๋ณต์ ์ธ ์ž‘์—…์„ ์œ„์ž„ํ•จ์œผ๋กœ์จ ๋”์šฑ ์˜๋ฏธ ์žˆ๊ณ  ํฅ๋ฏธ๋กœ์šด ์ผ์— ์ง‘์ค‘ํ•  ์ˆ˜ ์žˆ๊ฒŒ ๋ฉ๋‹ˆ๋‹ค. ์ด๋Š” ์ธ๊ฐ„๊ณผ ์—์ด์ „ํŠธ๊ฐ€ ํ˜‘๋ ฅํ•˜์—ฌ ์ „๋ฐ˜์ ์ธ ์ƒ์‚ฐ์„ฑ์„ ํ–ฅ์ƒ์‹œํ‚ค๋Š” ๋ชจ๋ธ์ด๋ฉฐ, ์—์ด์ „ํŠธ๊ฐ€ ์ธ๊ฐ„์„ ๋Œ€์ฒดํ•˜๋Š” ๊ฒƒ์ด ์•„๋‹ˆ๋ผ ์ธ๊ฐ„์˜ ์—ญ๋Ÿ‰์„ ๊ฐ•ํ™”ํ•˜๋Š” ๋ฐฉํ–ฅ์œผ๋กœ ๋‚˜์•„๊ฐˆ ๊ฒƒ์ž…๋‹ˆ๋‹ค. ๋˜ํ•œ, ์›น ์ ‘๊ทผ์„ฑ(accessibility)์„ ๋†’์ด๋Š” ๋ฐ๋„ ๊ธฐ์—ฌํ•ฉ๋‹ˆ๋‹ค. ์˜ˆ๋ฅผ ๋“ค์–ด, ๋‹ค์Šค CEO๋Š” ์ž์‹ ์˜ ๋ถ€๋ชจ๋‹˜์ด ๋งค๋ฒˆ ์ƒˆ๋กœ์šด ์›น์‚ฌ์ดํŠธ ์‚ฌ์šฉ๋ฒ•์„ ์ตํž ํ•„์š” ์—†์ด, AI ๋น„์„œ์—๊ฒŒ ์›ํ•˜๋Š” ๋ฐ”๋ฅผ ๋งํ•˜๋ฉด ์—์ด์ „ํŠธ๊ฐ€ ์‹ ๋ขฐ์„ฑ ์žˆ๊ฒŒ ์ž‘์—…์„ ์ฒ˜๋ฆฌํ•ด ์ค„ ์ˆ˜ ์žˆ๋‹ค๋ฉด โ€œ์ •๋ง ๋ฉ‹์ง„ ์ผโ€์ด ๋  ๊ฒƒ์ด๋ผ๊ณ  ์„ค๋ช…ํ•ฉ๋‹ˆ๋‹ค.

์ฐฝ์—…์ž์˜ ์—ฌ์ •: ์˜ํ•™์—์„œ AI๋กœ, ๊ทธ๋ฆฌ๊ณ  ์ง‘์ฐฉ์˜ ํž˜

์•„๋น„์‰‘ ๋‹ค์Šค CEO์˜ ๊ฐœ์ธ์ ์ธ ์—ฌ์ •์€ ๊ทธ์˜ ํ™•๊ณ ํ•œ ์ œํ’ˆ ์ฒ ํ•™์„ ์ดํ•ดํ•˜๋Š” ๋ฐ ์ค‘์š”ํ•œ ๋ฐฐ๊ฒฝ์ด ๋ฉ๋‹ˆ๋‹ค. ๊ทธ๋Š” ์˜์‚ฌ์™€ ์˜๋ฃŒ์ธ ๊ฐ€์กฑ์—์„œ ํƒœ์–ด๋‚ฌ์ง€๋งŒ, ์•„์ด๋Ÿฌ๋‹ˆํ•˜๊ฒŒ๋„ ํ”ผ๋ฅผ ๋ฌด์„œ์›Œํ•˜์—ฌ ์˜ํ•™์˜ ๊ธธ์„ ํฌ๊ธฐํ•˜๊ณ  ๊ณตํ•™์„ ์„ ํƒํ–ˆ์Šต๋‹ˆ๋‹ค. ๊ทธ๋Š” ๊ฐ€์„ค์„ ์„ธ์šฐ๊ณ , ์ด๋ฅผ ๊ฒ€์ฆํ•˜๊ฑฐ๋‚˜ ๋ฐ˜์ฆํ•˜๊ธฐ ์œ„ํ•œ ์‹คํ—˜์„ ์„ค๊ณ„ํ•˜๊ณ , ๊ฒฐ๋ก ์„ ๋„์ถœํ•˜์—ฌ ๋‹ค์Œ ๊ฐ€์„ค์„ ์„ธ์šฐ๋Š” โ€˜๊ณผํ•™์  ๊ณผ์ •(scientific process)โ€˜์— ๊นŠ์ด ๋งค๋ฃŒ๋˜์—ˆ์Šต๋‹ˆ๋‹ค.

์ธ๋„ ๊ณต๊ณผ๋Œ€ํ•™๊ต ๋ฃจ๋ฅดํ‚ค(IIT Rurki)์—์„œ ํ•™๋ถ€ ๊ณผ์ •์„ ๋ฐŸ๋˜ ์ฒซ ํ•ด, ๋‹ค์Šค๋Š” ์ข‹์€ ์„ฑ์ ์„ ๋ฐ›์•˜์ง€๋งŒ, ์ „๊ธฐ ๊ณตํ•™(ํŠนํžˆ ์ „๋ ฅ ์‹œ์Šคํ…œ ๋ถ„์•ผ)์ด ์ž์‹ ์˜ ํฅ๋ฏธ์™€ ๋งž์ง€ ์•Š๋Š”๋‹ค๋Š” ๊ฒƒ์„ ๋น ๋ฅด๊ฒŒ ๊นจ๋‹ฌ์•˜์Šต๋‹ˆ๋‹ค. ๊ทธ๋Š” ์ด๋ฅผ โ€œ์ธ์ƒ์˜ ์ค‘์š”ํ•œ ๊ฐˆ๋ฆผ๊ธธ(significant fork)โ€œ์ด์ž โ€œ์ฒซ ๋ฒˆ์งธ ๋ฐ˜ํ•ญ์ ์ธ ๊ธฐ์งˆ(rebellious streak)โ€œ์ด์—ˆ๋‹ค๊ณ  ํšŒ์ƒํ•ฉ๋‹ˆ๋‹ค. ๊ทธ๋Š” ์ „๊ธฐ ๊ณตํ•™์— ๋Œ€ํ•œ ๊ด€์‹ฌ์„ ์ค„์ด๊ณ  ๋Œ€์‹  ํ”„๋กœ๊ทธ๋ž˜๋ฐ๊ณผ ์†Œํ”„ํŠธ์›จ์–ด ๊ฐœ๋ฐœ์— ๋งŽ์€ ์‹œ๊ฐ„์„ ํ• ์• ํ•˜๊ธฐ ์‹œ์ž‘ํ–ˆ์Šต๋‹ˆ๋‹ค.

๊ทธ์—๊ฒŒ ํฐ ๋„์›€์ด ๋œ ๊ฒƒ์€ IIT ๋ฃจ๋ฅดํ‚ค์— 1314๋…„ ์ „๋ถ€ํ„ฐ ๊ฐ•๋ ฅํ•œ ํ”„๋กœ๊ทธ๋ž˜๋ฐ ํด๋Ÿฝ ๋ฌธํ™”๊ฐ€ ์žˆ์—ˆ๋‹ค๋Š” ์ ์ž…๋‹ˆ๋‹ค. ํŠนํžˆ โ€˜SDS Labsโ€™๋ผ๋Š” ๊ทธ๋ฃน์€ ๋งค๋…„ 1015๋ช…์˜ ์ฝ”๋”๋“ค์ด ๋ชจ์—ฌ ์บ ํผ์Šค ๋‚ด ์ธํ„ฐ๋„ท์„ ์œ„ํ•œ ์ˆ˜๋งŽ์€ ์• ํ”Œ๋ฆฌ์ผ€์ด์…˜์„ ๋งŒ๋“ค๊ณ  ์‹คํ—˜ํ–ˆ์Šต๋‹ˆ๋‹ค. ๋‹ค์Šค CEO๋Š” โ€œ์‚ฌ์šฉ์ž๋“ค์ด ์ง์ ‘ ๋งŒ๋“  ๊ฒƒ์„ ์‚ฌ์šฉํ•˜๊ณ , ์ƒํ˜ธ์ž‘์šฉํ•˜๋Š” ๋ชจ์Šต์„ ๋ณด๋Š” ๊ฒƒ์ด ๋„ํŒŒ๋ฏผ(dopamine)์ด ํ„ฐ์ง€๋Š” ๋“ฏํ•œ ์ฆ๊ฑฐ์›€์ด์—ˆ๋‹คโ€๊ณ  ๋งํ•ฉ๋‹ˆ๋‹ค. ๊ทธ๋Š” ๋™๋ฃŒ๋“ค๊ณผ ํ•จ๊ป˜ ๋ฐค์ƒ˜ ์ž‘์—…์„ ํ•˜๋ฉฐ ๊ธฐ๋Šฅ ์ถ”๊ฐ€์™€ ๊ฐœ์„ ์— ๋ชฐ๋‘ํ–ˆ๊ณ , ์ด๋Ÿฌํ•œ โ€˜์ง‘์ฐฉ(obsession)โ€˜์ด ๊ฐ•๋ ฅํ•œ ๋™๊ธฐ ๋ถ€์—ฌ๊ฐ€ ๋˜์—ˆ์Šต๋‹ˆ๋‹ค.

๋‹ค์Šค๋Š” ์˜ค๋žซ๋™์•ˆ ์ž์‹ ๋งŒ์˜ ์‚ฌ์—…์„ ์‹œ์ž‘ํ•˜๊ณ  ์‹ถ๋‹ค๋Š” ์—ด๋ง์„ ๊ฐ€์ง€๊ณ  ์žˆ์—ˆ์Šต๋‹ˆ๋‹ค. ํ•™๋ถ€ ์กธ์—… ํ›„์™€ ๋ฐ•์‚ฌ ํ•™์œ„ ์ทจ๋“ ํ›„์—๋„ ์—ฌ๋Ÿฌ ๋ฒˆ ์ฐฝ์—…์„ ๊ณ ๋ คํ–ˆ์ง€๋งŒ, ๊ฒฐ๊ตญ ์‹คํ–‰์— ์˜ฎ๊ธฐ์ง€ ๋ชปํ–ˆ์Šต๋‹ˆ๋‹ค. ๊ทธ๋Ÿฌ๋‚˜ ๊ทธ๋Š” โ€œ์„ธ์ƒ์—๋Š” ํ•ด๊ฒฐํ•ด์•ผ ํ•  ํฅ๋ฏธ๋กœ์šด ๋ฌธ์ œ๋“ค์ด ๋งŽ๊ณ , ๋‹ค๋ฅธ ์‚ฌ๋žŒ์˜ ๋น„์ „์ด ์•„๋‹Œ ๋‚ด๊ฐ€ ์ค‘์š”ํ•˜๊ฒŒ ์ƒ๊ฐํ•˜๋Š” ๋น„์ „์„ ์ถ”์ง„ํ•˜๊ณ  ์‹ถ์—ˆ๋‹คโ€๊ณ  ๋งํ•˜๋ฉฐ, ์œ ํ† ๋ฆฌ ์ฐฝ์—…์€ ์‹œ๊ฐ„ ๋ฌธ์ œ์˜€๋‹ค๊ณ  ์„ค๋ช…ํ•ฉ๋‹ˆ๋‹ค.

์ œํ’ˆ ์ฒ ํ•™: ์‹ ๋ขฐ, ์ง๊ด€, ๊ทธ๋ฆฌ๊ณ  โ€˜๋ณด์ด์ง€ ์•Š๋Š”โ€™ ๋…ธ๋ ฅ

์œ ํ† ๋ฆฌ๊ฐ€ ์ถ”๊ตฌํ•˜๋Š” ์ œํ’ˆ ์ฒ ํ•™์€ ์‹ ๋ขฐ์„ฑ๊ณผ ์‚ฌ์šฉ์ž ๊ฒฝํ—˜์„ ์ตœ์šฐ์„ ์œผ๋กœ ํ•ฉ๋‹ˆ๋‹ค.

์‹ ๋ขฐ์„ฑ ์ตœ์šฐ์„ : ์˜ค๋ฅ˜ ์ธ์‹๊ณผ ์—ญ์ถ”์ (Backtracking)

๋‹ค์Šค CEO๋Š” ์—์ด์ „ํŠธ์˜ ์‹ ๋ขฐ์„ฑ์„ ํ™•๋ณดํ•˜๊ธฐ ์œ„ํ•ด โ€˜์˜ค๋ฅ˜ ์ธ์‹โ€™๊ณผ โ€˜์—ญ์ถ”์ (backtracking)โ€™ ๋Šฅ๋ ฅ์ด ๋งค์šฐ ์ค‘์š”ํ•˜๋‹ค๊ณ  ๊ฐ•์กฐํ•ฉ๋‹ˆ๋‹ค. ์—์ด์ „ํŠธ๊ฐ€ ์‹ค์ˆ˜๋ฅผ ์ €์งˆ๋ €์„ ๋•Œ, ์ด๋ฅผ ์Šค์Šค๋กœ ์ธ์ง€ํ•˜๊ณ  ์ด์ „ ๋‹จ๊ณ„๋กœ ๋Œ์•„๊ฐ€ ๋‹ค๋ฅธ ๊ฒฝ๋กœ๋ฅผ ํƒ์ƒ‰ํ•˜์—ฌ ์˜ฌ๋ฐ”๋ฅธ ์ž‘์—…์„ ์ˆ˜ํ–‰ํ•˜๋Š” ๋Šฅ๋ ฅ์ด ํ•„์ˆ˜์ ์ด๋ผ๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค.

์œ ํ† ๋ฆฌ๋Š” ์ด๋ฅผ ์œ„ํ•ด โ€˜ํ‰๊ฐ€(evals)โ€˜์™€ โ€˜์•ˆ์ „์žฅ์น˜(guardrails)โ€™ ๊ตฌ์ถ•์— ๋งŽ์€ ๋…ธ๋ ฅ์„ ๊ธฐ์šธ์ž…๋‹ˆ๋‹ค. ์‚ฌ์šฉ์ž๊ฐ€ ์‹คํ–‰ํ•˜๋Š” ๋ชจ๋“  ํ”„๋กœ๋•์…˜ ์ฟผ๋ฆฌ(production query)๋Š” ํฌ๊ด„์ ์ธ ํ‰๊ฐ€ ์‹œ์Šคํ…œ์„ ๊ฑฐ์ณ ์—์ด์ „ํŠธ๊ฐ€ ์–ด๋–ค ๋ถ€๋ถ„์—์„œ ์ž˜ ์ž‘๋™ํ•˜๊ณ  ์–ด๋–ค ๋„๋ฉ”์ธ์—์„œ ๊ฐœ์„ ์ด ํ•„์š”ํ•œ์ง€ ๋น ๋ฅด๊ฒŒ ์‹๋ณ„ํ•ฉ๋‹ˆ๋‹ค.

์›น ์—์ด์ „ํŠธ์˜ ํŠน์„ฑ์ƒ ์„ธ์ƒ์˜ ๋ชจ๋“  ์›น์‚ฌ์ดํŠธ๋ฅผ ํ•™์Šต์‹œํ‚ค๋Š” ๊ฒƒ์€ ๋ถˆ๊ฐ€๋Šฅํ•ฉ๋‹ˆ๋‹ค. ์ƒˆ๋กœ์šด ์›น์‚ฌ์ดํŠธ๋Š” ๋Š์ž„์—†์ด ์ƒ๊ฒจ๋‚˜๊ณ , ๊ธฐ์กด ์›น์‚ฌ์ดํŠธ์˜ ์ˆ˜๋„ ๋ฐฉ๋Œ€ํ•˜๊ธฐ ๋•Œ๋ฌธ์ž…๋‹ˆ๋‹ค. ๋”ฐ๋ผ์„œ ์œ ํ† ๋ฆฌ๋Š” ์ œํ•œ๋œ ์›น์‚ฌ์ดํŠธ ์„ธํŠธ์—์„œ ๋ชจ๋ธ์„ ํ›ˆ๋ จํ•˜๊ณ  ๊ฐœ์„ ํ•ด๋‚˜๊ฐ‘๋‹ˆ๋‹ค. ์ธ๊ฐ„๋„ ์ƒˆ๋กœ์šด ์›น์‚ฌ์ดํŠธ์—์„œ ์‹ค์ˆ˜๋ฅผ ์ €์ง€๋ฅด๋“ฏ, ๋ชจ๋ธ ์—ญ์‹œ ์‹ค์ˆ˜๋ฅผ ํ•˜๋Š” ๊ฒƒ์€ ์ž์—ฐ์Šค๋Ÿฌ์šด ์ผ์ž…๋‹ˆ๋‹ค. ๊ทธ๋Ÿฌ๋‚˜ ์ค‘์š”ํ•œ ๊ฒƒ์€ ์‹ค์ˆ˜๋ฅผ ํ–ˆ์„ ๋•Œ, ์—์ด์ „ํŠธ๊ฐ€ ์ด๋ฅผ ์ธ์‹ํ•˜๊ณ  ์Šค์Šค๋กœ ์ˆ˜์ •ํ•˜์—ฌ ์˜ฌ๋ฐ”๋ฅธ ๋ฐฉํ–ฅ์œผ๋กœ ๋‚˜์•„๊ฐˆ ์ˆ˜ ์žˆ๋Š” ๋Šฅ๋ ฅ์ž…๋‹ˆ๋‹ค. ์ด๊ฒƒ์ด ์œ ํ† ๋ฆฌ๊ฐ€ ๋ชจ๋ธ์„ ํ›ˆ๋ จํ•˜๊ณ  ๊ตฌ์ถ•ํ•˜๋Š” ํ•ต์‹ฌ์ ์ธ ์š”์†Œ์ž…๋‹ˆ๋‹ค.

์‚ฌ์šฉ์ž ์ค‘์‹ฌ, ๊ทธ๋Ÿฌ๋‚˜ ์ง๊ด€์„ ๋”ํ•˜๋‹ค

์œ ํ† ๋ฆฌ๋Š” ์ œํ’ˆ ๊ฐœ๋ฐœ์— 80/20 ์ ‘๊ทผ ๋ฐฉ์‹์„ ์ทจํ•ฉ๋‹ˆ๋‹ค. ์ฆ‰, ์‚ฌ์šฉ์ž๊ฐ€ ์š”์ฒญํ•˜๋Š” ๊ธฐ๋Šฅ ์™ธ์—๋„ ๊ฐœ๋ฐœํŒ€์˜ โ€˜์ง๊ด€(intuition)โ€˜์„ ํ†ตํ•ด ์ž ์žฌ์ ์œผ๋กœ ์‚ฌ์šฉ์ž์—๊ฒŒ ํฐ ๊ฐ€์น˜๋ฅผ ์ œ๊ณตํ•  ์ˆ˜ ์žˆ๋Š” ๊ธฐ๋Šฅ์„ ๋ฐœ๊ตดํ•ฉ๋‹ˆ๋‹ค. ๋‹ค์Šค CEO๋Š” โ€œ์‚ฌ์šฉ์ž๋“ค์ด ๋ช…์‹œ์ ์œผ๋กœ ์š”์ฒญํ•˜์ง€ ์•Š์„ ์ˆ˜๋„ ์žˆ์ง€๋งŒ, ์ผ๋‹จ ๊ตฌํ˜„๋˜๋ฉด ์‚ฌ์šฉ์ž๋“ค์€ โ€˜์šฐ๋ฆฌ๋ฅผ ๋ณด๊ณ  ์žˆ๊ตฌ๋‚˜โ€™๋ผ๊ณ  ๋А๋ผ๊ฒŒ ๋  ๊ฒƒโ€์ด๋ผ๊ณ  ์„ค๋ช…ํ•ฉ๋‹ˆ๋‹ค.

๋Œ€ํ‘œ์ ์ธ ์˜ˆ์‹œ๋กœ iOS๋‚˜ ์•ˆ๋“œ๋กœ์ด๋“œ์˜ 2๋‹จ๊ณ„ ์ธ์ฆ(two-factor authentication) SMS ์ž๋™ ์™„์„ฑ ๊ธฐ๋Šฅ์„ ๊ผฝ์Šต๋‹ˆ๋‹ค. ์‚ฌ์šฉ์ž๊ฐ€ ์ด ๊ธฐ๋Šฅ์„ ์ง์ ‘ ์š”์ฒญํ•˜๊ธฐ๋Š” ์–ด๋ ต์ง€๋งŒ, ์ด ์ž‘์€ ๊ธฐ๋Šฅ์€ ์ „ ์„ธ๊ณ„ ์ˆ˜๋งŽ์€ ์‚ฌ๋žŒ๋“ค์˜ ์ผ์ƒ์—์„œ ํ•˜๋ฃจ์—๋„ ์—ฌ๋Ÿฌ ๋ฒˆ ๋ช‡ ์ดˆ์”ฉ ์‹œ๊ฐ„์„ ์ ˆ์•ฝํ•ด ์ค๋‹ˆ๋‹ค. ๊ทธ๋Š” ์ด๋Ÿฌํ•œ โ€˜์ž‘์€ ๋ถˆํŽธํ•จ(tiny paper cuts)โ€˜์„ ์ค„์—ฌ์ฃผ๋Š” ์„ธ์‹ฌํ•œ ๋ฐฐ๋ ค๊ฐ€ ์‚ฌ์šฉ์ž์—๊ฒŒ ํฐ ๋งŒ์กฑ๊ฐ์„ ์„ ์‚ฌํ•œ๋‹ค๊ณ  ๋งํ•ฉ๋‹ˆ๋‹ค. ๋”ฐ๋ผ์„œ ์œ ํ† ๋ฆฌ์˜ ์ œํ’ˆ ๊ฐœ๋ฐœ์€ ์‚ฌ์šฉ์ž์˜ ํ”ผ๋“œ๋ฐฑ๊ณผ ๊ฐœ๋ฐœํŒ€์˜ ์ง๊ด€์ด ์กฐํ™”๋ฅผ ์ด๋ฃจ๋Š” ๊ณผ์ •์ž…๋‹ˆ๋‹ค.

โ€˜๋ง›โ€™๊ณผ โ€˜์žฅ์ธ์ •์‹ โ€™์œผ๋กœ ์ฐจ๋ณ„ํ™”ํ•˜๋‹ค (Taste and Craft)

์ฝ”๋”ฉ LLM(coding LLM)์„ ํ™œ์šฉํ•ด ์ดˆ๊ธฐ ํ”„๋กœํ† ํƒ€์ž…์„ ๋น ๋ฅด๊ฒŒ ๋งŒ๋“œ๋Š” ๊ฒƒ์ด ์‰ฌ์›Œ์ง„ ์‹œ๋Œ€์—, ์œ ํ† ๋ฆฌ๋Š” โ€˜๋ง›(taste)โ€˜๊ณผ โ€˜์žฅ์ธ์ •์‹ (craft)โ€˜์„ ์ง„์ •ํ•œ ์ฐจ๋ณ„์ ์œผ๋กœ ์‚ผ์Šต๋‹ˆ๋‹ค. ์ œํ’ˆ์ด ์–ผ๋งˆ๋‚˜ ์ง๊ด€์ ์ด๊ณ  ์ž˜ ์„ค๊ณ„๋˜์—ˆ๋Š”์ง€๊ฐ€ ์ค‘์š”ํ•˜๋‹ค๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค.

์ด๋ฅผ ์œ„ํ•ด ์œ ํ† ๋ฆฌ ํŒ€์ด ์ค‘์š”ํ•˜๊ฒŒ ์—ฌ๊ธฐ๋Š” ๊ด€ํ–‰ ์ค‘ ํ•˜๋‚˜๋Š” โ€˜์ž์‚ฌ ์ œํ’ˆ ์‚ฌ์šฉ(dogfooding)โ€˜์ž…๋‹ˆ๋‹ค. ๋งค์ฃผ ํ•œ๋‘ ์‹œ๊ฐ„์”ฉ ์ƒˆ๋กœ์šด ๊ธฐ๋Šฅ์„ ์ง์ ‘ ์‚ฌ์šฉํ•ด๋ณด๋Š” ์‹œ๊ฐ„์„ ๊ฐ€์ง€๋ฉฐ, ํ•ญ์ƒ ์ˆ˜์‹ญ ๊ฐ€์ง€์˜ ๋‚ด๋ถ€ ์‹คํ—˜์„ ์ง„ํ–‰ํ•ฉ๋‹ˆ๋‹ค. ์ด ์ค‘ ์‹ค์ œ๋กœ ์™ธ๋ถ€ ์‚ฌ์šฉ์ž์—๊ฒŒ ์ œ๊ณต๋˜๋Š” ํ”„๋กœ๋•์…˜ ๋ฒ„์ „์œผ๋กœ ์ถœ์‹œ๋˜๋Š” ๊ฒƒ์€ ๊ทนํžˆ ์ผ๋ถ€์ž…๋‹ˆ๋‹ค. ์ด๋Ÿฌํ•œ ์ง€์†์ ์ธ ์ž์‚ฌ ์ œํ’ˆ ์‚ฌ์šฉ์€ โ€˜๋ฌด์—‡์ด ์ข‹๊ณ  ๋‚˜์œ์ง€โ€™, โ€˜์–ด๋–ค ๊ฒฝํ—˜์ด ๋†€๋ž๊ณ  ๋งˆ๋ฒ• ๊ฐ™์€์ง€โ€™์— ๋Œ€ํ•œ ํŒ€์˜ โ€˜๋ง›โ€™์„ ๋‹ค๋“ฌ๋Š” ๊ณผ์ •์ž…๋‹ˆ๋‹ค. ๋‹ค์Šค CEO๋Š” ๋‹ค๋ฅธ ๋ชจ๋“  ๊ธฐ์ˆ ๊ณผ ๋งˆ์ฐฌ๊ฐ€์ง€๋กœ, ์ด๋Ÿฌํ•œ โ€˜๊ทผ์œกโ€™์„ ํ‚ค์šฐ๊ธฐ ์œ„ํ•ด์„œ๋Š” ์ˆ˜๋งŽ์€ ๋ฐ˜๋ณต(a lot of reps)์ด ํ•„์š”ํ•˜๋‹ค๊ณ  ๊ฐ•์กฐํ•ฉ๋‹ˆ๋‹ค.

ํˆฌ๋ช…์„ฑ์œผ๋กœ ์Œ“๋Š” ์‹ ๋ขฐ: โ€˜์ž‘์—… ์ฆ๋ช…โ€™์˜ ์ค‘์š”์„ฑ

AI ๋ชจ๋ธ์˜ ๊ฒฝ์šฐ, ์ตœ์ข… ์˜ˆ์ธก์ด๋‚˜ ๋‹ต๋ณ€๋ฟ๋งŒ ์•„๋‹ˆ๋ผ โ€˜์ž‘์—… ์ฆ๋ช…(proof of work)โ€™, ์ฆ‰ ์ตœ์ข… ๊ฒฐ๊ณผ๊ฐ€ ๋„์ถœ๋˜๊ธฐ๊นŒ์ง€์˜ ๊ณผ์ •์„ ๋ณด์—ฌ์ฃผ๋Š” ๊ฒƒ์ด ์ค‘์š”ํ•ฉ๋‹ˆ๋‹ค. ๋‹ค์Šค CEO๋Š” ๊ณผ๊ฑฐ ์ž์‹ ์˜ ์—ฐ๊ตฌ ๊ฒฝํ—˜์—์„œ ๋‚˜์˜จ โ€˜๊ทธ๋ ˆ๋“œ์บ (Grad-CAM)โ€™ ํ”„๋กœ์ ํŠธ๋ฅผ ์˜ˆ๋กœ ๋“ค์–ด ์„ค๋ช…ํ•ฉ๋‹ˆ๋‹ค. ๊ทธ๋ ˆ๋“œ์บ ์€ ๋”ฅ๋Ÿฌ๋‹ ๋ชจ๋ธ์ด ์ด๋ฏธ์ง€ ๋ถ„๋ฅ˜ ์˜ˆ์ธก์„ ํ•  ๋•Œ, ์ด๋ฏธ์ง€์˜ ์–ด๋А ๋ถ€๋ถ„์„ ๋ณด๊ณ  ๊ทธ๋Ÿฌํ•œ ํŒ๋‹จ์„ ๋‚ด๋ ธ๋Š”์ง€ ์‹œ๊ฐ์ ์œผ๋กœ ๋ณด์—ฌ์ฃผ๋Š” ๊ธฐ์ˆ ์ž…๋‹ˆ๋‹ค. ๋‹น์‹œ 25์„ธ์˜€๋˜ ๋‹ค์Šค๊ฐ€ ๊ณต๋™ ์ €์ž๋กœ ์ฐธ์—ฌํ•œ ์ด ๋…ผ๋ฌธ์€ ํ•ด์„ ๊ฐ€๋Šฅ์„ฑ(interpretability) ๋ถ„์•ผ์—์„œ 2๋งŒ~3๋งŒ ํšŒ์˜ ์ธ์šฉ์„ ๊ธฐ๋กํ•˜๋ฉฐ ํฐ ๋ฐ˜ํ–ฅ์„ ์ผ์œผ์ผฐ์Šต๋‹ˆ๋‹ค.

๊ทธ๋Š” ๊ทธ๋ ˆ๋“œ์บ ์ฒ˜๋Ÿผ AI ๋ชจ๋ธ์ด โ€˜์™œโ€™ ๊ทธ๋Ÿฌํ•œ ๊ฒฐ๋ก ์— ๋„๋‹ฌํ–ˆ๋Š”์ง€ ๋ณด์—ฌ์ฃผ๋Š” ๊ฒƒ์ด ์‚ฌ์šฉ์ž ์‹ ๋ขฐ๋ฅผ ๊ตฌ์ถ•ํ•˜๋Š” ๋ฐ ํ•ต์‹ฌ์ด๋ผ๊ณ  ๋งํ•ฉ๋‹ˆ๋‹ค. ์œ ํ† ๋ฆฌ์˜ ์ œํ’ˆ์—์„œ๋„ ์ด๋Ÿฌํ•œ ์ฒ ํ•™์ด ๋ฐ˜์˜๋˜์–ด ์žˆ์Šต๋‹ˆ๋‹ค. ์‚ฌ์šฉ์ž๊ฐ€ ์›น์„ ๋ชจ๋‹ˆํ„ฐ๋งํ•˜๋„๋ก ์„ค์ •ํ•œ โ€˜์ •์ฐฐ ์—์ด์ „ํŠธ(scouts and agents)โ€˜๊ฐ€ ๋ณด๊ณ ์„œ๋ฅผ ์ƒ์„ฑํ•  ๋•Œ, ์‚ฌ์šฉ์ž ์ธํ„ฐํŽ˜์ด์Šค(UI) ๋‚ด์˜ ๋ฒ„ํŠผ์„ ํ†ตํ•ด ์—์ด์ „ํŠธ๊ฐ€ ์–ด๋–ค ์›น์‚ฌ์ดํŠธ๋ฅผ ๋ฐฉ๋ฌธํ–ˆ๊ณ , ์–ด๋–ค ์ •๋ณด๋ฅผ ๋ฐ”ํƒ•์œผ๋กœ ๋ณด๊ณ ์„œ๋ฅผ ์ž‘์„ฑํ–ˆ๋Š”์ง€ ์ง์ ‘ ํ™•์ธํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.

๋‹ค์Šค CEO๋Š” โ€œ์‚ฌ์šฉ์ž๋“ค์ด ๋ณผ ์ˆ˜ ์žˆ๋Š” ์ œํ’ˆ์˜ ๋ถ€๋ถ„์— ์„ธ์‹ฌํ•œ ์ฃผ์˜๋ฅผ ๊ธฐ์šธ์ด๋ฉด, ์‚ฌ์šฉ์ž๋Š” ๋ณด์ด์ง€ ์•Š๋Š” ์ œํ’ˆ์˜ ๋ถ€๋ถ„๋„ ๋” ์‹ ๋ขฐํ•  ๊ฐ€๋Šฅ์„ฑ์ด ๋†’๋‹คโ€๊ณ  ๋งํ•ฉ๋‹ˆ๋‹ค. ์„ธ์ƒ์˜ ๋ชจ๋“  ์œ„๋Œ€ํ•œ ๊ฒƒ๋“ค์ด ๊ฐœ์ธ์ด๋‚˜ ๊ทธ๋ฃน์˜ ์—„์ฒญ๋‚œ ๋…ธ๋ ฅ๊ณผ ์„ธ์‹ฌํ•œ ์ฃผ์˜๋ฅผ ํ†ตํ•ด ๋งŒ๋“ค์–ด์ง€๋“ฏ, ์œ ํ† ๋ฆฌ ์—ญ์‹œ ๋ชจ๋“  ์ œํ’ˆ ๊ฐœ๋ฐœ์— ์ด๋Ÿฌํ•œ ์ฒ ํ•™์œผ๋กœ ์ ‘๊ทผํ•ฉ๋‹ˆ๋‹ค. ์˜๋ฏธ ์žˆ๋Š” ๊ฒƒ์„ ๋งŒ๋“ค๊ณ , ์˜ฌ๋ฐ”๋ฅด๊ฒŒ ๋งŒ๋“ค๊ณ , ๋ฏธ๋ž˜์˜ ๋น„์ „์„ ํ˜„์‹ค๋กœ ๋งŒ๋“ค๋ฉฐ, ์ฆ๊ฒ๊ณ  ์‹ ๋ขฐํ•  ์ˆ˜ ์žˆ๋Š” ์ œํ’ˆ ๊ฒฝํ—˜์„ ๊ตฌ์ถ•ํ•˜๋Š” ๋ฐ๋Š” ์‹œ๊ฐ„๊ณผ ๋…ธ๋ ฅ์ด ํ•„์š”ํ•˜๋ฉฐ, ๊ฒฐ์ฝ” ์šฐ์—ฐํžˆ ๋‚˜ํƒ€๋‚˜๋Š” ๊ฒƒ์ด ์•„๋‹ˆ๋ผ๋Š” ๊ฒƒ์ด ๊ทธ์˜ ๋ณ€์น˜ ์•Š๋Š” ์‹ ๋…์ž…๋‹ˆ๋‹ค.


โ€œThe internal AI tool thatโ€™s transforming how Stripe designs products | Owen Williamsโ€ โ€” How I AI ๊ธฐ๋ฐ˜ ๊ธฐ์‚ฌ ์›๋ณธ ์˜์ƒ ๋ณด๊ธฐ

์ŠคํŠธ๋ผ์ดํ”„์˜ ๋””์ž์ธ ํ˜์‹ : AI ๊ธฐ๋ฐ˜ ๋‚ด๋ถ€ ํˆด โ€˜ํ”„๋กœํ† ๋Œ€์‹œโ€™๊ฐ€ ์ œํ’ˆ ๊ฐœ๋ฐœ ๋ฌธํ™”๋ฅผ ๋ฐ”๊พธ๋Š” ๋ฐฉ๋ฒ•

์ŠคํŠธ๋ผ์ดํ”„(Stripe)๋Š” ์‚ฌ์šฉ์ž ๊ฒฝํ—˜(UX)๊ณผ ๋””์ž์ธ ํ’ˆ์งˆ์— ๋Œ€ํ•œ ๋†’์€ ๊ธฐ์ค€์œผ๋กœ ์ž˜ ์•Œ๋ ค์ ธ ์žˆ์Šต๋‹ˆ๋‹ค. ํ•˜์ง€๋งŒ ์ด๋Ÿฌํ•œ ๋ช…์„ฑ์—๋„ ๋ถˆ๊ตฌํ•˜๊ณ , ์ œํ’ˆ ๋””์ž์ธ ๊ณผ์ •์—์„œ ํ”„๋กœํ† ํƒ€์ดํ•‘์€ ์—ฌ์ „ํžˆ ๋งŽ์€ ์–ด๋ ค์›€์„ ์•ˆ๊ณ  ์žˆ์—ˆ์Šต๋‹ˆ๋‹ค. ํŠนํžˆ ๋ฐ์ดํ„ฐ๊ฐ€ ๋ณต์žกํ•˜๊ฒŒ ์–ฝํžŒ ๋Œ€์‹œ๋ณด๋“œ ๊ฐ™์€ ์ œํ’ˆ์€ ๊ธฐ์กด ํˆด๋กœ๋Š” ๊ฑฐ์˜ ๊ตฌํ˜„ ๋ถˆ๊ฐ€๋Šฅ์— ๊ฐ€๊นŒ์› ์Šต๋‹ˆ๋‹ค. ์ŠคํŠธ๋ผ์ดํ”„์˜ ๋””์ž์ธ ๋งค๋‹ˆ์ €์ธ ์˜ค์›ฌ ์œŒ๋ฆฌ์—„์Šค(Owen Williams)๋Š” ์ด๋Ÿฌํ•œ ๋ฌธ์ œ๋ฅผ ํ•ด๊ฒฐํ•˜๊ธฐ ์œ„ํ•ด AI ๊ธฐ๋ฐ˜์˜ ๋‚ด๋ถ€ ํ”„๋กœํ† ํƒ€์ดํ•‘ ํˆด โ€˜ํ”„๋กœํ† ๋Œ€์‹œ(Protodash)โ€˜๋ฅผ ๊ฐœ๋ฐœํ–ˆ์Šต๋‹ˆ๋‹ค. ์ด ํˆด์€ ๋‹จ์ˆœํžˆ ๋””์ž์ธ ํ”„๋กœ์„ธ์Šค๋ฅผ ํšจ์œจํ™”ํ•˜๋Š” ๊ฒƒ์„ ๋„˜์–ด, ์ŠคํŠธ๋ผ์ดํ”„์˜ ์ œํ’ˆ ๊ฐœ๋ฐœ ๋ฌธํ™”๋ฅผ ๊ทผ๋ณธ์ ์œผ๋กœ ๋ณ€ํ™”์‹œํ‚ค๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค.

๋ฌธ์ œ ์ œ๊ธฐ: โ€œํ…Œ์ผ์œˆ๋“œ ์ธ๋””๊ณ  ์Šฌ๋กญโ€๊ณผ ํ”ผ๊ทธ๋งˆ์˜ ํ•œ๊ณ„

์˜ค์›ฌ์€ ๋””์ž์ธ ๋ฆฌ๋ทฐ์— ์ฐธ์—ฌํ•˜๋ฉฐ ๋””์ž์ด๋„ˆ๋“ค์ด ๊ฒช๋Š” ๊ณ ์ถฉ์„ ์ง์ ‘ ๋ชฉ๊ฒฉํ–ˆ์Šต๋‹ˆ๋‹ค. ๋””์ž์ด๋„ˆ๋“ค์€ Vzero์™€ ๊ฐ™์€ ์™ธ๋ถ€ AI ํˆด์„ ํ™œ์šฉํ•ด ํ”„๋กœํ† ํƒ€์ž…์„ ๋งŒ๋“ค๋ ค ํ–ˆ์ง€๋งŒ, ๊ทธ ๊ฒฐ๊ณผ๋ฌผ์€ ์ŠคํŠธ๋ผ์ดํ”„์˜ ์—„๊ฒฉํ•œ ๋””์ž์ธ ์‹œ์Šคํ…œ๊ณผ๋Š” ๋™๋–จ์–ด์ง„ โ€˜ํ…Œ์ผ์œˆ๋“œ ์ธ๋””๊ณ  ์Šฌ๋กญ(Tailwind indigo slop)โ€˜์— ๋ถˆ๊ณผํ–ˆ์Šต๋‹ˆ๋‹ค. ์ด๋Š” ์™ธ๋ถ€ AI ํˆด์ด ์ŠคํŠธ๋ผ์ดํ”„์˜ ๊ณ ์œ ํ•œ ๋””์ž์ธ ์‹œ์Šคํ…œ์„ ์ดํ•ดํ•˜์ง€ ๋ชปํ•ด ๋ฐœ์ƒํ•˜๋Š”, ์–ด๋”˜๊ฐ€ ์–ด์ƒ‰ํ•˜๊ณ  ์ผ๋ฐ˜์ ์ธ ๋А๋‚Œ์˜ ๊ฒฐ๊ณผ๋ฌผ์„ ์˜๋ฏธํ•ฉ๋‹ˆ๋‹ค. ๋‚ด๋น„๊ฒŒ์ด์…˜, ํฐํŠธ ๋“ฑ์ด ๋ฏธ๋ฌ˜ํ•˜๊ฒŒ ์–ด๊ธ‹๋‚˜ ๋ชฐ์ž…์„ ๋ฐฉํ•ดํ•˜๋Š” ๊ฒฝ์šฐ๊ฐ€ ๋งŽ์•˜์Šต๋‹ˆ๋‹ค.

๋” ํฐ ๋ฌธ์ œ๋Š” ๋ฐ์ดํ„ฐ๊ฐ€ ํ•ต์‹ฌ์ธ ์ŠคํŠธ๋ผ์ดํ”„์˜ ์ œํ’ˆ ํŠน์„ฑ์ด์—ˆ์Šต๋‹ˆ๋‹ค. ์ˆ˜๋งŽ์€ ์ƒํ˜ธ์ž‘์šฉ, ํ•„ํ„ฐ, ๋‹ค์–‘ํ•œ ๋ฐ์ดํ„ฐ ์ƒํƒœ(๋ฐ์ดํ„ฐ๊ฐ€ ์ „ํ˜€ ์—†๋Š” โ€˜์ œ๋กœ ์ƒํƒœโ€™๋ถ€ํ„ฐ ๋ฐฉ๋Œ€ํ•œ ๋ฐ์ดํ„ฐ, ํŠน์ • ๋น„์ฆˆ๋‹ˆ์Šค ๋ชจ๋ธ์— ๋”ฐ๋ฅธ ๋ฐ์ดํ„ฐ ๋ณ€ํ™”, ๊ตญ์ œํ™”์— ๋”ฐ๋ฅธ ํ…์ŠคํŠธ ๊ธธ์ด ๋ณ€ํ™” ๋“ฑ)๋ฅผ ๊ฐ€์ง„ ๋Œ€์‹œ๋ณด๋“œ๋ฅผ ํ”ผ๊ทธ๋งˆ(Figma) ๊ฐ™์€ ๊ธฐ์กด ๋””์ž์ธ ํˆด๋กœ ํ”„๋กœํ† ํƒ€์ดํ•‘ํ•˜๋Š” ๊ฒƒ์€ โ€œ๊ฑฐ์˜ ๋ถˆ๊ฐ€๋Šฅ์— ๊ฐ€๊นŒ์šดโ€ ์ผ์ด์—ˆ์Šต๋‹ˆ๋‹ค. ๊ณผ๊ฑฐ์—๋Š” โ€˜๋กœ๋ ˜ ์ž…์ˆจ(Lorem Ipsum)โ€™ ํ…์ŠคํŠธ๋‚˜ ๊ฐ€์งœ ๊ธˆ์•ก์„ ์ˆ˜๋™์œผ๋กœ ์ž…๋ ฅํ•ด์•ผ ํ–ˆ๊ณ , ๋‹ค๋‹จ๊ณ„ ํ”Œ๋กœ์šฐ๋‚˜ ์—๋Ÿฌ ์ƒํƒœ๋ฅผ ์‹œ๊ฐํ™”ํ•˜๋Š” ๊ฒƒ์€ ์—„์ฒญ๋‚œ ๋…ธ๋ ฅ์„ ์š”๊ตฌํ–ˆ์Šต๋‹ˆ๋‹ค. ์ด๋Ÿฌํ•œ ํ•œ๊ณ„๋Š” ๋””์ž์ด๋„ˆ๋“ค์ด ์‹ค์ œ ์‚ฌ์šฉ ์‹œ๋‚˜๋ฆฌ์˜ค๋ฅผ ์ถฉ๋ถ„ํžˆ ํƒ์ƒ‰ํ•˜๊ณ , ์ œํ’ˆ์˜ ํ˜„์‹ค์ ์ธ ๋ฌธ์ œ๋ฅผ ์กฐ๊ธฐ์— ๋ฐœ๊ฒฌํ•˜๋Š” ๊ฒƒ์„ ์–ด๋ ต๊ฒŒ ๋งŒ๋“ค์—ˆ์Šต๋‹ˆ๋‹ค.

๋˜ํ•œ, ๋Œ€๋ถ€๋ถ„์˜ ๋””์ž์ธ ๋ฆฌ๋ทฐ๊ฐ€ ์ •์ ์ธ JPEG ์ด๋ฏธ์ง€๋‚˜ ์Šฌ๋ผ์ด๋“œ์‡ผ ํ˜•ํƒœ๋กœ ์ง„ํ–‰๋˜๋ฉด์„œ, ์ œํ’ˆ์˜ ์‹ค์ œ ๋™์ž‘์„ ํŒŒ์•…ํ•˜๊ณ  ์‹ฌ์ธต์ ์ธ ํ”ผ๋“œ๋ฐฑ์„ ์ฃผ๊ณ ๋ฐ›๋Š” ๋ฐ ํ•œ๊ณ„๊ฐ€ ์žˆ์—ˆ์Šต๋‹ˆ๋‹ค. ์˜ค์›ฌ์€ ์ด๋Ÿฌํ•œ ๋น„ํšจ์œจ์„ฑ์„ ๊ทน๋ณตํ•˜๊ณ  ์‹ถ์—ˆ์Šต๋‹ˆ๋‹ค.

ํ•ด๊ฒฐ์ฑ…: ์˜ค์›ฌ์˜ โ€˜ํ”„๋กœํ† ๋Œ€์‹œโ€™ ํƒ„์ƒ

์—”์ง€๋‹ˆ์–ด๋ง ๋ฐฐ๊ฒฝ์„ ๊ฐ€์ง„ ์˜ค์›ฌ ์œŒ๋ฆฌ์—„์Šค๋Š” ์ด๋Ÿฌํ•œ ๋ฌธ์ œ์— ๊นŠ์ด ๊ณต๊ฐํ–ˆ์Šต๋‹ˆ๋‹ค. ๊ทธ๋Š” ๊ธฐ์ˆ ์  ์†Œ์–‘์„ ๊ฐ–์ถ˜ ๋””์ž์ด๋„ˆ๋ฅผ ์„ ํ˜ธํ•˜๋ฉฐ, โ€˜ํ„ฐ๋ฏธ๋„์„ ๋‹ค๋ฃฐ ์ค„ ์•„๋Š”โ€™ ๋””์ž์ด๋„ˆ๊ฐ€ ๊ฐ•๋ ฅํ•œ ์ŠˆํผํŒŒ์›Œ๋ฅผ ๊ฐ€์งˆ ์ˆ˜ ์žˆ๋‹ค๊ณ  ๋ฏฟ์—ˆ์Šต๋‹ˆ๋‹ค. ํ•˜์ง€๋งŒ ๋™์‹œ์— ๋””์ž์ด๋„ˆ๋“ค์ด ์›น ๊ฐœ๋ฐœ์˜ ๋ณต์žกํ•œ ๊ธฐ์ˆ  ์Šคํƒ(npm, Vit, React ๋“ฑ)์— ์ขŒ์ ˆํ•˜๋Š” ๊ฒƒ์„ ๋ณด๋ฉฐ, ๊ธฐ์ˆ  ์žฅ๋ฒฝ์„ ๋‚ฎ์ถ”๋Š” ๊ฒƒ์ด ์ค‘์š”ํ•˜๋‹ค๊ณ  ์ƒ๊ฐํ–ˆ์Šต๋‹ˆ๋‹ค. AI์˜ ๋“ฑ์žฅ์€ ์ด๋Ÿฌํ•œ ๊ทธ์˜ ์ฒ ํ•™์— ์ „ํ™˜์ ์ด ๋˜์—ˆ์Šต๋‹ˆ๋‹ค. ์ด์ œ ๋””์ž์ด๋„ˆ๋“ค์€ ๋ณต์žกํ•œ ๋ช…๋ น์–ด๋ฅผ ๋ชฐ๋ผ๋„ AI์—๊ฒŒ ์ง์ ‘ ์งˆ๋ฌธํ•˜๋ฉฐ ํ•„์š”ํ•œ ์ž‘์—…์„ ์ˆ˜ํ–‰ํ•  ์ˆ˜ ์žˆ๊ฒŒ ๋œ ๊ฒƒ์ž…๋‹ˆ๋‹ค.

์˜ค์›ฌ์€ ์ด๋Ÿฌํ•œ ๋ฐฐ๊ฒฝ์„ ๋ฐ”ํƒ•์œผ๋กœ ๋‚ด๋ถ€ ํ”„๋กœํ† ํƒ€์ดํ•‘ ํˆด์ธ โ€˜ํ”„๋กœํ† ๋Œ€์‹œ(Protodash)โ€˜๋ฅผ ๊ฐœ๋ฐœํ•˜๊ธฐ ์‹œ์ž‘ํ–ˆ์Šต๋‹ˆ๋‹ค. ์ด๋Š” ์ŠคํŠธ๋ผ์ดํ”„์˜ ๋””์ž์ธ ์‹œ์Šคํ…œ์ธ โ€˜์„ธ์ผ(Sale)โ€™ ์ปดํฌ๋„ŒํŠธ์™€ MCP(Message Control Plane) ์„œ๋ฒ„ ํ†ตํ•ฉ, ๊ทธ๋ฆฌ๊ณ  ์ปค์„œ(Cursor)์™€ ๊ฐ™์€ AI ์ฝ”๋”ฉ ํˆด์— ํŠนํ™”๋œ ๊ทœ์น™ ๋ฒˆ๋“ค๋กœ ๊ตฌ์„ฑ๋œ ๋ฆฌ์•กํŠธ(React) ๊ธฐ๋ฐ˜์˜ ์›น ์•ฑ์ด์—ˆ์Šต๋‹ˆ๋‹ค. ์ด ๊ทœ์น™ ๋ฒˆ๋“ค์€ AI์—๊ฒŒ ํ”„๋กœ์ ํŠธ ์‚ฌ์šฉ๋ฒ•, Figma ๋งํฌ ์ฒ˜๋ฆฌ ๋ฐฉ๋ฒ•, ๋””์ž์ธ ์‹œ์Šคํ…œ ์ปดํฌ๋„ŒํŠธ ํ™œ์šฉ๋ฒ•, ํ”ํ•œ ์‹ค์ˆ˜ ๋ฐฉ์ง€๋ฒ• ๋“ฑ์„ ๊ฐ€๋ฅด์ณ, ์ŠคํŠธ๋ผ์ดํ”„์˜ ๋†’์€ ๋””์ž์ธ ํ’ˆ์งˆ ๊ธฐ์ค€์„ ์ถฉ์กฑํ•˜๋Š” ํ”„๋กœํ† ํƒ€์ž…์„ 90% ์ •๋„ ์ž๋™์œผ๋กœ ์ƒ์„ฑํ•  ์ˆ˜ ์žˆ๋„๋ก ํ–ˆ์Šต๋‹ˆ๋‹ค.

๋ฐ๋ธŒ๋ฐ•์Šค(Devbox) ์ธํ”„๋ผ๋ฅผ ํ†ตํ•œ ์ ‘๊ทผ์„ฑ ํ–ฅ์ƒ

์ดˆ๊ธฐ์—๋Š” ํ”„๋กœํ† ๋Œ€์‹œ๋ฅผ ๋กœ์ปฌ์—์„œ ์‹คํ–‰ํ–ˆ์ง€๋งŒ, ์ŠคํŠธ๋ผ์ดํ”„์˜ โ€˜๋ฐ๋ธŒ๋ฐ•์Šค(Devbox)โ€™ ์ธํ”„๋ผ๋ฅผ ํ™œ์šฉํ•˜๋ฉด์„œ ์ ‘๊ทผ์„ฑ์ด ํ˜์‹ ์ ์œผ๋กœ ๊ฐœ์„ ๋˜์—ˆ์Šต๋‹ˆ๋‹ค. ๋ฐ๋ธŒ๋ฐ•์Šค๋Š” ํด๋ผ์šฐ๋“œ ๊ธฐ๋ฐ˜์˜ ๊ฐœ๋ฐœ ํ™˜๊ฒฝ์œผ๋กœ, ๋””์ž์ด๋„ˆ๋Š” ๋ณต์žกํ•œ ๋กœ์ปฌ ์„ค์ •์ด๋‚˜ npm run dev์™€ ๊ฐ™์€ ๋ช…๋ น์–ด๋ฅผ ์ง์ ‘ ์ž…๋ ฅํ•  ํ•„์š” ์—†์ด ์›น URL(์˜ˆ: go/protodash)๋งŒ์œผ๋กœ ๋ช‡ ๋ถ„ ์•ˆ์— ํ”„๋กœํ† ํƒ€์ž… ํ™˜๊ฒฝ์„ ์ƒ์„ฑํ•˜๊ณ  ์‹คํ–‰ํ•  ์ˆ˜ ์žˆ๊ฒŒ ๋˜์—ˆ์Šต๋‹ˆ๋‹ค. ์ด๋Š” ๋กœ์ปฌ ํ™˜๊ฒฝ์˜ ๋А๋ฆฐ ์†๋„๋‚˜ ๋Œ€๊ทœ๋ชจ ๋ชจ๋…ธ๋ ˆํฌ(monorepo) ๊ด€๋ฆฌ์˜ ์–ด๋ ค์›€์„ ํ•ด์†Œํ–ˆ์œผ๋ฉฐ, ์ƒ์„ฑ๋œ URL์„ ํ†ตํ•ด ๋””์ž์ธ ๋ฆฌ๋ทฐ์—์„œ ํ™”๋ฉด ๊ณต์œ  ์—†์ด๋„ ์‹ค์ œ ๋™์ž‘ํ•˜๋Š” ํ”„๋กœํ† ํƒ€์ž…์„ ๊ณต์œ ํ•˜๊ณ  ํด๋ฆญํ•˜๋ฉฐ ํ”ผ๋“œ๋ฐฑ์„ ๋ฐ›์„ ์ˆ˜ ์žˆ๊ฒŒ ํ–ˆ์Šต๋‹ˆ๋‹ค. ์˜ค์›ฌ์€ ์ด๋ฅผ โ€œ๋ฐ๋ชจ๊ฐ€ ๊ณง ๋ฉ”๋ชจ(demos not memos)โ€œ๋ผ๋Š” ์ŠคํŠธ๋ผ์ดํ”„์˜ ๋””์ž์ธ ๋ฆฌ๋”์‹ญ ์ฒ ํ•™์„ ์‹คํ˜„ํ•˜๋Š” ํ•ต์‹ฌ์ ์ธ ์š”์†Œ๋กœ ๊ผฝ์•˜์Šต๋‹ˆ๋‹ค.

๋ฐ์ดํ„ฐ ์ค‘์‹ฌ ๋””์ž์ธ์˜ ํ˜์‹ 

ํ”„๋กœํ† ๋Œ€์‹œ๋Š” ํ”ผ๊ทธ๋งˆ๋กœ๋Š” ์—„๋‘๋„ ๋‚ด์ง€ ๋ชปํ–ˆ๋˜ ๋‹ค์–‘ํ•œ ๋ฐ์ดํ„ฐ ์‹œ๋‚˜๋ฆฌ์˜ค๋ฅผ ๋น ๋ฅด๊ฒŒ ์‹œ๋ฎฌ๋ ˆ์ด์…˜ํ•  ์ˆ˜ ์žˆ๊ฒŒ ํ–ˆ์Šต๋‹ˆ๋‹ค. ์˜ˆ๋ฅผ ๋“ค์–ด, ๋ฐ์ดํ„ฐ๊ฐ€ ์ „ํ˜€ ์—†๋Š” โ€˜์ œ๋กœ ์ƒํƒœโ€™๋ถ€ํ„ฐ, ๋Œ€๊ธฐ์—…๊ณผ ์Šคํƒ€ํŠธ์—…์˜ ๋น„์ฆˆ๋‹ˆ์Šค ๋ชจ๋ธ์— ๋”ฐ๋ฅธ ๋ฐ์ดํ„ฐ ๋ณ€ํ™”, ๋ณต์žกํ•œ ๊ตญ์ œํ™”(internationalization)์— ๋”ฐ๋ฅธ ํ…์ŠคํŠธ ๊ธธ์ด ๋ณ€ํ™”, ์‹ฌ์ง€์–ด โ€˜์ง€์ €๋ถ„ํ•œ ๋ฐ์ดํ„ฐ(messy data)โ€™ ์‹œ๋‚˜๋ฆฌ์˜ค๊นŒ์ง€ ์ฆ‰๊ฐ์ ์œผ๋กœ ํ™•์ธํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค. ์ด๋Š” ๋””์ž์ด๋„ˆ๊ฐ€ ์‹ค์ œ ์‚ฌ์šฉ์ž๊ฐ€ ์ œํ’ˆ์„ ์–ด๋–ป๊ฒŒ ๊ฒฝํ—˜ํ• ์ง€ ๋” ๋นจ๋ฆฌ ์ดํ•ดํ•˜๊ณ , ๋‹ค์–‘ํ•œ ๋น„์ฆˆ๋‹ˆ์Šค ์ผ€์ด์Šค๋ฅผ ๊ณ ๋ คํ•œ ๊ฒฌ๊ณ ํ•œ ๋””์ž์ธ์„ ๋งŒ๋“ค ์ˆ˜ ์žˆ๋„๋ก ๋•์Šต๋‹ˆ๋‹ค.

์˜ค์›ฌ์€ โ€œ๋กœ๋ ˜ ์ž…์ˆจ(Lorem Ipsum) ํ…์ŠคํŠธ๋ฅผ ์ˆ˜๋™์œผ๋กœ ๋ถ™์—ฌ ๋„ฃ๊ณ  ๊ฐ€์งœ ๊ธˆ์•ก์„ ์ž…๋ ฅํ•˜๋˜ ๊ณผ๊ฑฐ์™€๋Š” ๋น„๊ตํ•  ์ˆ˜ ์—†๋Š” ํ˜์‹ โ€์ด๋ผ๊ณ  ๊ฐ•์กฐํ•ฉ๋‹ˆ๋‹ค. ์ด์ œ ๋””์ž์ด๋„ˆ๋Š” ๋‹ค๋‹จ๊ณ„ ํ”Œ๋กœ์šฐ, ์—๋Ÿฌ ์ƒํƒœ, ๋‹ค์–‘ํ•œ ์ƒํ˜ธ์ž‘์šฉ ๋“ฑ ๋ณต์žกํ•œ ์‚ฌ์šฉ์ž ์—ฌ์ •์„ ํ›จ์”ฌ ๋” ์‰ฝ๊ณ  ํ˜„์‹ค์ ์œผ๋กœ ํ”„๋กœํ† ํƒ€์ดํ•‘ํ•  ์ˆ˜ ์žˆ๊ฒŒ ๋˜์—ˆ์œผ๋ฉฐ, ์ด๋Š” ์ œํ’ˆ์˜ ๋ชจ๋“  ์š”์†Œ๊ฐ€ ์˜๋„์ ์œผ๋กœ ์„ค๊ณ„๋˜์—ˆ๋Š”์ง€ ๊ฒ€ํ† ํ•˜๋Š” ๋ฐ ํฐ ๋„์›€์ด ๋ฉ๋‹ˆ๋‹ค.

์ง„ํ™”ํ•˜๋Š” โ€˜ํ”„๋กœํ† ๋Œ€์‹œ ์ŠคํŠœ๋””์˜คโ€™: ๋ธŒ๋ผ์šฐ์ € ๊ธฐ๋ฐ˜์˜ ํ˜‘์—… ํ”Œ๋žซํผ

์˜ค์›ฌ์˜ ๊ถ๊ทน์ ์ธ ๋ชฉํ‘œ๋Š” Vzero์™€ ๊ฐ™์ด ๋ธŒ๋ผ์šฐ์ €์—์„œ ์ง์ ‘ ์ŠคํŠธ๋ผ์ดํ”„ ๋งž์ถคํ˜• ํ”„๋กœํ† ํƒ€์ž…์„ ๊ตฌ์ถ•ํ•  ์ˆ˜ ์žˆ๋Š” ํ™˜๊ฒฝ์„ ๋งŒ๋“œ๋Š” ๊ฒƒ์ด์—ˆ์Šต๋‹ˆ๋‹ค. ์ด๋ฅผ ์œ„ํ•ด ๊ทธ๋Š” โ€˜ํ”„๋กœํ† ๋Œ€์‹œ ์ŠคํŠœ๋””์˜ค(Protodash Studio)โ€˜๋ผ๋Š” ์ถ”๊ฐ€ ๋ ˆ์ด์–ด๋ฅผ ๊ฐœ๋ฐœํ–ˆ์Šต๋‹ˆ๋‹ค. ์ด๋Š” ๋ฐ๋ธŒ๋ฐ•์Šค ์œ„์—์„œ ๋™์ž‘ํ•˜๋ฉฐ, ๋””์ž์ด๋„ˆ๊ฐ€ ๋ณ„๋„์˜ ์•ฑ ์„ค์น˜ ์—†์ด ์›น ๋ธŒ๋ผ์šฐ์ €๋งŒ์œผ๋กœ ํ”„๋กœํ† ํƒ€์ž…์„ ๋งŒ๋“ค๊ณ  ๊ด€๋ฆฌํ•  ์ˆ˜ ์žˆ๊ฒŒ ํ•ฉ๋‹ˆ๋‹ค.

์ฃผ์š” ๊ธฐ๋Šฅ ๋ฐ ํ˜‘์—… ๋ฐฉ์‹

  • ๋ธŒ๋ผ์šฐ์ € ๋‚ด ์ž„๋ฒ ๋””๋“œ LLM(Large Language Model) ์—๋””ํ„ฐ: ๋””์ž์ด๋„ˆ๋Š” ์ฝ”๋“œ๋ฅผ ์ง์ ‘ ๋‹ค๋ฃจ์ง€ ์•Š๊ณ ๋„ ์ž์—ฐ์–ด๋กœ ํ”„๋กœํ† ํƒ€์ž…์„ ์ˆ˜์ •ํ•˜๊ณ  ๋‹ค์–‘ํ•œ ๋ณ€ํ˜•(variant)์„ ๋งŒ๋“ค ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค. ์˜ˆ๋ฅผ ๋“ค์–ด, โ€œ๋ผ์ธ ์ฐจํŠธ(line chart)๊ฐ€ ์žˆ๋Š” ์ƒˆ๋กœ์šด ํ”„๋กœํ† ํƒ€์ž… ๋ณ€ํ˜•์„ ์ถ”๊ฐ€ํ•ด์ค˜โ€๋ผ๊ณ  ์š”์ฒญํ•˜๋ฉด AI๊ฐ€ ์ž๋™์œผ๋กœ ์ฝ”๋“œ๋ฅผ ์ƒ์„ฑํ•˜๊ณ  ์ ์šฉํ•ฉ๋‹ˆ๋‹ค.
  • ํ”„๋กœํ† ํƒ€์ž… ํ”ผ๋“œ ๋ฐ ๋ฆฌ๋ฏน์Šค: ๋‹ค๋ฅธ ๋””์ž์ด๋„ˆ๊ฐ€ ๋งŒ๋“  ํ”„๋กœํ† ํƒ€์ž…์„ ์‰ฝ๊ฒŒ ์ฐพ์•„๋ณด๊ณ , ๊ทธ๋“ค์˜ ์ ‘๊ทผ ๋ฐฉ์‹์—์„œ ์˜๊ฐ์„ ์–ป๊ฑฐ๋‚˜ ์ฝ”๋“œ๋ฅผ โ€˜๋ฆฌ๋ฏน์Šค(remix)โ€˜ํ•˜์—ฌ ์ž์‹ ์˜ ์ž‘์—…์— ํ™œ์šฉํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค. ์ด๋Š” ์ง€์‹ ๊ณต์œ ์™€ ํ˜‘์—…์„ ์ด‰์ง„ํ•ฉ๋‹ˆ๋‹ค.
  • ๋‹ค์–‘ํ•œ ๋””์ž์ธ ์‹œ์Šคํ…œ ์ง€์›: ์ŠคํŠธ๋ผ์ดํ”„์˜ โ€˜์„ธ์ผโ€™ ๋””์ž์ธ ์‹œ์Šคํ…œ ์™ธ์—๋„ ํ…Œ์ผ์œˆ๋“œ(Tailwind)์™€ ๊ฐ™์€ ์™ธ๋ถ€ ํ”„๋ ˆ์ž„์›Œํฌ๋ฅผ LLM์— ํ—ˆ์šฉํ•˜์—ฌ, ๋””์ž์ธ ์‹œ์Šคํ…œ์˜ ์ œ์•ฝ ์—†์ด ์™„์ „ํžˆ ์ƒˆ๋กœ์šด ๊ฐœ๋…์„ ํƒ์ƒ‰ํ•  ์ˆ˜ ์žˆ๋Š” ์œ ์—ฐ์„ฑ์„ ์ œ๊ณตํ•ฉ๋‹ˆ๋‹ค.
  • AI ํ”ผ๋“œ๋ฐฑ ๋ฐ ์ž๋™ ์ˆ˜์ •: โ€˜์ฃผ์„ AI(annotate AI)โ€™ ๊ธฐ๋Šฅ์„ ํ†ตํ•ด ๋””์ž์ด๋„ˆ๋Š” ํ”„๋กœํ† ํƒ€์ž… ์บ”๋ฒ„์Šค์—์„œ ํŠน์ • ์š”์†Œ๋ฅผ ์ง์ ‘ ์„ ํƒํ•˜๊ณ , ํˆดํŒ ์ถ”๊ฐ€, ํŒจ๋”ฉ ์กฐ์ • ๋“ฑ ๊ตฌ์ฒด์ ์ธ ํ”ผ๋“œ๋ฐฑ์„ AI์— ์ „๋‹ฌํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค. AI๋Š” ์ด ํ”ผ๋“œ๋ฐฑ์„ ๋ฐ”ํƒ•์œผ๋กœ ์ฝ”๋“œ๋ฅผ ์ˆ˜์ •ํ•˜๊ณ , ์‹ฌ์ง€์–ด ์Šค์Šค๋กœ ์˜ค๋ฅ˜๋ฅผ ๊ฐ์ง€ํ•˜๊ณ  ์ˆ˜์ •ํ•˜๋Š” โ€˜์…€ํ”„ ํ…Œ์ŠคํŠธ(self-test)โ€™ ๊ธฐ๋Šฅ๊นŒ์ง€ ๊ฐ–์ถ”๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค.
  • ๋””์ž์ธ ๋ฆฌ๋ทฐ ๋ชจ๋“œ: ์ŠคํŠธ๋ผ์ดํ”„์˜ ๊ณ ์œ ํ•œ ๋””์ž์ธ ๋ฆฌ๋ทฐ ๋ฌธํ™”๋ฅผ ๋ฐ˜์˜ํ•˜์—ฌ, ์—ฌ๋Ÿฌ ์ฐธ์—ฌ์ž๊ฐ€ ํ”„๋กœํ† ํƒ€์ž…์— ์ง์ ‘ ๋Œ“๊ธ€์„ ๋‹ฌ๊ณ  ํ”ผ๋“œ๋ฐฑ์„ ๊ณต์œ ํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค. AI๋Š” ์ด ํ”ผ๋“œ๋ฐฑ์„ ์ž๋™์œผ๋กœ ์š”์•ฝํ•˜์—ฌ ์ˆ˜์ • ์ž‘์—… ๋Œ€๊ธฐ์—ด์— ์ถ”๊ฐ€ํ•˜๋ฉฐ, ๋””์ž์ด๋„ˆ๋Š” ๋ฆฌ๋ทฐ ํ›„์† ์ž‘์—…์„ ๋Œ€ํญ ์ค„์ด๊ณ , โ€˜์ˆ˜์ • ์™„๋ฃŒโ€™ ์ฆ๋ช…์„œ๊นŒ์ง€ AI์—๊ฒŒ ๋ฐ›์„ ์ˆ˜ ์žˆ์–ด ์—…๋ฌด ๋ถ€๋‹ด์„ ํฌ๊ฒŒ ์ค„์˜€์Šต๋‹ˆ๋‹ค.

๋‚ด๋ถ€ ํˆด ๊ตฌ์ถ•์˜ ๊ฐ€์น˜์™€ ๋ฌธํ™”์  ๋ณ€ํ™”

์˜ค์›ฌ์€ ํ”„๋กœํ† ๋Œ€์‹œ ์ŠคํŠœ๋””์˜ค๋ฅผ ๊ฐœ๋ฐœํ•˜๋ฉฐ โ€œ๋‚ด๋ถ€ ํˆด์€ ์ƒ์‚ฐ ๋“ฑ๊ธ‰(production grade)์ผ ํ•„์š”๊ฐ€ ์—†๋‹คโ€๋Š” ์ฒ ํ•™์„ ๊ฐ€์ง€๊ณ  ์žˆ์—ˆ์Šต๋‹ˆ๋‹ค. ์ด๋Š” ์‹คํ—˜๊ณผ ๋น ๋ฅธ ๋ฐ˜๋ณต์„ ๊ฐ€๋Šฅํ•˜๊ฒŒ ํ–ˆ๊ณ , ๋ฒ„๊ทธ๊ฐ€ ๋ฐœ์ƒํ•ด๋„ ์šฉ์ธ๋˜๋Š” ์œ ์—ฐํ•œ ๊ฐœ๋ฐœ ํ™˜๊ฒฝ์„ ์กฐ์„ฑํ–ˆ์Šต๋‹ˆ๋‹ค.

๋””์ž์ด๋„ˆ์˜ ๊ธฐ์—ฌ์™€ ๋ฌธํ™” ์ง„ํ™”

ํ”„๋กœํ† ๋Œ€์‹œ๋Š” ๋‹จ์ˆœํžˆ ํšจ์œจ์„ฑ์„ ๋†’์ด๋Š” ๊ฒƒ์„ ๋„˜์–ด, ์กฐ์ง์˜ ๋ฌธํ™”์  ๋ณ€ํ™”๋ฅผ ์ด๋Œ์–ด๋ƒˆ์Šต๋‹ˆ๋‹ค. ์˜ค์›ฌ์€ โ€œ๋ˆ„๊ตฌ๋‚˜ ๊ธฐ์—ฌํ•  ์ˆ˜ ์žˆ๋‹คโ€๋Š” ์›์น™ ์•„๋ž˜, ๋””์ž์ด๋„ˆ๋“ค์ด ์ง์ ‘ ํ’€ ๋ฆฌํ€˜์ŠคํŠธ(pull request)๋ฅผ ๋ณด๋‚ด ํˆด์„ ๊ฐœ์„ ํ•˜๋„๋ก ์žฅ๋ คํ•ฉ๋‹ˆ๋‹ค. ์ด๋Š” ๋””์ž์ธ ๋ฆฌ๋ทฐ ๋ฐฉ์‹์ด๋‚˜ ์ž‘์—… ํ๋ฆ„์— ๋Œ€ํ•œ ๋ถˆ๋งŒ์„ ํ•ด์†Œํ•˜๊ณ , ๋ฌธํ™”๋ฅผ ๋Šฅ๋™์ ์œผ๋กœ ์ง„ํ™”์‹œํ‚ค๋Š” ๊ฐ•๋ ฅํ•œ ๋™๊ธฐ๊ฐ€ ๋ฉ๋‹ˆ๋‹ค. ๋””์ž์ด๋„ˆ๋“ค์€ ์ด์ œ โ€œ๋””์ž์ธ ๋ฆฌ๋ทฐ๊ฐ€ ์ œ๋Œ€๋กœ ์ž‘๋™ํ•˜์ง€ ์•Š์•„. ๋ฌด์—‡์ด ํ•„์š”ํ• ๊นŒ?โ€๋ผ๋Š” ์งˆ๋ฌธ์— ๋Œ€ํ•ด ์ง์ ‘ ์†”๋ฃจ์…˜์„ ๊ตฌํ˜„ํ•  ์ˆ˜ ์žˆ๊ฒŒ ๋œ ๊ฒƒ์ž…๋‹ˆ๋‹ค.

PM(Product Manager)์˜ ์—ญ๋Ÿ‰ ๊ฐ•ํ™”

๊ฐ€์žฅ ํฅ๋ฏธ๋กœ์šด ๋ณ€ํ™” ์ค‘ ํ•˜๋‚˜๋Š” PM(Product Manager)์˜ ์—ญํ•  ๋ณ€ํ™”์ž…๋‹ˆ๋‹ค. ์˜ค์›ฌ์€ ์ฒ˜์Œ์—๋Š” PM์ด ๋””์ž์ธํ•˜๋Š” ๊ฒƒ์— ๋Œ€ํ•ด โ€œ์•ฝ๊ฐ„ ๋ถˆ์•ˆ๊ฐโ€์„ ๋А๊ผˆ์ง€๋งŒ, ์ด์ œ๋Š” ์ด๋ฅผ ๊ธ์ •์ ์œผ๋กœ ํ‰๊ฐ€ํ•ฉ๋‹ˆ๋‹ค. PM์€ ๋””์ž์ด๋„ˆ์˜ ๋„์›€ ์—†์ด๋„ ์ŠคํŠธ๋ผ์ดํ”„ ๊ธฐ์ค€์— ๋งž๋Š” ๊ณ ํ’ˆ์งˆ ํ”„๋กœํ† ํƒ€์ž…์„ ๋งŒ๋“ค ์ˆ˜ ์žˆ๊ฒŒ ๋˜์—ˆ๊ณ , ์ด๋Š” PM ์Šค์Šค๋กœ๊ฐ€ ์•„์ด๋””์–ด๋ฅผ ํƒ์ƒ‰ํ•˜๊ณ  ๊ตฌ์ฒดํ™”ํ•˜๋Š” ๋ฐ ํฐ ๋„์›€์ด ๋ฉ๋‹ˆ๋‹ค.

  • ์†Œํ†ต ๊ฐœ์„ : PM์€ ์ž์‹ ์˜ ์•„์ด๋””์–ด๋ฅผ ์‹œ๊ฐ์ ์œผ๋กœ ๊ตฌํ˜„ํ•จ์œผ๋กœ์จ ๋””์ž์ด๋„ˆ์™€ ๋” ํšจ๊ณผ์ ์œผ๋กœ ์†Œํ†ตํ•  ์ˆ˜ ์žˆ๊ฒŒ ๋ฉ๋‹ˆ๋‹ค.
  • UXR(User Experience Research) ์กฐ๊ธฐ ์ง„ํ–‰: ์ •์ ์ธ ๊ทธ๋ฆผ์ด ์•„๋‹Œ ์‹ค์ œ ๋™์ž‘ํ•˜๋Š” ํ”„๋กœํ† ํƒ€์ž…์œผ๋กœ ์‚ฌ์šฉ์ž ํ…Œ์ŠคํŠธ๋ฅผ ์กฐ๊ธฐ์— ์ง„ํ–‰ํ•˜์—ฌ, ์ œํ’ˆ ์•„์ด๋””์–ด๋ฅผ ๊ฒ€์ฆํ•˜๊ณ  ๋ถˆํ•„์š”ํ•œ ๊ฐœ๋ฐœ ๋ฆฌ์Šคํฌ๋ฅผ ์ค„์ผ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
  • ํ”„๋กœ์ ํŠธ ์˜นํ˜ธ: ๊ณ ํ’ˆ์งˆ ํ”„๋กœํ† ํƒ€์ž…์€ PM์ด ์ž์‹ ์˜ ํ”„๋กœ์ ํŠธ์— ํ•„์š”ํ•œ ๋ฆฌ์†Œ์Šค(์˜ˆ: ๋””์ž์ด๋„ˆ ์ธ๋ ฅ)๋ฅผ ๋” ์„ค๋“๋ ฅ ์žˆ๊ฒŒ ์˜นํ˜ธํ•˜๊ณ  ํ™•๋ณดํ•˜๋Š” ๋ฐ ๊ธฐ์—ฌํ•ฉ๋‹ˆ๋‹ค.
  • ๋ณต์žกํ•œ ๊ฐœ๋… ์‹œ๊ฐํ™”: ๊ฐœ๋ฐœ์ž ๊ฒฝํ—˜(Developer Experience) ๋ถ„์•ผ์˜ ์›นํ›…(webhooks)์ฒ˜๋Ÿผ ๋ณต์žกํ•œ ๊ธฐ์ˆ ์  ๊ฐœ๋…๋„ ๋‹ค์–‘ํ•œ ์ƒํƒœ์™€ ํ”Œ๋กœ์šฐ๋ฅผ ํ†ตํ•ด ์‹œ๊ฐ์ ์œผ๋กœ ๋ช…ํ™•ํ•˜๊ฒŒ ์„ค๋ช…ํ•  ์ˆ˜ ์žˆ๊ฒŒ ๋˜์—ˆ์Šต๋‹ˆ๋‹ค.

๊ฒฐ๋ก : ์ŠคํŠธ๋ผ์ดํ”„์˜ ๋ฏธ๋ž˜ ๋””์ž์ธ, AI์™€ ๋‚ด๋ถ€ ํˆด์ด ์ด๋„๋Š” ํ˜์‹ 

์ŠคํŠธ๋ผ์ดํ”„์˜ โ€˜ํ”„๋กœํ† ๋Œ€์‹œโ€™ ์‚ฌ๋ก€๋Š” AI ์‹œ๋Œ€์— ๋‚ด๋ถ€ ํˆด์ด ๋‹จ์ˆœํ•œ ํšจ์œจ์„ฑ ์ฆ๋Œ€๋ฅผ ๋„˜์–ด, ์กฐ์ง์˜ ํ˜‘์—… ๋ฐฉ์‹, ๋ฌธํ™”, ๊ทธ๋ฆฌ๊ณ  ๊ถ๊ทน์ ์œผ๋กœ ์ œํ’ˆ ๊ฐœ๋ฐœ์˜ ๋ณธ์งˆ์„ ์–ด๋–ป๊ฒŒ ๋ณ€ํ™”์‹œํ‚ฌ ์ˆ˜ ์žˆ๋Š”์ง€ ๋ณด์—ฌ์ฃผ๋Š” ๊ฐ•๋ ฅํ•œ ์ฆ๊ฑฐ์ž…๋‹ˆ๋‹ค. ์™ธ๋ถ€ ์†”๋ฃจ์…˜์ด ์•„๋‹Œ, ํŒ€์˜ ๊ณ ์œ ํ•œ ๋ฌธํ™”์™€ ์ž‘์—… ๋ฐฉ์‹์— ์ •ํ™•ํžˆ ๋งž์ถฐ์ง„ ํˆด์„ ๊ตฌ์ถ•ํ•จ์œผ๋กœ์จ ์ŠคํŠธ๋ผ์ดํ”„๋Š” ๋” ๋†’์€ ํ’ˆ์งˆ๊ณผ ํšจ์œจ์„ฑ์„ ๋‹ฌ์„ฑํ•˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค.

์˜ค์›ฌ ์œŒ๋ฆฌ์—„์Šค๋Š” ํ”„๋กœํ† ๋Œ€์‹œ๋ฅผ ํ†ตํ•ด ๋””์ž์ด๋„ˆ์™€ PM์ด ๊ธฐ์ˆ ์  ์žฅ๋ฒฝ ์—†์ด ์ฐฝ์˜์ ์ธ ์•„์ด๋””์–ด๋ฅผ ์ž์œ ๋กญ๊ฒŒ ํƒ์ƒ‰ํ•˜๊ณ , ์ œํ’ˆ์„ ๋” ํ˜„์‹ค์ ์œผ๋กœ ์‹œ๋ฎฌ๋ ˆ์ด์…˜ํ•˜๋ฉฐ, ํ˜‘์—…์˜ ์งˆ์„ ๋†’์ด๋Š” ์ƒˆ๋กœ์šด ์‹œ๋Œ€๋ฅผ ์—ด์—ˆ์Šต๋‹ˆ๋‹ค. ์ด๋Š” ์ŠคํŠธ๋ผ์ดํ”„๊ฐ€ ์•ž์œผ๋กœ๋„ ํ˜์‹ ์ ์ธ ์ œํ’ˆ์„ ์ง€์†์ ์œผ๋กœ ๋งŒ๋“ค์–ด๋‚ผ ์ˆ˜ ์žˆ๋Š” ๊ฐ•๋ ฅํ•œ ๊ธฐ๋ฐ˜์ด ๋  ๊ฒƒ์ž…๋‹ˆ๋‹ค.

์ด ์—ํ”ผ์†Œ๋“œ๋Š” ์…€๋ฆฌ๊ณ (Celigo)์™€ ์ปค์„œ(Cursor)์˜ ํ›„์›์œผ๋กœ ์ œ์ž‘๋˜์—ˆ์Šต๋‹ˆ๋‹ค.


โ€œWhat Drives Political Violence in Americaโ€ โ€” New York Times Podcasts ๊ธฐ๋ฐ˜ ๊ธฐ์‚ฌ ์›๋ณธ ์˜์ƒ ๋ณด๊ธฐ

๋ฏธ๊ตญ์„ ๋’คํ”๋“œ๋Š” โ€˜ํญ๋ ฅ์  ํฌํ“ฐ๋ฆฌ์ฆ˜โ€™: ์ˆจ๊ฒจ์ง„ ๋‘ ๊ฐ€์ง€ ์›์ธ๊ณผ ํ•ด๊ฒฐ์ฑ…

์ตœ๊ทผ ๋ฏธ๊ตญ์€ ์ •์น˜์  ์ธ๋ฌผ๋“ค์„ ํ–ฅํ•œ ํญ๋ ฅ ํ–‰์œ„์˜ ๋ฌผ๊ฒฐ์— ํœฉ์‹ธ์—ฌ ์žˆ์Šต๋‹ˆ๋‹ค. ๋ฏธ๋„ค์†Œํƒ€์—์„œ๋Š” ์ฃผ ์˜์› ๋ถ€๋ถ€๊ฐ€ โ€œ์ •์น˜์  ๋™๊ธฐ๊ฐ€ ์žˆ๋Š” ๊ณต๊ฒฉโ€์œผ๋กœ ์‚ดํ•ด๋‹นํ–ˆ๊ณ , ํŽœ์‹ค๋ฒ ์ด๋‹ˆ์•„ ์ฃผ์ง€์‚ฌ ๊ด€์ €์—๋Š” ๋ฐฉํ™”๊ฐ€ ๋ฐœ์ƒํ–ˆ์œผ๋ฉฐ, ๋„๋„๋“œ ํŠธ๋Ÿผํ”„ ์ „ ๋Œ€ํ†ต๋ น์— ๋Œ€ํ•œ ์„ธ ์ฐจ๋ก€์˜ ์•”์‚ด ์‹œ๋„๊นŒ์ง€ ์ด์–ด์กŒ์Šต๋‹ˆ๋‹ค. ์ด์ฒ˜๋Ÿผ ์ถฉ๊ฒฉ์ ์ธ ์‚ฌ๊ฑด๋“ค์ด ์—ฐ์ด์–ด ํ„ฐ์ง€๋ฉด์„œ, โ€œ์–ด์ฉŒ๋‹ค ์—ฌ๊ธฐ๊นŒ์ง€ ์™”๋Š”๊ฐ€, ๊ทธ๋ฆฌ๊ณ  ์ƒํ™ฉ์€ ์–ผ๋งˆ๋‚˜ ๋” ๋‚˜๋น ์งˆ ์ˆ˜ ์žˆ๋Š”๊ฐ€?โ€๋ผ๋Š” ์งˆ๋ฌธ์ด ๋ฏธ๊ตญ ์‚ฌํšŒ ์ „๋ฐ˜์— ๊ฑธ์ณ ์ œ๊ธฐ๋˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค.

์ด๋Ÿฌํ•œ ๋ฌผ์Œ์— ๋‹ตํ•˜๊ธฐ ์œ„ํ•ด, ๋‰ด์š•ํƒ€์ž„์Šค ํŒŸ์บ์ŠคํŠธ โ€˜๋” ๋ฐ์ผ๋ฆฌ(The Daily)โ€˜๋Š” ์ˆ˜์‹ญ ๋…„๊ฐ„ ์ •์น˜ ํญ๋ ฅ ๋ถ„์•ผ์˜ ์ตœ๊ณ  ์ „๋ฌธ๊ฐ€๋กœ ํ™œ๋™ํ•˜๋ฉฐ 2001๋…„๋ถ€ํ„ฐ 2024๋…„๊นŒ์ง€ ๋ชจ๋“  ๋ฐฑ์•…๊ด€์— ์ž๋ฌธํ•ด์˜จ ์‹œ์นด๊ณ ๋Œ€ํ•™๊ต ๋กœ๋ฒ„ํŠธ ํŽ˜์ดํ”„(Robert Pape) ๊ต์ˆ˜๋ฅผ ์ธํ„ฐ๋ทฐํ–ˆ์Šต๋‹ˆ๋‹ค. ํŽ˜์ดํ”„ ๊ต์ˆ˜๋Š” ์ขŒ์šฐ๋ฅผ ๋ง‰๋ก ํ•˜๊ณ  ํญ๋ ฅ์ด ์ฆ๊ฐ€ํ•˜๋Š” ์ด์œ ์™€ ์ด๋ฅผ ๋ฉˆ์ถ”๊ธฐ ์œ„ํ•œ ๋ฐฉ์•ˆ์— ๋Œ€ํ•ด ์‹ฌ์ธต์ ์ธ ๋ถ„์„์„ ๋‚ด๋†“์•˜์Šต๋‹ˆ๋‹ค.

โ€ํญ๋ ฅ์  ํฌํ“ฐ๋ฆฌ์ฆ˜(Violent Populism)โ€œ์˜ ์‹œ๋Œ€

ํŽ˜์ดํ”„ ๊ต์ˆ˜๋Š” ํ˜„์žฌ ๋ฏธ๊ตญ์ด โ€œํญ๋ ฅ์  ํฌํ“ฐ๋ฆฌ์ฆ˜(Violent Populism)โ€œ์˜ ์‹œ๋Œ€์— ์ง„์ž…ํ–ˆ๋‹ค๊ณ  ์ง„๋‹จํ•ฉ๋‹ˆ๋‹ค. ์ด๋Š” ๊ทธ ์–ด๋–ค ํ…Œ๋Ÿฌ ์ง‘๋‹จ์ด๋‚˜ ์™ธ๊ตญ ๊ตญ๊ฐ€๋ณด๋‹ค๋„ ๋ฏธ๊ตญ ๋ฏผ์ฃผ์ฃผ์˜์— ๋” ํฐ ์œ„ํ˜‘์ด ๋œ๋‹ค๋Š” ์ฃผ์žฅ์ž…๋‹ˆ๋‹ค. ๊ทธ์— ๋”ฐ๋ฅด๋ฉด, ์˜ค๋Š˜๋‚  ๋ฏธ๊ตญ ์ •์น˜ ํญ๋ ฅ์˜ ๊ฐ€์žฅ ์ค‘์š”ํ•œ ์‚ฌ์‹ค์€ **โ€œ์ˆ˜์ฒœ๋งŒ ๋ช…์˜ ๋ฏธ๊ตญ์ธ์ด ์–‘์ชฝ ์ง„์˜ ๋ชจ๋‘์—์„œ ์ •์น˜์  ํญ๋ ฅ์„ ์šฉ๋‚ฉํ•  ๋งŒํ•˜๋‹ค๊ณ  ๋ณธ๋‹คโ€**๋Š” ์ ์ž…๋‹ˆ๋‹ค.

์ด๋Š” ์†Œ์ˆ˜์˜ ๊ทน๋‹จ์ฃผ์˜์ž๋‚˜ ๋ฏผ๋ณ‘๋Œ€ ์ง‘๋‹จ์— ๊ตญํ•œ๋œ ํ˜„์ƒ์ด ์•„๋‹™๋‹ˆ๋‹ค. ์ˆ˜์ฒœ๋งŒ ๋ช…์˜ ์‹œ๋ฏผ์ด ํญ๋ ฅ์„ ์šฉ์ธํ•˜๋Š” ์ƒํ™ฉ์€ ๊ณต๊ฒฉ์˜ ์œ„ํ—˜์— ๋Œ€ํ•œ ๋ชจ๋“  ๊ฒƒ์„ ๋ณ€ํ™”์‹œํ‚ต๋‹ˆ๋‹ค. ๊ฐœ์ธ์ ์ธ ์‹ฌ๋ฆฌ ์‚ฌํšŒ์  ์ด์œ ๋กœ ๋ถˆ์•ˆ์ •ํ•œ ์‚ฌ๋žŒ๋“ค์ด ์‚ฌํšŒ์  ์Šน์ธ์„ ๊ธฐ๋Œ€ํ•˜๋ฉฐ ํญ๋ ฅ์˜ ๊ฒฝ๊ณ„์„ ์„ ๋„˜๊ฒŒ ๋งŒ๋“ญ๋‹ˆ๋‹ค. ๋˜ํ•œ, ์ˆ˜์ฒœ๋งŒ ๋ช…์ด ์ •์น˜ ํญ๋ ฅ์„ ์ง€์ง€ํ•˜๋ฉด FBI์— ์ œ๋ณดํ•˜๋Š” ๊ฒƒ์กฐ์ฐจ ๊ทน๋„๋กœ ์–ด๋ ค์›Œ์ง‘๋‹ˆ๋‹ค. ํญ๋ ฅ์— ๋Œ€ํ•œ ์ง€์ง€๊ฐ€ ๋งŽ์„์ˆ˜๋ก ์‚ฌ๋žŒ๋“ค์€ ๋‹จ์ˆœํžˆ ๋ชป ๋ณธ ์ฒ™ํ•˜๊ฑฐ๋‚˜, ๋ชฉ๊ฒฉํ•œ ๊ฒƒ์„ ๋Œ€์ˆ˜๋กญ์ง€ ์•Š๊ฒŒ ์—ฌ๊ธฐ๊ธฐ ๋•Œ๋ฌธ์ž…๋‹ˆ๋‹ค.

ํŽ˜์ดํ”„ ๊ต์ˆ˜๋Š” ์ด๋Ÿฌํ•œ ํ˜„์ƒ์ด ์ขŒํŒŒ์™€ ์šฐํŒŒ ์–‘์ชฝ ๋ชจ๋‘์—์„œ ๋‚˜ํƒ€๋‚˜๋ฉฐ, ์ด์ œ๋Š” ๋‹จ์ˆœํžˆ ์ฆ๊ฐ€ํ•˜๋Š” ๊ฒƒ์„ ๋„˜์–ด **โ€œ๋‚˜์„ ํ˜•์œผ๋กœ ์ฆํญ๋˜๊ธฐ ์‹œ์ž‘ํ–ˆ๋‹คโ€**๊ณ  ๊ฒฝ๊ณ ํ•ฉ๋‹ˆ๋‹ค. ๊ทธ๋Š” ์ตœ๊ทผ 12๊ฐœ์›”๊ฐ„์˜ ์‚ฌ๋ก€๋กœ 2025๋…„ 6์›” ๋ฏธ๋„ค์†Œํƒ€ ๋ฏผ์ฃผ๋‹น ์ง€๋„์ž ์•”์‚ด, ๋ช‡ ๋‹ฌ ํ›„ ์ฐฐ๋ฆฌ ์ปคํฌ(Charlie Kirk) ์•”์‚ด, ํŽœ์‹ค๋ฒ ์ด๋‹ˆ์•„ ์ฃผ์ง€์‚ฌ ๊ด€์ € ๋ฐฉํ™”, ๊ทธ๋ฆฌ๊ณ  ์—ฌ๋Ÿฌ ์ฐจ๋ก€์˜ ํŠธ๋Ÿผํ”„ ์ „ ๋Œ€ํ†ต๋ น ์•”์‚ด ์‹œ๋„๋ฅผ ์–ธ๊ธ‰ํ•˜๋ฉฐ, ์ •์น˜ ํญ๋ ฅ์ด ์–‘์ชฝ ์ง„์˜์—์„œ ๋ถ„๋ช…ํžˆ ์ฆ๊ฐ€ํ•˜๊ณ  ์žˆ์Œ์„ ๊ฐ•์กฐํ–ˆ์Šต๋‹ˆ๋‹ค.

๋ฐ์ดํ„ฐ๋กœ ๋ณธ ์ •์น˜ ํญ๋ ฅ์˜ ํ™•์‚ฐ

ํŽ˜์ดํ”„ ๊ต์ˆ˜ ์—ฐ๊ตฌํŒ€์€ ์ง€๋‚œ 5๋…„๊ฐ„ 20์ฐจ๋ก€์— ๊ฑธ์ณ ์ „๊ตญ ๋Œ€ํ‘œ์„ฑ ์žˆ๋Š” ๋ฏธ๊ตญ์ธ ์„ค๋ฌธ์กฐ์‚ฌ๋ฅผ ์‹ค์‹œํ–ˆ์Šต๋‹ˆ๋‹ค. ์‹œ์นด๊ณ ๋Œ€ํ•™๊ต ์•ˆ๋ณด ๋ฐ ์œ„ํ˜‘ ํ”„๋กœ์ ํŠธ(CPOS, University of Chicago Project on Security and Threats)์—์„œ 2021๋…„ ์—ฌ๋ฆ„๋ถ€ํ„ฐ ์‹œ์ž‘๋œ ์ด ์กฐ์‚ฌ๋Š” ๋ฏธ๊ตญ์ธ์˜ ์ •์น˜ ํญ๋ ฅ์— ๋Œ€ํ•œ ํƒœ๋„๋ฅผ ์‹ฌ์ธต์ ์œผ๋กœ ํŒŒ์•…ํ–ˆ์Šต๋‹ˆ๋‹ค.

์กฐ์‚ฌ ๊ฒฐ๊ณผ๋Š” ์ถฉ๊ฒฉ์ ์ž…๋‹ˆ๋‹ค.

  • ๋ฐ”์ด๋“  ํ–‰์ •๋ถ€ ์ดˆ๊ธฐ ๋ช‡ ๋…„๊ฐ„, ๋ฏธ๊ตญ ์ •์น˜ ์ฐธ์—ฌ ์ธ๊ตฌ์˜ **10%**๊ฐ€ ๋„๋„๋“œ ํŠธ๋Ÿผํ”„๋ฅผ ๋Œ€ํ†ต๋ น์œผ๋กœ ๋ณต๊ถŒ์‹œํ‚ค๊ธฐ ์œ„ํ•œ ์ •์น˜ ํญ๋ ฅ์„ ์ง€์ง€ํ–ˆ์Šต๋‹ˆ๋‹ค.
  • ์ด ์ค‘ 55%๋Š” โ€˜๋ฌด๋ ฅ ์‚ฌ์šฉ(use of force)โ€˜์ด โ€˜์•”์‚ด(assassination)โ€˜์„ ์˜๋ฏธํ•œ๋‹ค๊ณ  ์ƒ๊ฐํ–ˆ์Šต๋‹ˆ๋‹ค. ์ฆ‰, ๋‹น์‹œ ์ •์น˜ ์ฐธ์—ฌ ์ธ๊ตฌ์˜ ์ ˆ๋ฐ˜์ธ 10%๊ฐ€ ์•”์‚ด์„ ์ง€์ง€ํ•œ ๊ฒƒ์ž…๋‹ˆ๋‹ค.
  • ํ˜„์žฌ ๊ทธ ๋น„์œจ์€ 14%์—์„œ 21% ์‚ฌ์ด๋กœ ๊ฑฐ์˜ ๋‘ ๋ฐฐ ์ฆ๊ฐ€ํ–ˆ์Šต๋‹ˆ๋‹ค. ํŠธ๋Ÿผํ”„ ์ „ ๋Œ€ํ†ต๋ น ๋‹น์„  ์งํ›„๋ถ€ํ„ฐ ์ด ์ˆ˜์น˜๋Š” ๋‘ ๋ฐฐ๊ฐ€ ๋˜์—ˆ๊ณ , ์ดํ›„์—๋„ ์ค„์–ด๋“ค์ง€ ์•Š๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค.
  • 2025๋…„ 9์›”, ์ฐฐ๋ฆฌ ์ปคํฌ ์•”์‚ด๊ณผ ๋‚ธ์‹œ ํŽ ๋กœ์‹œ ์•”์‚ด ์‹œ๋„์— ๋Œ€ํ•ด โ€œ์šฉ๋‚ฉํ•  ๋งŒํ•˜๋‹คโ€๊ณ  ์‘๋‹ตํ•œ ๋น„์œจ์€ ๊ฐ๊ฐ 10%์˜€์Šต๋‹ˆ๋‹ค. ์ด๋Š” ์•ฝ 2์ฒœ๋งŒ ๋ช…์˜ ์„ฑ์ธ ๋ฏธ๊ตญ์ธ์ด ๊ฐ ์‚ฌ๊ฑด์„ ์šฉ์ธํ•œ๋‹ค๋Š” ์˜๋ฏธ์ž…๋‹ˆ๋‹ค.

์ข…ํ•ฉ์ ์œผ๋กœ ๋ณผ ๋•Œ, ํ˜„์žฌ ๋ฏธ๊ตญ์ธ 5๋ช… ์ค‘ 1๋ช…๊ผด๋กœ ์–ด๋–ค ํ˜•ํƒœ์˜ ์ •์น˜ ํญ๋ ฅ์„ ์ง€์ง€ํ•˜๊ณ  ์žˆ์œผ๋ฉฐ, ์ด ์ˆ˜์น˜๋Š” ์ŠคํŽ™ํŠธ๋Ÿผ์˜ ์–‘์ชฝ์—์„œ ๋ชจ๋‘ ์ฆ๊ฐ€ํ•˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค.

๊ณผ๊ฑฐ์™€๋Š” ๋‹ค๋ฅธ ํ˜„์žฌ์˜ ํญ๋ ฅ

ํŽ˜์ดํ”„ ๊ต์ˆ˜๋Š” ๋ฏธ๊ตญ์ด ๊ณผ๊ฑฐ์—๋„ โ€œํญ๋ ฅ์  ํฌํ“ฐ๋ฆฌ์ฆ˜โ€ ์‹œ๊ธฐ๋ฅผ ๊ฒช์—ˆ๋‹ค๊ณ  ์„ค๋ช…ํ•ฉ๋‹ˆ๋‹ค.

  • 1920๋…„๋Œ€: ์ œ1์ฐจ ์„ธ๊ณ„๋Œ€์ „ ์ดํ›„ ์ œ2์ฐจ ์ฟ  ํด๋Ÿญ์Šค ํด๋žœ(Ku Klux Klan)์ด ์ˆ˜์‹ญ๋งŒ ๋ช…์—์„œ 400๋งŒ~600๋งŒ ๋ช…์œผ๋กœ ๊ธ‰์ฆํ•˜๋ฉฐ ์ˆ˜๋ฐฑ๋งŒ ๋ช…์˜ ๋ฏธ๊ตญ์ธ์ด ์ •์น˜ ํญ๋ ฅ์„ ์ง€์ง€ํ–ˆ์Šต๋‹ˆ๋‹ค.
  • 1960๋…„๋Œ€: ์ผ๋ จ์˜ ์ •์น˜์  ์•”์‚ด ์‚ฌ๊ฑด๊ณผ ํ•จ๊ป˜ 1968๋…„ ์‹œ์นด๊ณ  ์ „๋‹น๋Œ€ํšŒ์™€ ๊ฐ™์€ ์ง‘๋‹จ ์ •์น˜ ํญ๋ ฅ๊ณผ ํญ๋™์ด ๋ฐœ์ƒํ–ˆ์œผ๋ฉฐ, ๋‹น์‹œ ์—ฌ๋ก ์กฐ์‚ฌ์—์„œ๋„ ์ˆ˜์ฒœ๋งŒ ๋ช…์˜ ํญ๋ ฅ ์ง€์ง€๊ฐ€ ๋‚˜ํƒ€๋‚ฌ์Šต๋‹ˆ๋‹ค.

๊ทธ๋Ÿฌ๋‚˜ ํ˜„์žฌ์˜ ์ƒํ™ฉ์€ ๊ณผ๊ฑฐ์™€ ์ค‘์š”ํ•œ ์ฐจ์ด์ ์ด ์žˆ์Šต๋‹ˆ๋‹ค. ๊ณผ๊ฑฐ์—๋Š” ์ฃผ๋กœ ํ•œ์ชฝ ์ง„์˜์—์„œ ํญ๋ ฅ์  ํฌํ“ฐ๋ฆฌ์ฆ˜์ด ๋ฐœ์ƒํ–ˆ์ง€๋งŒ, ํ˜„์žฌ๋Š” ์–‘์ชฝ ์ง„์˜์—์„œ ๋™์‹œ์— ๋ฐœ์ƒํ•˜๊ณ  ์žˆ๋‹ค๋Š” ์ ์ž…๋‹ˆ๋‹ค. 1960๋…„๋Œ€๋Š” ์ฃผ๋กœ ์ขŒํŒŒ์—์„œ, 1920๋…„๋Œ€ ์ดˆ๋ฐ˜์€ ์ฃผ๋กœ ์šฐํŒŒ์—์„œ ํญ๋ ฅ์ด ๋ฐœ์ƒํ–ˆ์Šต๋‹ˆ๋‹ค.

์ˆ˜์น˜๋กœ ๋ด๋„ ์ •์น˜ ํญ๋ ฅ์˜ ์ฆ๊ฐ€๋Š” ๋ช…ํ™•ํ•ฉ๋‹ˆ๋‹ค. ํŽ˜์ดํ”„ ๊ต์ˆ˜์˜ ์—ฐ๊ตฌํŒ€์ด 2001๋…„๋ถ€ํ„ฐ 2024๋…„๊นŒ์ง€ 25๋…„๊ฐ„ ์—ฐ๋ฐฉ์ˆ˜์‚ฌ๊ตญ(FBI) ๋ฐ์ดํ„ฐ๋ฅผ ํฌํ•จํ•œ ๋‹ค์–‘ํ•œ ๋ฐ์ดํ„ฐ๋ฒ ์ด์Šค๋ฅผ ๋ถ„์„ํ•œ ๊ฒฐ๊ณผ, ํŠธ๋Ÿผํ”„ ๋Œ€ํ†ต๋ น ์žฌ์ž„ ์ฒซํ•ด์ธ 2017๋…„๋ถ€ํ„ฐ ์˜์›๋“ค์— ๋Œ€ํ•œ ์œ„ํ˜‘์ด 5๋ฐฐ ์ฆ๊ฐ€ํ–ˆ์Šต๋‹ˆ๋‹ค. ์ด ์ฆ๊ฐ€๋Š” ๋ฏผ์ฃผ๋‹น๊ณผ ๊ณตํ™”๋‹น ์˜์› ๋ชจ๋‘์—๊ฒŒ ํ•ด๋‹นํ•˜๋ฉฐ, ํŠธ๋Ÿผํ”„ ์ „ ๋Œ€ํ†ต๋ น์˜ ๋ถ€์ƒ์ด ์–‘์ชฝ ์ง„์˜ ๋ชจ๋‘์—๊ฒŒ โ€˜ํ”ผ๋ขฐ์นจโ€™ ์—ญํ• ์„ ํ–ˆ์Œ์„ ๋ณด์—ฌ์ค๋‹ˆ๋‹ค. ํŠธ๋Ÿผํ”„ ์˜์ œ๋ฅผ ๊ณผ๋„ํ•˜๊ฒŒ ์ง€์ง€ํ•˜๊ฑฐ๋‚˜(์˜ˆ: ๋งˆ์กฐ๋ฆฌ ํ…Œ์ผ๋Ÿฌ ๊ทธ๋ฆฐ), ์ถฉ๋ถ„ํžˆ ์ง€์ง€ํ•˜์ง€ ์•Š๋Š” ์˜์›๋“ค ๋ชจ๋‘๊ฐ€ ํ‘œ์ ์ด ๋˜๋ฉด์„œ, ์œ„ํ˜‘ ์ˆ˜์ค€์€ ๊ธ‰๊ฒฉํžˆ ์ƒ์Šนํ•˜์—ฌ ์•ฝ 8๋…„๊ฐ„ ํŠธ๋Ÿผํ”„์™€ ๋ฐ”์ด๋“  ํ–‰์ •๋ถ€ ๋‚ด๋‚ด ๊ทธ ๋†’์€ ์ˆ˜์ค€์„ ์œ ์ง€ํ–ˆ์Šต๋‹ˆ๋‹ค. ๋Œ€ํ†ต๋ น์ด ๋ฐ”๋€Œ์–ด๋„ ์œ„ํ˜‘์ด ์ค„์–ด๋“ค์ง€ ์•Š์•˜๋‹ค๋Š” ๊ฒƒ์€ ๋‹จ์ˆœํžˆ ํŠน์ • ์ธ๋ฌผ(ํŠธ๋Ÿผํ”„) ์ด์ƒ์˜ ๋ฌธ์ œ๊ฐ€ ์žˆ์Œ์„ ์‹œ์‚ฌํ•ฉ๋‹ˆ๋‹ค.

์ •์น˜ ํญ๋ ฅ์˜ ๊ทผ๋ณธ ์›์ธ: ๋‘ ๊ฐ€์ง€ ๊ฑฐ๋Œ€ํ•œ ์‚ฌํšŒ ๋ณ€ํ™”

ํŽ˜์ดํ”„ ๊ต์ˆ˜๋Š” ์ด๋Ÿฌํ•œ ํญ๋ ฅ ์ฆ๊ฐ€์˜ ๊ทผ๋ณธ ์›์ธ์ด ๋‹จ์ˆœํžˆ ๊ฐœ์ธ์ด๋‚˜ ์†Œ์…œ ๋ฏธ๋””์–ด์˜ ์˜ํ–ฅ์ด ์•„๋‹ˆ๋ผ๊ณ  ๊ฐ•์กฐํ•ฉ๋‹ˆ๋‹ค. ๊ทธ๋Š” **โ€œ์‚ฌํšŒ ๋ณ€ํ™”์˜ ์˜ํ–ฅโ€**์ด ๊ฐ€์žฅ ์ค‘์š”ํ•˜๋‹ค๊ณ  ๋งํ•ฉ๋‹ˆ๋‹ค. ๊ตญ๊ฐ€์—์„œ ์ฃผ์š” ์‚ฌํšŒ ๋ณ€ํ™”๊ฐ€ ๋ฐœ์ƒํ•˜๋ฉด ์ •์น˜์˜ ๊ธ‰์ง„ํ™”๋ฅผ ์ดˆ๋ž˜ํ•˜๊ณ , ์ด๋Š” ๋‹ค์‹œ ์ •์น˜ ํญ๋ ฅ์„ ์œ ๋ฐœํ•  ์ˆ˜ ์žˆ๋‹ค๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค. ํ˜„์žฌ ๋ฏธ๊ตญ์—์„œ ์ผ์–ด๋‚˜๊ณ  ์žˆ๋Š” ๋‘ ๊ฐ€์ง€ ์ฃผ์š” ์‚ฌํšŒ ๋ณ€ํ™”๋Š” ๋‹ค์Œ๊ณผ ๊ฐ™์Šต๋‹ˆ๋‹ค.

1. ์ธ๊ตฌํ†ต๊ณ„ํ•™์  ๋ณ€ํ™” (Demographic Shift): ๋ฐฑ์ธ ๋‹ค์ˆ˜์—์„œ ์†Œ์ˆ˜๋กœ

๋ฏธ๊ตญ 250๋…„ ์—ญ์‚ฌ์ƒ ์ฒ˜์Œ์œผ๋กœ ๋ฐฑ์ธ ๋‹ค์ˆ˜ ๋ฏผ์ฃผ์ฃผ์˜์—์„œ ๋ฐฑ์ธ ์†Œ์ˆ˜ ๋ฏผ์ฃผ์ฃผ์˜๋กœ ์ „ํ™˜ํ•˜๊ณ  ์žˆ๋‹ค๋Š” ์ ์ž…๋‹ˆ๋‹ค. 1960๋…„์—๋Š” ๋ฏธ๊ตญ์ธ์˜ ์•ฝ 88%๊ฐ€ ๋น„ํžˆ์ŠคํŒจ๋‹‰๊ณ„ ๋ฐฑ์ธ์ด์—ˆ์ง€๋งŒ, ์˜ค๋Š˜๋‚ ์—๋Š” 57%์— ๋ถˆ๊ณผํ•˜๋ฉฐ, ์•ฝ 20๋…„ ํ›„์—๋Š” 49%๊ฐ€ ๋  ๊ฒƒ์œผ๋กœ ์˜ˆ์ƒ๋ฉ๋‹ˆ๋‹ค. ์ด๋ฏผ์ด ์ด๋Ÿฌํ•œ ๋ณ€ํ™”์˜ ์ฃผ๋œ ์›์ธ ์ค‘ ํ•˜๋‚˜์ž…๋‹ˆ๋‹ค.

ํŽ˜์ดํ”„ ๊ต์ˆ˜๋Š” ์ด โ€˜ํ‹ฐํ•‘ ํฌ์ธํŠธ(tipping point)โ€™ ์„ธ๋Œ€, ์ฆ‰ ๋ฐฑ์ธ ๋‹ค์ˆ˜์—์„œ ์†Œ์ˆ˜๋กœ ์ „ํ™˜๋˜๋Š” 20๋…„ ๋™์•ˆ ์ด๋ฏผ ๋ฌธ์ œ๊ฐ€ ๋ฏธ๊ตญ์—์„œ ๊ฐ€์žฅ ํฐ โ€˜ํ”ผ๋ขฐ์นจโ€™์ด ๋˜๋Š” ๊ฒƒ์€ ๋‹น์—ฐํ•˜๋‹ค๊ณ  ๋งํ•ฉ๋‹ˆ๋‹ค.

  • ๊ธฐ์กด ์ง‘๋‹จ(๋ฐฑ์ธ ๋‹ค์ˆ˜): ์ด๋Ÿฌํ•œ ๋ณ€ํ™”๋ฅผ ๋ฉˆ์ถ”๊ฑฐ๋‚˜ ๋˜๋Œ๋ฆฌ๊ธฐ๋ฅผ ์›ํ•ฉ๋‹ˆ๋‹ค. ๋ณ€ํ™”๊ฐ€ ์ง„ํ–‰๋˜๋ฉด ์ •์น˜์  ๊ถŒ๋ ฅ์„ ์žƒ๊ณ  ๊ฒฝ์ œ์  ์ „๋ง์ด ์•…ํ™”๋  ๊ฒƒ์ด๊ธฐ ๋•Œ๋ฌธ์ž…๋‹ˆ๋‹ค.
  • ์ƒˆ๋กœ์šด ์ง‘๋‹จ(์†Œ์ˆ˜ ์ง‘๋‹จ): ์ด๋Ÿฌํ•œ ๋ณ€ํ™”๊ฐ€ ๊ณ„์†๋˜๊ธฐ๋ฅผ ๊ฐ„์ ˆํžˆ ์›ํ•˜๋ฉฐ, ์˜คํžˆ๋ ค ๊ฐ€์†ํ™”ํ•˜๊ธฐ๋ฅผ ๋ฐ”๋ž๋‹ˆ๋‹ค. ๊ทธ๋“ค์€ ์ด ๋ณ€ํ™”๋กœ๋ถ€ํ„ฐ ํ˜œํƒ์„ ๋ฐ›์„ ๊ฒƒ์ด๊ธฐ ๋•Œ๋ฌธ์ž…๋‹ˆ๋‹ค.

์ด๋Ÿฌํ•œ ์ธ๊ตฌํ†ต๊ณ„ํ•™์  ๋ณ€ํ™”๋Š” ์ •์น˜ ํญ๋ ฅ ์ง€์ง€๋ฅผ ์ง์ ‘์ ์œผ๋กœ ๋ถ€์ถ”๊ธฐ๋Š” ํ•ต์‹ฌ ์Ÿ์ ์ด ๋˜๊ณ  ์žˆ์œผ๋ฉฐ, ์ด๋Š” ์„ค๋ฌธ์กฐ์‚ฌ์—์„œ๋„ ๋ช…ํ™•ํžˆ ๋“œ๋Ÿฌ๋‚ฉ๋‹ˆ๋‹ค.

2. ๋ถ€์˜ ์ง‘์ค‘ (Wealth Shift): ์ƒ์œ„ 1%๋กœ์˜ ๋ถ€์˜ ์ด๋™

์ธ๊ตฌํ†ต๊ณ„ํ•™์  ๋ณ€ํ™”์™€ ๋”๋ถˆ์–ด, 1980๋…„๋Œ€ ์ค‘๋ฐ˜๋ถ€ํ„ฐ ์‹œ์ž‘๋œ ์ƒ์œ„ 1%๋กœ์˜ ๋ถ€์˜ ์ง‘์ค‘ ํ˜„์ƒ์ด ์žˆ์Šต๋‹ˆ๋‹ค. ์ง‘๊ถŒ๋‹น์ด ์–ด๋А ๋‹น์ด๋“  ์ƒ๊ด€์—†์ด, ํ•˜์œ„ 90%๋กœ๋ถ€ํ„ฐ ์ƒ์œ„ 1%๋กœ ๋ถ€๊ฐ€ ์ด๋™ํ•˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค. ์ด๋Š” ํ•˜์œ„ 90%์— ์†ํ•˜๋Š” ๊ณตํ™”๋‹น ์ง€์ง€์ž๋“ค๊ณผ ๋ฏผ์ฃผ๋‹น ์ง€์ง€์ž๋“ค ๋ชจ๋‘๊ฐ€ ์ž์‹ ๋“ค์˜ ๊ฒฝ์ œ์  ์ „๋ง์ด ๋”์šฑ ๋‚˜๋น ์งˆ ์ˆ˜ ์žˆ๋‹ค๋Š” ์šฐ๋ ค๋ฅผ ๊ฐ€์ง€๊ฒŒ ๋งŒ๋“ญ๋‹ˆ๋‹ค.

์ด๋Ÿฌํ•œ ๋ถ€์˜ ์ง‘์ค‘ ํ˜„์ƒ์„ ์–ด๋А ์ •๋‹น๋„ ํ•ด๊ฒฐํ•˜์ง€ ๋ชปํ•˜๊ณ  ์žˆ๋‹ค๋Š” ์ ์ด ๋ฌธ์ œ์ž…๋‹ˆ๋‹ค. ์–‘๋‹น ๋ชจ๋‘ ๋ณธ์งˆ์ ์œผ๋กœ ๊ธฐ๋ถ€์ž๋“ค์—๊ฒŒ ๋ˆ์„ ์ฃผ๊ณ  ์žˆ์œผ๋ฉฐ, ์ด ์ƒํ™ฉ์€ ๋ณ€ํ•˜์ง€ ์•Š๊ณ  ๊ณ„์†๋˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค. ์ด๋Š” ์ •์น˜ ํญ๋ ฅ์˜ ์ดˆ๋‹นํŒŒ์ (bipartisan) ์ฆ๊ฐ€์— ์ถ”๊ฐ€์ ์ธ ์—ฐ๋ฃŒ๊ฐ€ ๋ฉ๋‹ˆ๋‹ค.

๊ฒฐ๋ก ์ ์œผ๋กœ, ํŽ˜์ดํ”„ ๊ต์ˆ˜๋Š” ์ด๋Ÿฌํ•œ ๋‘ ๊ฐ€์ง€ ๊ฑฐ๋Œ€ํ•œ ์‚ฌํšŒ ๋ณ€ํ™”๊ฐ€ ํŠน์ • ์‚ฌ๋žŒ๋“ค์˜ ๊ถŒ๋ ฅ๊ณผ ์˜ํ–ฅ๋ ฅ์˜ ๋ณ€ํ™”, ๊ทธ๋ฆฌ๊ณ  ๋ฏธ๋ž˜์— ๋Œ€ํ•œ ๋ถˆ์•ˆ๊ฐ์œผ๋กœ ์ด์–ด์ง€๋ฉฐ, **โ€œ๋งŽ์€ ์‚ฌ๋žŒ๋“ค์ด ์ž์‹ ๋“ค์˜ ๊ถŒ๋ ฅ์„ ์žƒ๊ณ  ์žˆ๋‹ค๊ณ  ๋А๋ผ๋Š” ๋Œ€๊ทœ๋ชจ์˜ ๋ณ€ํ™”โ€**๊ฐ€ ์ •์น˜ ํญ๋ ฅ์˜ ํ•ต์‹ฌ ๋™๋ ฅ์ด๋ผ๊ณ  ์ง€์ ํ•ฉ๋‹ˆ๋‹ค.

โ€˜์žƒ์„ ๊ฒƒ ๋งŽ์€โ€™ ์ด๋“ค์˜ ํญ๋ ฅ: ์ •์น˜์  ๋ฐฐ์ œ์— ๋Œ€ํ•œ ๋‘๋ ค์›€

์šฐ๋ฆฌ๋Š” ํ”ํžˆ ์ •์น˜ ํญ๋ ฅ์„ ์ €์ง€๋ฅด๋Š” ์‚ฌ๋žŒ๋“ค์„ โ€˜์žƒ์„ ๊ฒƒ์ด ์—†๋Š”โ€™ ์‚ฌํšŒ์˜ ์ฃผ๋ณ€์ธ, ์ฆ‰ ์‚ฌํšŒ์  ์•ฝ์ž๋ผ๊ณ  ์ƒ๊ฐํ•˜๋Š” ๊ฒฝํ–ฅ์ด ์žˆ์Šต๋‹ˆ๋‹ค. ๊ทธ๋Ÿฌ๋‚˜ ํŽ˜์ดํ”„ ๊ต์ˆ˜์˜ ์—ฐ๊ตฌ๋Š” ์ด๋Ÿฌํ•œ ํ†ต๋…์„ ๋’ค์ง‘์Šต๋‹ˆ๋‹ค. ๊ทธ์˜ ์„ค๋ฌธ์กฐ์‚ฌ์— ๋”ฐ๋ฅด๋ฉด, ๊ต์œก ์ˆ˜์ค€์ด ๋†’๊ณ  ์ค‘์‚ฐ์ธต ๋ฐ ์ƒ๋ฅ˜์ธต์— ์†ํ•˜๋Š” ์ƒ๋‹น์ˆ˜์˜ ์‚ฌ๋žŒ๋“ค์ด ์ •์น˜ ํญ๋ ฅ์„ ์ง€์ง€ํ•˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค. ๊ทธ๋“ค์€ ํ˜„์žฌ ๊ฐ€์ง„ ๊ฒƒ์ด ๋งŽ์œผ๋ฉฐ, ๋ฏธ๋ž˜์— ๋” ๋งŽ์€ ๊ฒƒ์„ ์žƒ์„๊นŒ ๋ด ๋‘๋ ค์›Œํ•˜๋Š” ์‚ฌ๋žŒ๋“ค์ž…๋‹ˆ๋‹ค.

๊ทธ๋“ค์ด ํญ๋ ฅ์— ์˜์กดํ•˜๋Š” ์ด์œ ๋Š” ๋ฌด์—‡์ผ๊นŒ์š”? ํŽ˜์ดํ”„ ๊ต์ˆ˜๋Š” ํ•ต์‹ฌ์ ์ธ ๋‘๋ ค์›€์œผ๋กœ **โ€œ์ •์น˜์  ๋ฐฐ์ œ(Political Exclusion)โ€œ**๋ฅผ ๊ผฝ์Šต๋‹ˆ๋‹ค. ์†Œ์ˆ˜์ž๊ฐ€ ๋˜์–ด ์ •์น˜ ์‹œ์Šคํ…œ ๋‚ด์—์„œ ์ž์‹ ๋“ค์˜ ๋ถˆ๋งŒ์„ ํ•ด๊ฒฐํ•˜๊ฑฐ๋‚˜ ์ด์ต์„ ๊ด€์ฒ ํ•  ์ˆ˜ ์—†๊ฒŒ ๋  ๊ฒƒ์ด๋ผ๋Š” ๋‘๋ ค์›€์ž…๋‹ˆ๋‹ค. ์ด๋Ÿฌํ•œ ๋‘๋ ค์›€์€ ์–‘์ชฝ ์ง„์˜ ๋ชจ๋‘์—์„œ ๋ฐœ์ƒํ•ฉ๋‹ˆ๋‹ค. ์˜ˆ๋ฅผ ๋“ค์–ด, ํŠธ๋Ÿผํ”„ ์ „ ๋Œ€ํ†ต๋ น์ด ํˆฌํ‘œ ๊ธฐ๊ณ„ ์••๋ฅ˜์— ๋Œ€ํ•ด ์ด์•ผ๊ธฐํ•˜๋Š” ๊ฒƒ์€ ๋ฏผ์ฃผ๋‹น ์ง€์ง€์ž๋“ค์—๊ฒŒ โ€˜์ •์น˜์  ๋ด‰์‡„(lockout)โ€˜์— ๋Œ€ํ•œ ๋‘๋ ค์›€์„ ์‹ฌ์–ด์ค๋‹ˆ๋‹ค. ์ „ํ™˜๊ธฐ์ธ 10~20๋…„ ๋™์•ˆ ์–‘์ชฝ ๋ชจ๋‘ ์ •์น˜์  ๋ด‰์‡„๋ฅผ ๋‘๋ ค์›Œํ•  ์ˆ˜ ์žˆ์œผ๋ฉฐ, ์ด๋Š” ๋Œ์ดํ‚ฌ ์ˆ˜ ์—†๊ฒŒ ๋˜๊ธฐ ์ „ ๋งˆ์ง€๋ง‰ ๊ธฐํšŒ๋ผ๋Š” ์ƒ๊ฐ์— ๋”์šฑ ๊ฒฉ๋ ฌํ•˜๊ฒŒ ์‹ธ์šฐ๊ฒŒ ๋งŒ๋“ญ๋‹ˆ๋‹ค.

๊ฒฐ๊ตญ, ์ด๋Š” ํ˜„์žฌ์˜ ์ •์น˜ ์‹œ์Šคํ…œ์— ๋Œ€ํ•œ ๋น„ํŒ์ž…๋‹ˆ๋‹ค. ์–‘๋‹น ๋ชจ๋‘ ๋งŽ์€ ์‚ฌ๋žŒ๋“ค์—๊ฒŒ ์ •์น˜์  ๋ณ€ํ™”๊ฐ€ ์‹ค์ œ๋กœ ๊ฐ€๋Šฅํ•˜๋‹ค๋Š” ๋А๋‚Œ์„ ์ฃผ์ง€ ๋ชปํ•˜๊ณ  ์žˆ์œผ๋ฉฐ, ์ด๋Ÿฌํ•œ ๋‘๋ ค์›€์€ ํŠน์ • ์‚ฌ๋žŒ๋“ค์—๊ฒŒ ์‹ค์กด์ (existential)์ธ ๋ฌธ์ œ๋กœ ๋‹ค๊ฐ€์˜ต๋‹ˆ๋‹ค. ๊ทธ๋“ค์€ ์ž์‹ ์ด ๋ฐฐ์ œ๋  ๊ฒƒ์ด๋ฉฐ, ๊ณง ์ž์‹ ๋“ค์˜ ๋ชฉ์†Œ๋ฆฌ๊ฐ€ ๋“ค๋ฆด ๋‹ค๋ฅธ ๋ฐฉ๋ฒ•์ด ์—†์„ ๊ฒƒ์ด๋ผ๊ณ  ๋А๋‚๋‹ˆ๋‹ค. ์ด๋Š” ์‚ฌํšŒ๋ณด์žฅ์„ธ์œจ๊ณผ ๊ฐ™์€ ์ •์ฑ… ๋…ผ์Ÿ์„ ๋„˜์–ด์„ , **โ€œ์ง„์ •ํ•œ ์‹ค์กด์  ์ •์น˜(existential politics)์˜ ๋ถ€์ƒโ€**์œผ๋กœ ์ด์–ด์ง€๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค.

ํญ๋ ฅ์œผ๋กœ ๊ฐ€๋Š” ๊ธธ: ์ง€๋„์ž์˜ ์ˆ˜์‚ฌ(Rhetoric)์™€ ์‚ฌํšŒ์  ์Šน์ธ

๊ทธ๋ ‡๋‹ค๋ฉด, ๋ฌด๋ ฅ๊ฐ์„ ๋А๋ผ๋Š” ์‚ฌ๋žŒ๋“ค์ด ์–ด๋–ป๊ฒŒ ์‹ค์ œ๋กœ ํญ๋ ฅ์„ ํ–‰์‚ฌํ•˜๋Š” ์ง€์ ๊นŒ์ง€ ๋‚˜์•„๊ฐˆ๊นŒ์š”? ํŽ˜์ดํ”„ ๊ต์ˆ˜๋Š” **โ€œ์ง€๋„์ž๋“ค์˜ ๋ฐœ์–ธโ€**์ด ์ค‘์š”ํ•˜๋‹ค๊ณ  ์ง€์ ํ•ฉ๋‹ˆ๋‹ค. ๋„๋„๋“œ ํŠธ๋Ÿผํ”„ ์ „ ๋Œ€ํ†ต๋ น์ด โ€œ๊ทธ๋ฅผ ์–ผ๊ตด์— ์ฃผ๋จน์œผ๋กœ ๋•Œ๋ฆฌ๊ณ  ์‹ถ๋‹คโ€๊ณ  ๋งํ•˜๋ฉฐ ํญ๋ ฅ์„ ์•”์‹œํ•˜๋Š” ๋ฐœ์–ธ์„ ํ•œ ๊ฒƒ์ด ๋Œ€ํ‘œ์ ์ž…๋‹ˆ๋‹ค. ๋˜ํ•œ ๋ฒ ํ†  ์˜ค๋ฃจํฌ(Beto Oโ€™Rourke)๋‚˜ ๊ฐœ๋นˆ ๋‰ด์„ฌ(Gavin Newsom) ๊ฐ™์€ ๋ฏผ์ฃผ๋‹น ์ง€๋„์ž๋“ค๋„ โ€œ๋จผ์ € ์ฃผ๋จน์„ ๋‚ ๋ ค์•ผ ํ•œ๋‹คโ€๊ฑฐ๋‚˜ โ€œ๊นกํŒจ์—๊ฒŒ ๋งž์„œ ์‹ธ์šด๋‹คโ€๋Š” ์‹์˜ ๊ฐ•๊ฒฝํ•œ ํ‘œํ˜„์„ ์‚ฌ์šฉํ•˜๋ฉฐ ์ง€์ง€์ž๋“ค์—๊ฒŒ ์ž์‹ ๋“ค์ด ์‹ธ์šธ ๊ฒƒ์ž„์„ ๋ณด์—ฌ์ค๋‹ˆ๋‹ค.

์ด๋Ÿฌํ•œ ์ง€๋„์ž๋“ค์€ ๊ถŒ์œ„ ์žˆ๋Š” ์ธ๋ฌผ์ด๋ฉฐ, ๊ทธ๋“ค์ด ํญ๋ ฅ์„ ์šฉ์ธํ•  ๋งŒํ•˜๋‹ค๊ณ  ๋งํ•˜๋ฉด(ํ˜น์€ ์˜๋„์น˜ ์•Š๊ฒŒ ๊ทธ๋ ‡๊ฒŒ ๋“ค๋ฆฌ๊ฒŒ ํ•˜๋ฉด), ๋‹ค๋ฅธ ์‚ฌ๋žŒ๋“ค๋„ ๊ทธ ๋ฐฉํ–ฅ์œผ๋กœ ์›€์ง์ผ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค. ํŽ˜์ดํ”„ ๊ต์ˆ˜๋Š” ์ด๋ฅผ โ€˜์œ„์—์„œ ์•„๋ž˜๋กœ(top-down)โ€™ ์ „๋‹ฌ๋˜๋Š” ์ˆ˜์‚ฌ๋ผ๊ณ  ํ‘œํ˜„ํ•ฉ๋‹ˆ๋‹ค. ํŠน์ • ๊ฐœ์ธ๋“ค์€ ์ด๋ฏธ ํ™”๋ฅผ ์ž˜ ๋‚ด๊ณ  ์ถฉ๋™์ ์œผ๋กœ ํ–‰๋™ํ•˜๋Š” ๊ฒฝํ–ฅ์ด ์žˆ๋Š”๋ฐ, ์ด๋Ÿฌํ•œ ์ง€๋„์ž๋“ค์˜ ๋ฐœ์–ธ์€ ๊ทธ๋“ค์„ ํ–‰๋™์œผ๋กœ ์˜ฎ๊ธฐ๊ฒŒ ํ•˜๋Š” ์ด‰๋งค์ œ๊ฐ€ ๋ฉ๋‹ˆ๋‹ค.

๋˜ํ•œ โ€˜์•„๋ž˜์—์„œ ์œ„๋กœ(bottom-up)โ€™ ์ „๋‹ฌ๋˜๋Š” ์ˆ˜์‚ฌ๋„ ์ค‘์š”ํ•œ ์—ญํ• ์„ ํ•ฉ๋‹ˆ๋‹ค. ์†Œ์…œ ๋ฏธ๋””์–ด๋Š” ์ด๋Ÿฌํ•œ ์ƒํ™ฉ์— โ€œํœ˜๋ฐœ์œ ๋ฅผ ๋ถ“๋Š”โ€ ์—ญํ• ์„ ํ•ฉ๋‹ˆ๋‹ค. ์˜ˆ๋ฅผ ๋“ค์–ด, ์œ ๋‚˜์ดํ‹ฐ๋“œ ํ—ฌ์Šค์ผ€์–ด(United Healthcare) CEO๊ฐ€ ์‚ดํ•ด๋œ ํ›„, ์‚ฌ๋žŒ๋“ค์ด X(๊ตฌ ํŠธ์œ„ํ„ฐ)์—์„œ ๋ฒ”์ธ์„ ๊ณต๊ฐœ์ ์œผ๋กœ ์ฐฌ์–‘ํ•˜๊ณ  ์˜์›…์‹œํ•˜๋Š” ํ˜„์ƒ์ด ๋‚˜ํƒ€๋‚ฌ์Šต๋‹ˆ๋‹ค. ์ด๋Š” ์ •์น˜ ํญ๋ ฅ์„ ์ง€์ง€ํ•˜๋Š” ์ˆ˜์ฒœ๋งŒ ๋ช…์˜ ๋ฏธ๊ตญ์ธ์ด ์†Œ์…œ ๋ฏธ๋””์–ด์— ์กด์žฌํ•˜๋ฉฐ, ๊ทธ๋“ค์˜ ๋ฐ˜์‘์€ ์ž ์žฌ์  ๊ณต๊ฒฉ์ž๋“ค์—๊ฒŒ โ€œ์‚ฌํšŒ์  ์Šน์ธโ€์— ๋Œ€ํ•œ ํ™•์‹ ์„ ์ค๋‹ˆ๋‹ค. ๋„๋„๋“œ ํŠธ๋Ÿผํ”„ ์ฃผ๋‹ˆ์–ด๊ฐ€ ํด ํŽ ๋กœ์‹œ ํ”ผ์Šต ์‚ฌ๊ฑด์„ ํ•ผ๋Ÿฌ์œˆ ์˜์ƒ์œผ๋กœ ์กฐ๋กฑํ•˜๋Š” ๊ฒŒ์‹œ๋ฌผ์„ ์˜ฌ๋ฆฐ ํ›„ ์—„์ฒญ๋‚œ ์ง€์ง€๋ฅผ ๋ฐ›์€ ์‚ฌ๋ก€์ฒ˜๋Ÿผ, ์ƒํ–ฅ์‹๊ณผ ํ•˜ํ–ฅ์‹ ์ˆ˜์‚ฌ๊ฐ€ ์ƒํ˜ธ์ž‘์šฉํ•˜๋ฉฐ ํญ๋ ฅ์„ ๋ถ€์ถ”๊น๋‹ˆ๋‹ค.

๊ทธ๋Ÿฌ๋‚˜ ํŽ˜์ดํ”„ ๊ต์ˆ˜๋Š” ์†Œ์…œ ๋ฏธ๋””์–ด๊ฐ€ ์ƒํ™ฉ์„ ์•…ํ™”์‹œํ‚ค์ง€๋งŒ, ๊ทธ ์ž์ฒด๊ฐ€ โ€œ๋ถˆ(fire)โ€œ์€ ์•„๋‹ˆ๋ผ๊ณ  ๊ฐ•์กฐํ•ฉ๋‹ˆ๋‹ค. 1920๋…„๋Œ€์™€ 1960๋…„๋Œ€์—๋Š” ์†Œ์…œ ๋ฏธ๋””์–ด๊ฐ€ ์—†์—ˆ์Œ์—๋„ ๋ถˆ๊ตฌํ•˜๊ณ  ์—„์ฒญ๋‚œ ์ •์น˜ ํญ๋ ฅ์ด ์กด์žฌํ–ˆ๊ธฐ ๋•Œ๋ฌธ์ž…๋‹ˆ๋‹ค. ๊ทธ๋Š” ์šฐ๋ฆฌ๊ฐ€ ์ธํ„ฐ๋„ท๊ณผ ๊ฐ™์€ ๋ถ€์ฐจ์ ์ธ ๋ฌธ์ œ์—๋งŒ ์ง‘์ค‘ํ•˜์—ฌ ๊ทผ๋ณธ์ ์ธ ์‚ฌํšŒ ๋ณ€ํ™”๋ผ๋Š” ํ•ต์‹ฌ ๋ฌธ์ œ๋ฅผ ์™ธ๋ฉดํ•˜๋Š” ๊ฒฝํ–ฅ์ด ์žˆ๋‹ค๊ณ  ์ง€์ ํ•ฉ๋‹ˆ๋‹ค.

๋‹ฌ๋ผ์ง„ ๊ณต๊ฒฉ์ž์˜ ํ”„๋กœํ•„: ์šฐ๋ฆฌ ์˜†์ง‘์˜ โ€˜๊ดด๋ฌผโ€™

ํŽ˜์ดํ”„ ๊ต์ˆ˜๋Š” ์ •์น˜ ํญ๋ ฅ์„ ์ €์ง€๋ฅด๋Š” ์‚ฌ๋žŒ๋“ค์˜ ํ”„๋กœํ•„์ด ํฌ๊ฒŒ ๋ฐ”๋€Œ์—ˆ๋‹ค๋Š” ์ ์„ ๊ฐ•์กฐํ•ฉ๋‹ˆ๋‹ค. ๊ทธ์˜ ์—ฐ๊ตฌํŒ€์€ 1์›” 6์ผ ๊ตญํšŒ์˜์‚ฌ๋‹น ๋‚œ์ž… ์‚ฌ๊ฑด์— ๊ฐ€๋‹ดํ•œ ์‚ฌ๋žŒ๋“ค์˜ ์ดˆ๊ธฐ ํ”„๋กœํ•„์„ ์ง€๋‚œ ์ˆ˜์‹ญ ๋…„๊ฐ„ FBI ๋ฐ์ดํ„ฐ์— ๋‚˜ํƒ€๋‚œ ์šฐ์ต ์ •์น˜ ํญ๋ ฅ ๋ฒ”์ฃ„์ž๋“ค์˜ ํ”„๋กœํ•„๊ณผ ๋น„๊ตํ–ˆ์Šต๋‹ˆ๋‹ค.

  • ๊ณผ๊ฑฐ์˜ ํ”„๋กœํ•„: ์‹ค์—…์ž, ๋ฏผ๋ณ‘๋Œ€์› ๋“ฑ ์‚ฌํšŒ์˜ ์ฃผ๋ณ€๋ถ€(fringe)์— ์†ํ•˜๋Š” ์‚ฌ๋žŒ๋“ค์ด์—ˆ์Šต๋‹ˆ๋‹ค.
  • 1์›” 6์ผ ์‚ฌ๊ฑด ๊ฐ€๋‹ด์ž ํ”„๋กœํ•„: ํ˜„์ €ํžˆ ๋‹ฌ๋ž์Šต๋‹ˆ๋‹ค. ์‹ค์—…๋ฅ ์€ ์ „๊ตญ ํ‰๊ท ์ธ 7%์— ๋ถˆ๊ณผํ–ˆ์œผ๋ฉฐ, ์‚ฌ์—…๊ฐ€, ์˜์‚ฌ, ๋ณ€ํ˜ธ์‚ฌ, CEO ๋“ฑ ์ฃผ๋ฅ˜ ์‚ฌํšŒ ๊ตฌ์„ฑ์›์ด ๋งŽ์•˜์Šต๋‹ˆ๋‹ค. ๋ฏผ๋ณ‘๋Œ€ ์†Œ์†์€ 10%์— ๋ถˆ๊ณผํ–ˆ์Šต๋‹ˆ๋‹ค(๊ณผ๊ฑฐ์—๋Š” 50%). ํŽ˜์ดํ”„ ๊ต์ˆ˜๋Š” ์ด๋“ค์„ โ€œ๋ฉ‹์ง„ ๊ต์™ธ์— ์‚ฌ๋Š” ๋น„์ฆˆ๋‹ˆ์Šค ์ •์žฅ ์ฐจ๋ฆผ์˜ ๋ฐ˜๋ž€์žโ€๋ผ๊ณ  ๋ฌ˜์‚ฌํ•ฉ๋‹ˆ๋‹ค.

์ด๋Ÿฌํ•œ ๋ณ€ํ™”๋Š” ์šฐ๋ฆฌ๊ฐ€ ํญ๋ ฅ์ ์ธ ์‚ฌ๋žŒ๋“ค์„ ์ž˜๋ชป๋œ ๊ณณ์—์„œ ์ฐพ๊ณ  ์žˆ์„ ์ˆ˜ ์žˆ์Œ์„ ์˜๋ฏธํ•ฉ๋‹ˆ๋‹ค. ํŽ˜์ดํ”„ ๊ต์ˆ˜๋Š” ์šฐ๋ฆฌ๊ฐ€ โ€˜๊ดด๋ฌผโ€™์„ ์šฐ๋ฆฌ์—๊ฒŒ์„œ ๋ฉ€๋ฆฌ ๋–จ์–ด์ง„ ์•…๋‹น์œผ๋กœ ์—ฌ๊ธฐ๊ธฐ๋ฅผ ์ข‹์•„ํ•œ๋‹ค๊ณ  ์„ค๋ช…ํ•ฉ๋‹ˆ๋‹ค. ์ •์น˜์  ์•”์‚ด๊ณผ ์‚ด์ธ์„ ์ €์ง€๋ฅผ ์ˆ˜ ์žˆ๋Š” โ€˜๊ดด๋ฌผโ€™์ด ์šฐ๋ฆฌ์™€ ๋ฉ€๋ฆฌ ๋–จ์–ด์ ธ ์žˆ๊ณ , ์˜ค๋Š˜ ์ผ์–ด๋‚œ ์ผ์ด ๋‹ค์‹œ๋Š” ์ผ์–ด๋‚˜์ง€ ์•Š์„ ๊ฒƒ์ด๋ผ๊ณ  ์ƒ๊ฐํ•˜๋Š” ๊ฒƒ์ด ๋งˆ์Œ ํŽธํ•˜๊ธฐ ๋•Œ๋ฌธ์ž…๋‹ˆ๋‹ค. ์ด๋Ÿฌํ•œ โ€œ์ž˜๋ชป๋œ ํ”„๋กœํ•„์— ๋Œ€ํ•œ ๊ฑฐ์ง“ ์ด์•ผ๊ธฐโ€๋Š” ๋Œ€์ค‘์—๊ฒŒ ์•ˆ์‹ฌ์„ ์ฃผ๊ธฐ ์œ„ํ•ด ์ƒ๊ฒจ๋‚˜๊ณ  ์ธ๊ธฐ๋ฅผ ์–ป์ง€๋งŒ, ํ˜„์‹ค๊ณผ๋Š” ๊ฑฐ๋ฆฌ๊ฐ€ ๋ฉ€๋‹ค๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค.

ํ•ด๊ฒฐ์ฑ…: ๋‘ ๋‹จ๊ณ„ ์ ‘๊ทผ๊ณผ 75%์˜ ์—ญํ• 

ํŽ˜์ดํ”„ ๊ต์ˆ˜๋Š” ์ด ์œ„๊ธฐ๋ฅผ ํ•ด๊ฒฐํ•˜๊ธฐ ์œ„ํ•ด ๋‘ ๊ฐ€์ง€ ๋‹จ๊ณ„์˜ ์ ‘๊ทผ ๋ฐฉ์‹์ด ํ•„์š”ํ•˜๋‹ค๊ณ  ์ œ์•ˆํ•ฉ๋‹ˆ๋‹ค.

1. ์žฅ๊ธฐ์  ํ•ด๊ฒฐ์ฑ… (์ค‘๊ฐ„์„ ๊ฑฐ ์ดํ›„)

  • ์ธ๊ตฌํ†ต๊ณ„ํ•™์  ๋ณ€ํ™” ์†๋„ ์กฐ์ ˆ: ๊ตญ๊ฐ€์˜ ์ธ๊ตฌํ†ต๊ณ„ํ•™์  ๋ณ€ํ™”๋ฅผ ๋ง‰์„ ์ˆ˜๋Š” ์—†์ง€๋งŒ, ๊ทธ ์†๋„๋ฅผ ๋Šฆ์ถ”์–ด ์ •์น˜ ์ œ๋„๊ฐ€ ๋”ฐ๋ผ์žก์„ ์ˆ˜ ์žˆ๋„๋ก ํ•˜๋Š” ๊ฒƒ์ด ํ˜„๋ช…ํ•ฉ๋‹ˆ๋‹ค. ์ด๋Š” ๋ณธ์งˆ์ ์œผ๋กœ ์˜ค๋ฐ”๋งˆ ์‹œ๋Œ€์˜ ์ด๋ฏผ ์ •์ฑ…์„ ์ฑ„ํƒํ•˜๋Š” ๊ฒƒ์„ ์˜๋ฏธํ•ฉ๋‹ˆ๋‹ค.
  • ๋ถ€์˜ ์ง‘์ค‘ ์™„ํ™”: ์ƒ์œ„ 1%๋กœ์˜ ๋ถ€์˜ ์ด๋™์„ ๋ฉˆ์ถ”๊ฑฐ๋‚˜ ์—ญ์ „์‹œํ‚ค๋Š” ์ •์ฑ…์ด ํ•„์š”ํ•ฉ๋‹ˆ๋‹ค. ํ˜„์žฌ ์–‘๋‹น ๋ชจ๋‘ ์ด ๋ฌธ์ œ๋ฅผ ํ•ด๊ฒฐํ•˜์ง€ ๋ชปํ•˜๊ณ  ์žˆ์œผ๋ฉฐ, ๊ธฐ๋ถ€์ž๋“ค์—๊ฒŒ ๋ˆ์„ ์ฃผ๊ณ  ์žˆ๋Š” ์ƒํ™ฉ์ž…๋‹ˆ๋‹ค. ์šฐ๋ฆฌ๋Š” ์ด์— ๋Œ€ํ•ด ์†”์งํ•ด์•ผ ํ•ฉ๋‹ˆ๋‹ค.

2. ๋‹จ๊ธฐ์  ํ•ด๊ฒฐ์ฑ… (์ค‘๊ฐ„์„ ๊ฑฐ๊นŒ์ง€)

  • ์ง€๋„์ž๋“ค์˜ ์—ญํ• : ๋‹น์žฅ ์ค‘๊ฐ„์„ ๊ฑฐ๊นŒ์ง€๋Š” ์œ„์˜ ์žฅ๊ธฐ์  ์ •์ฑ…๋“ค์„ ๋ฐ”๊ฟ€ ์ˆ˜ ์—†์Šต๋‹ˆ๋‹ค. ๋”ฐ๋ผ์„œ ์•ž์œผ๋กœ ๋ช‡ ๋‹ฌ ๋™์•ˆ ์šฐ๋ฆฌ๊ฐ€ ํ•  ์ˆ˜ ์žˆ๋Š” ์œ ์ผํ•œ ์ •์ฑ…์€ ์ง€๋„์ž๋“ค์˜ ๋ณ€ํ™”์ž…๋‹ˆ๋‹ค. ์ง€๋„์ž๋“ค์€ ํ˜„์žฌ ๋ฐœ์ƒํ•˜๊ณ  ์žˆ๋Š” ํญ๋ ฅ์˜ ์†Œ์šฉ๋Œ์ด๋ฅผ ์ค„์ผ ์ˆ˜ ์žˆ๋Š” ์œ ์ผํ•œ ์‹ค์ œ ์›์ฒœ์ž…๋‹ˆ๋‹ค.
  • ํŽ˜์ดํ”„ ๊ต์ˆ˜๋Š” ํ•˜ํ‚ด ์ œํ”„๋ฆฌ์Šค(Hakeem Jeffries) ํ•˜์› ๋ฏผ์ฃผ๋‹น ์›๋‚ด๋Œ€ํ‘œ๊ฐ€ ๋„๋„๋“œ ํŠธ๋Ÿผํ”„ ์ „ ๋Œ€ํ†ต๋ น์—๊ฒŒ ์—ฐ๋ฝํ•˜์—ฌ โ€œ์–‘์ชฝ ์ง€๋„์ž๋“ค์ด ๋ชจ๋“  ํ˜•ํƒœ์˜ ์ •์น˜ ํญ๋ ฅ์„ ๊ทœํƒ„ํ•˜๋Š” ๊ณต๋™ ์˜์ƒ ์„ฑ๋ช…์„ ๋ฐœํ‘œํ•  ์ˆ˜ ์žˆ์„์ง€โ€ ์ œ์•ˆํ•˜๋Š” ๊ณผ๊ฐํ•œ ํ–‰๋™์„ ํ•ด์•ผ ํ•œ๋‹ค๊ณ  ๋งํ•ฉ๋‹ˆ๋‹ค.

์ด๋Ÿฌํ•œ ์ง€๋„์ž๋“ค์˜ ์ˆ˜์‚ฌ๊ฐ€ ํšจ๊ณผ๊ฐ€ ์žˆ์„๊นŒ์š”? ํŽ˜์ดํ”„ ๊ต์ˆ˜๋Š” 2024๋…„ ํŠธ๋Ÿผํ”„ ์ „ ๋Œ€ํ†ต๋ น์— ๋Œ€ํ•œ ์ฒซ ๋ฒˆ์งธ ์•”์‚ด ์‹œ๋„ ํ›„ ์กฐ ๋ฐ”์ด๋“  ๋Œ€ํ†ต๋ น์ด ๋ฐฑ์•…๊ด€ ์ง‘๋ฌด์‹ค์—์„œ ํญ๋ ฅ์„ ๊ทœํƒ„ํ•˜๋Š” ์—ฐ์„ค์„ ํ–ˆ์„ ๋•Œ, ๋ช‡ ๋‹ฌ ํ›„ ์ •์น˜ ํญ๋ ฅ ์ง€์ง€์œจ์ด ์ „๊ตญ์ ์œผ๋กœ 20% ๊ฐ์†Œํ–ˆ๋˜ ์‚ฌ๋ก€๋ฅผ ์–ธ๊ธ‰ํ•ฉ๋‹ˆ๋‹ค. ๋‘ ๋ฒˆ์งธ ์•”์‚ด ์‹œ๋„ ํ›„ ๋ฐ”์ด๋“  ๋Œ€ํ†ต๋ น๊ณผ ์นด๋ง๋ผ ํ•ด๋ฆฌ์Šค ๋ถ€ํ†ต๋ น ํ›„๋ณด๊ฐ€ ๋‹ค์‹œ ๊ฐ™์€ ์ˆ˜์‚ฌ๋ฅผ ์‚ฌ์šฉํ–ˆ์„ ๋•Œ๋„ ์ง€์ง€์œจ์ด ์ถ”๊ฐ€๋กœ ํ•˜๋ฝํ–ˆ์Šต๋‹ˆ๋‹ค. ์ด๋Š” ์ง€๋„์ž๋“ค์˜ ๋ฉ”์‹œ์ง€๊ฐ€ ์ตœ์†Œํ•œ ์ƒํ™ฉ์„ ์•…ํ™”์‹œํ‚ค์ง€ ์•Š๊ณ  ๊ธ์ •์ ์ธ ์˜ํ–ฅ์„ ๋ฏธ ๋ฏธ์ณค์Œ์„ ์‹œ์‚ฌํ•ฉ๋‹ˆ๋‹ค.

๊ตญ๋ฏผ์˜ ์„ ํƒ: 75%์˜ ์นจ๋ฌต์„ ๊นจ์•ผ ํ•  ๋•Œ

ํŽ˜์ดํ”„ ๊ต์ˆ˜๋Š” ์ด ๋ชจ๋“  ๊ตฌ์กฐ์  ์š”์ธ๊ณผ ์••๋ ฅ์—๋„ ๋ถˆ๊ตฌํ•˜๊ณ  **โ€œ์ •์น˜ ํญ๋ ฅ์˜ ๋ฏธ๋ž˜๋Š” ์„ ํƒ์˜ ๋ฌธ์ œโ€**๋ผ๊ณ  ๊ฐ•์กฐํ•ฉ๋‹ˆ๋‹ค. ์šฐ๋ฆฌ๋Š” ์ด๋Ÿฌํ•œ ๊ตฌ์กฐ์  ์••๋ ฅ์„ ๋ฌด์‹œํ•  ์ˆ˜๋„ ์žˆ๊ณ , ์ดํ•ดํ•˜๊ณ  ์•ž์œผ๋กœ ๋‚˜์•„๊ฐˆ ์ˆ˜๋„ ์žˆ์Šต๋‹ˆ๋‹ค.

๊ทธ๋Š” ์ •์น˜ ํญ๋ ฅ์„ ์ง€์ง€ํ•˜๋Š” ์ˆ˜์ฒœ๋งŒ ๋ช…์˜ ์‚ฌ๋žŒ๋“ค์—๊ฒŒ ์ดˆ์ ์„ ๋งž์ถฐ์™”์ง€๋งŒ, ๋” ์ค‘์š”ํ•œ ๊ฒƒ์€ **โ€œ๋ฏธ๊ตญ ์ธ๊ตฌ์˜ 75%๊ฐ€ ์ •์น˜ ํญ๋ ฅ์„ ํ˜์˜คํ•œ๋‹คโ€**๋Š” ์‚ฌ์‹ค์ž…๋‹ˆ๋‹ค. ์ด 75%๋Š” ํ–‰๋™์— ๋‚˜์„œ์•ผ ํ•ฉ๋‹ˆ๋‹ค. ๊ทธ๋“ค์€ ์ •์น˜ ํญ๋ ฅ์— ๋Œ€ํ•ด ๋ถˆํ‰ํ•˜๊ณ , ์ƒ์›์˜์›, ํ•˜์›์˜์›, ์‹œ์žฅ, ๋Œ€ํ†ต๋ น์—๊ฒŒ ํŽธ์ง€๋ฅผ ๋ณด๋‚ด์•ผ ํ•ฉ๋‹ˆ๋‹ค. โ€œ๋ฏผ์ฃผ๋‹น๊ณผ ๊ณตํ™”๋‹น ์ง€๋„์ž๋“ค์ด ๊ณต๋™์œผ๋กœ ์ •์น˜ ํญ๋ ฅ์„ ๊ทœํƒ„ํ•ด์•ผ ํ•  ๋•Œ์ž…๋‹ˆ๋‹คโ€๋ผ๊ณ  ๋ชฉ์†Œ๋ฆฌ๋ฅผ ๋‚ด์•ผ ํ•ฉ๋‹ˆ๋‹ค.

โ€œ์šฐ๋ฆฌ๋Š” ๋” ๋‚˜์€ ๋ฏธ๋ž˜๋ฅผ ๊ฐ€์งˆ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค. ๊ทธ๊ฒƒ์€ ์„ ํƒ์ž…๋‹ˆ๋‹ค. ๊ทธ๋Ÿฌ๋‚˜ ๊ทธ๊ฒƒ์€ ์šฐ๋ฆฌ ๊ตญ๋ฏผ์ด, ๊ทธ๋ฆฌ๊ณ  75%์˜