YouTube Digest

April 21, 2026

Listen to this issue

English

Based on โ€œHow Intercom 2Xโ€™d engineering velocity with Claude Code | Brian Scanlanโ€ from How I AI Watch the original video

Unleashing the Code: How Intercom Doubled Engineering Velocity with AI and Reimagined the Software Factory

In the rapidly evolving landscape of artificial intelligence, many companies are grappling with how to integrate these powerful tools effectively. While some express skepticism about AIโ€™s true impact on productivity, Intercom, the customer messaging platform, stands as a compelling counter-narrative. Under the guidance of senior principal engineer Brian Scanlan, Intercomโ€™s R&D department has achieved a remarkable feat: doubling their engineering throughput in just nine months, largely thanks to a full-throttle adoption of AI, particularly Claude Code.

This isnโ€™t just about incremental gains; itโ€™s a fundamental shift in how software is built, challenging long-held beliefs about developer productivity, code quality, and the very nature of engineering work.

The Moment of Inflection: Imagination, Not Tools, Is the Barrier

For years, the promise of AI in software development felt just out of reach. Code completion tools offered minor conveniences, but nothing truly transformative. Brian Scanlan recounts a pivotal moment around November-December last year, coinciding with the release of more advanced models like Anthropicโ€™s Opus 46 and GPT-4U. โ€œSuddenly you started realizing that you have to think bigger about things,โ€ Scanlan explains, โ€œor that your imagination is now the barrier, not the tool.โ€

This breakthrough wasnโ€™t just internal; it was a collective awakening. Scanlan describes the atmosphere during the Christmas break, with โ€œeverybody go wild on Twitter X,โ€ witnessing developers achieve unprecedented levels of output. The message was clear: โ€œEverythingโ€™s changed.โ€

Intercomโ€™s journey was fueled by an internal AI-first mandate that permeated their product strategy. Having already transformed their customer-facing products with AI, the expectation for similar innovation within engineering was high. This created an environment where impatience for AI adoption was a driving force, pushing the team to go โ€œall-inโ€ rather than dabble.

Measuring the Unmeasurable: 2x Velocity and the Path to 10x

To quantify their progress, Intercomโ€™s CTO, Darra, set an ambitious goal: double the throughput of R&D. They chose a simple, albeit sometimes debated, metric: merged Pull Requests (PRs) per R&D head. This metric, encompassing product managers, designers, and TPMs alongside engineers, revealed a dramatic โ€œnumber goes upโ€ chart. Within nine months, the team saw a 2x increase in PR throughput.

This isnโ€™t merely about engineers typing faster; itโ€™s about unlocking โ€œthe physical limits of my ability to type code,โ€ as Claire Vote, the host of โ€œHow I AI,โ€ eloquently puts it. The shift is so profound that Scanlan now asks, โ€œWhy canโ€™t it be 10x?โ€ This ambitious outlook stems from a belief that if an organization fully commits, prepares its team, and refines its codebase, it will see significant improvements in product quality and developer experience.

Addressing the common skepticism that such metrics might be โ€œgamedโ€ by shipping smaller, less impactful changes, Scanlan emphasizes Intercomโ€™s high-trust environment. The goal isnโ€™t just speed; itโ€™s about enabling engineers to do more meaningful work, faster and with greater enjoyment. Indeed, Scanlan shares, โ€œIโ€™ve been having the most amount of fun in my career over the last 3 months.โ€

Beyond Speed: The Surprising Uplift in Code Quality

A frequent concern among engineers and leaders alike is whether increased velocity comes at the cost of quality. Will AI-generated code be โ€œslopโ€ or โ€œgarbageโ€? Intercomโ€™s experience suggests the opposite.

Theyโ€™ve observed a consistent decrease in the time it takes from the first line of code written to a feature being announced on their news channel. More significantly, Intercom collaborated with a research group at Stanford, providing them with their internal data. The Stanford teamโ€™s independent measures of code quality indicated that code quality was actually improving.

This counter-intuitive finding highlights a crucial benefit of AI-driven development: the capacity to tackle technical debt and improve the codebase. As Scanlan and Vote discuss, businesses often have limited capacity for โ€œinternal projectsโ€ like improving code quality, as these donโ€™t directly generate revenue. However, when the โ€œcost of doing that compressesโ€ due to AIโ€™s efficiency, organizations can afford to invest in developer experience, security, compliance, maintainability, flaky tests, and CI/CD improvements.

This leads to a powerful piece of advice for CTOs and VPs of engineering: โ€œEverything you hate about the codebase, go spend a month fixing and see how fast we can speedrun that. Thatโ€™s going to feel really good.โ€ This strategic investment in core engineering health, made feasible by AI, ultimately unlocks even greater velocity and higher quality.

The Software Factory: Engineering the Golden Path

Intercomโ€™s success isnโ€™t just about adopting AI; itโ€™s about intentionally engineering their entire software delivery process around it, moving towards what Scanlan calls a โ€œsoftware factory.โ€ This involves treating the engineering organization itself like a product, complete with rigorous design, measurement, and continuous improvement.

One striking example is how Intercom addressed the issue of pull request (PR) descriptions. Initially, AI-generated PR descriptions were โ€œterrible,โ€ merely regurgitating code changes rather than explaining intent โ€“ the truly valuable information for human reviewers. Intercomโ€™s solution involved:

  1. Defining Quality: They established what a โ€œgoodโ€ PR description should look like.
  2. LLM Judge: An AI model was trained to evaluate PR descriptions, revealing a negative trend in quality with early AI adoption.
  3. Custom Skill: They developed a โ€œcreate PRโ€ skill for Claude Code that leverages session context to generate high-quality, intent-driven descriptions.
  4. Enforcement: This skill was integrated as a mandatory hook. If an engineer or agent tries to open a PR without using the skill, itโ€™s blocked.

This approach mirrors the determinism of modern CI/CD pipelines, but applied upstream to the code-writing process itself. โ€œWeโ€™re on this movement towards a software factory,โ€ Scanlan explains, where predictability, reliability, and consistent quality are paramount. This ensures that while engineers move faster, they still adhere to the high standards of a mature, 15-year-old SaaS company.

The Invisible Infrastructure: Telemetry and Personalized Feedback

Underpinning Intercomโ€™s โ€œsoftware factoryโ€ is a sophisticated system for monitoring and improving AI usage. They donโ€™t โ€œfly blind.โ€

  1. Skill Telemetry with Honeycomb: Every internal AI skill is instrumented with event-level telemetry, sending data to Honeycomb. This allows individual skill developers to see how often their skills are invoked, by whom, and when, fostering a data-driven approach to skill development.
  2. Session Data Analysis with S3: All raw Claude Code session data (the chat logs) is anonymized, uploaded to S3, and then analyzed. This allows Intercom to identify broader organizational trends, common pitfalls, and areas where training or new skills are needed.
  3. Personalized Feedback: Based on this session data, Intercom developed a simple internal tool that provides personalized feedback to individual engineers on their Claude Code usage. This helps new hires or those struggling to understand how to optimize their interactions, supporting a culture of continuous learning and self-improvement.

This robust telemetry system ensures that Intercom can identify bottlenecks, improve their internal tools, and provide targeted support, preventing the common problem of โ€œthrowing an API key and saying best of luck.โ€

The Elephant in the Room: AI Costs

The exponential growth in AI usage inevitably raises questions about cost. Scanlan readily admits that their AI bill โ€œlooks exactly like thisโ€ (referencing the steep velocity growth chart). โ€œItโ€™s like hiring whole new offices of people,โ€ he notes.

Intercomโ€™s current strategy is to prioritize speed over cost optimization. Their attitude is, โ€œeveryone just turn on Opus for everythingโ€ฆ going as fast as possible and caring about the bill later.โ€ This reflects a strategic investment mindset, believing that the significant benefits gained from rapid innovation and increased throughput outweigh the immediate costs. However, Scanlan acknowledges this might not be feasible for every business and that future optimization will be necessary if costs continue at the current rate.

The Future: Agent-First and Higher-Level Concerns

Looking ahead, Intercom envisions a future where โ€œall technical work will become agent first.โ€ This means AI agents will handle the foundational, repetitive tasks, freeing human engineers to operate at a higher level, focusing on complex problem-solving, strategic design, and truly innovative features.

The shift is profound: instead of engineers spending time on boilerplate code or debugging minor issues, agents will take on this โ€œbasic work,โ€ allowing humans to โ€œmove up to higher level to be able to like work on higher level concerns or just getting more stuff built more stuff out there or higher quality.โ€

Intercomโ€™s journey with Claude Code is more than just a case study in productivity; itโ€™s a blueprint for reimagining the entire software development lifecycle. By embracing AI comprehensively, treating the organization as a product, meticulously measuring outcomes, and proactively engineering for quality, Intercom has not only doubled its engineering velocity but has also cultivated a more engaging, productive, and ultimately more fun environment for its engineers. For any organization looking to leverage AI in engineering, Intercomโ€™s experience offers invaluable lessons and a compelling vision for the future.


Based on โ€œInside the Five Days That Remade the Supreme Courtโ€ from New York Times Podcasts Watch the original video

The Secret Birth of the Supreme Courtโ€™s Shadow Power

For the better part of a decade, the U.S. Supreme Court has operated with a parallel, often perplexing system for issuing major rulings โ€“ one that bypasses the traditional, deliberate processes that define American justice. This increasingly powerful mechanism, dubbed the โ€œshadow docket,โ€ has been responsible for consequential decisions on everything from immigration policy to presidential authority, often with little to no public explanation. Now, a groundbreaking New York Times investigation has pierced this veil of secrecy, revealing the precise, dramatic five days in 2016 when this expedited system was born, effectively remaking the nationโ€™s highest court.

At the heart of this revelation are 16 pages of confidential correspondence among the justices themselves, offering an unprecedented look into their private deliberations. These documents, unearthed by Times journalists Jodie Caner and Adam Liptac, allow us to โ€œeavesdrop on the justices at the exact moment that they are abandoning time-tested norms of judicial procedure and backing themselves into a new way of doing business.โ€

Understanding the Shadow Docket

To truly grasp the significance of this shift, one must understand the stark contrast between the Supreme Courtโ€™s traditional โ€œmerits docketโ€ and its shadowy counterpart. The merits docket is the familiar image of the court: justices meticulously select cases, receive multiple rounds of detailed briefs, hear extensive oral arguments, engage in in-person deliberations, and then exchange numerous drafts of lengthy, reasoned opinions, concurrences, and dissents. This painstaking process, often stretching over a year, culminates in decisions that can be hundreds of pages long, providing binding law and clear guidance to lower courts and the nation.

The shadow docket, however, short-circuits all of this. It operates with extreme speed, relying on thin briefs, no oral arguments, and no in-person deliberations. Its rulings are typically brief, often consisting of scant or even no reasoning at all. Yet, over the last ten years, it has become a major part of the courtโ€™s business, particularly during the Trump administration, where it was used to grant the president enormous leeway on issues like immigration, government spending, and agency power. Critics have consistently warned that this exponential increase in use, combined with the lack of transparent reasoning, raises critical questions: Is the court rushing to rule based on gut instinct, personal pique, or partisan impulse โ€“ precisely the elements a slow, deliberate, and judicious system is designed to avoid?

Defenders of the shadow docket argue that emergency orders have a long history, typically reserved for truly time-sensitive matters like death penalty cases or election disputes. They also contend that these orders are merely temporary, designed to maintain the status quo while cases proceed through lower courts. However, as Caner and Liptac point out, this โ€œtemporaryโ€ label often masks a very different reality. If the court allows the deportation of thousands of people, or the withholding of aid money, or the firing of employees, those actions, though nominally temporary, are often irreversible and conclusive in their practical effect. They are, in essence, final decisions made without the benefit of the courtโ€™s usual rigorous process.

The Unprecedented Request: Obamaโ€™s Clean Power Plan

The crucible for this seismic shift was a case in 2016 concerning President Barack Obamaโ€™s Clean Power Plan. Facing legislative gridlock on climate change, Obama had instructed the Environmental Protection Agency (EPA) to issue regulations that would fundamentally move the American power system away from coal. This initiative, predictably, enraged industry groups and โ€œred states,โ€ who challenged it in the DC Circuit Court. They asked the court not only to declare the plan unlawful but also to immediately halt its implementation. The DC Circuit agreed to fast-track arguments on the planโ€™s legality but refused to pause its rollout.

In an โ€œunprecedented request,โ€ the challengers then turned directly to the Supreme Court. They asked the justices to freeze the Clean Power Plan before any lower court had even ruled on its lawfulness โ€“ a move that everyone involved recognized as extraordinary. At this moment, the Supreme Court was evenly split, a 5-4 court leaning conservative, but with Justice Anthony Kennedy, a Republican appointee, serving as the crucial swing vote. Kennedy, famously โ€œpersuadable,โ€ had recently authored the majority opinion legalizing gay marriage, making the courtโ€™s direction unpredictable.

Five Days That Shook the Court

The emergency request landed in the chambers of Chief Justice John Roberts, who oversees the DC Circuit. The ordinary expectation was a swift denial, or at most, a collective denial by his colleagues. What transpired instead was a five-day sprint of internal memos that would forever alter the courtโ€™s operational landscape.

Day One: Robertsโ€™s Salvo Chief Justice Roberts initiated the debate with a three-page, single-spaced memo. He argued forcefully that Obamaโ€™s plan must be halted due to the โ€œenormous burdensโ€ it would impose on states and the coal industry. He claimed there was โ€œno time to wasteโ€ and questioned the EPAโ€™s authority under the Clean Air Act, invoking the โ€œmajor questions doctrineโ€ โ€“ a principle suggesting that unless Congress clearly grants an agency vast power, that power doesnโ€™t exist.

But beyond legal arguments, Robertsโ€™s memo betrayed a deep-seated grievance. He felt the EPA had โ€œtrickedโ€ the court just months earlier in a mercury emissions case. In that instance, the court had ruled against the EPA after a three-year litigation, but by then, the regulation had effectively gone into effect, rendering the ruling somewhat moot. Roberts was โ€œpeeved,โ€ โ€œirked,โ€ and determined โ€œnot to let it happen again.โ€ He viewed the traditional slow legal process as enabling the Obama administration to be โ€œsneakyโ€ in implementing its regulations.

Day One: Breyerโ€™s Counterpoint Justice Stephen Breyer, a Democratic appointee, immediately responded, laying out a contrasting chronology. He pointed out that the Clean Power Plan didnโ€™t require industry action for six years, with total compliance not due until 2030. There was, in his view, โ€œplenty of time to do this in the usual course.โ€ Breyer found the courtโ€™s intervention โ€œquite unusualโ€ and saw no urgent need to bypass the DC Circuit, advocating for the lower court to be allowed to finish its work.

Day Two: Robertsโ€™s Insistence Robertsโ€™s reply the next day revealed a growing irritation and an unyielding insistence on blocking the plan. He dismissed Breyerโ€™s procedural concerns: โ€œI recognize that the posture of this stay request is not typical, but review is sought of what has been described as the most expensive regulation ever imposed on the power sector.โ€ The Chief Justiceโ€™s impatience was palpable; he was โ€œready to rule now.โ€ He believed the court would ultimately strike down the Clean Power Plan, stating it was โ€œhighly unlikely to surviveโ€ a full review. He saw no reason to let the process play out when he already knew the answer.

This memo made it clear that Roberts was engaged in a power struggle with the Obama administration. He explicitly stated, โ€œI am of the mind that a rule designed to transform a substantial swath of the nationโ€™s economy should be tested by this court before it is presented as a fait accompli.โ€ As Liptac observes, this was not the โ€œeven-handed, magisterial toneโ€ typically seen in public; here, Roberts was โ€œacting as a bulldozer.โ€

Day Three: Kagan Sounds the Alarm Justice Elena Kagan, another Democratic appointee, filed an even more direct memo, opposing the Chief Justiceโ€™s stance. She called the relief sought โ€œuniqueโ€ and used the word โ€œunprecedentedโ€ to describe what Roberts wanted to do. Kagan stressed the complexity of the case, arguing that it involved โ€œa complex statutory and regulatory regimeโ€ that demanded more time and consideration. โ€œThis is weird. Are we sure we want to do this?โ€ she seemed to ask, highlighting the gravity of the situation.

Day Four: Alitoโ€™s Existential Threat Justice Samuel Alito, a Republican appointee, then weighed in, echoing Robertsโ€™s sense of insult and grievance. He declared that a โ€œfailure to stay this rule threatens to render our ability to provide meaningful judicial review and by extension our institutional legitimacy a nullity.โ€ Alito framed the issue as an โ€œalmost existential threatโ€ to the court itself, portraying the Obama administration as attempting to sideline the justices.

Whatโ€™s extraordinary about these private exchanges, the journalists emphasize, is how frankly the justices reveal their โ€œactual agendas.โ€ Unlike their public decisions, which often present a โ€œmaskโ€ of confining themselves to facts and legal materials, these memos expose the human element: grievances with an administration, fears about the courtโ€™s institutional place, and concerns about legitimacy โ€“ factors not typically associated with purely legal reasoning.

Day Five: Kennedyโ€™s Decisive Whisper The entire debate, as predicted, came down to Justice Anthony Kennedy. On February 9th, the fifth day, Kennedy sent a terse, three-sentence note: he was voting with the Chief. The debate was over. Within hours, the Supreme Court issued its order, a single paragraph of โ€œlegal boilerplateโ€ with no explanation, blocking President Obamaโ€™s signature environmental initiative.

The Enduring Legacy of the Shadow Docket

The insights gleaned from these confidential memos profoundly vindicate the long-standing criticisms of the shadow docket. This was โ€œnot the court doing A+ work,โ€ Caner and Liptac conclude. Instead, it was a court โ€œthrowing ideas around, seeming to be motivated by grievances against a presidentโ€ฆgetting snippy with each other and just in general not doing the kind of work we associate with the nationโ€™s highest court.โ€ The justices were disregarding time-tested procedures without seemingly considering the long-term implications. As Kagan noted, it was โ€œunprecedented,โ€ but no one seemed to ask: โ€œwhere is it going to lead?โ€

Where it led was to an explosion of emergency applications under the Trump administration, with the court often quickly ruling in favor of presidential initiatives on major questions, a stark contrast to the Obama era. Political scientists have observed more partisan voting on the shadow docket than on the merits docket, suggesting that when acting fast, justices may rely more heavily on partisan impulses. For instance, in the Biden years, the court initially voted against Biden on three emergency applications, only to rule for him when those same cases returned to the merits docket for full consideration. This pattern suggests that deliberation truly โ€œdampens partisan impulses.โ€

Beyond partisan concerns, the rise of the shadow docket poses a fundamental risk to the courtโ€™s legitimacy. Unlike elected officials, Supreme Court justices serve for life and derive their legitimacy not from votes, but from public trust. The act of writing a reasoned opinion is a judgeโ€™s way of saying, โ€œHereโ€™s why you should trust me; hereโ€™s my work.โ€ When the court increasingly abandons this practice, issuing barely explained rulings on major national issues, it erodes that trust. As the Supreme Courtโ€™s public approval ratings plumb historic lows, its increasing reliance on the shadow docket to make pivotal decisions only exacerbates this critical problem, threatening the very foundation of its authority.


ํ•œ๊ตญ์–ด

โ€œHow Intercom 2Xโ€™d engineering velocity with Claude Code | Brian Scanlanโ€ โ€” How I AI ๊ธฐ๋ฐ˜ ๊ธฐ์‚ฌ ์›๋ณธ ์˜์ƒ ๋ณด๊ธฐ

AI๊ฐ€ ์—”์ง€๋‹ˆ์–ด๋ง์˜ โ€˜ํƒ€์ดํ•‘ ํ•œ๊ณ„โ€™๋ฅผ ํ—ˆ๋ฌผ๋‹ค: ์ธํ„ฐ์ฝค, ํด๋กœ๋“œ ์ฝ”๋“œ๋กœ ๊ฐœ๋ฐœ ์†๋„ 2๋ฐฐ ๋‹ฌ์„ฑ ๋น„๊ฒฐ

[์„œ์šธ, ๋Œ€ํ•œ๋ฏผ๊ตญ] ์—”์ง€๋‹ˆ์–ด๋ง ํŒ€์˜ ์ƒ์‚ฐ์„ฑ์„ ๋‘ ๋ฐฐ๋กœ ๋Œ์–ด์˜ฌ๋ฆฐ๋‹ค๋Š” ๊ฒƒ์€ ๋Œ€๋‹ค์ˆ˜ ๊ธฐ์ˆ  ๊ธฐ์—…์˜ ๊ฟˆ์ผ ๊ฒƒ์ž…๋‹ˆ๋‹ค. ํ•˜์ง€๋งŒ ๊ณ ๊ฐ ๊ด€๊ณ„ ๊ด€๋ฆฌ(CRM) ๋ฐ ๋ฉ”์‹œ์ง• ํ”Œ๋žซํผ ๋ถ„์•ผ์˜ ์„ ๋‘ ์ฃผ์ž์ธ ์ธํ„ฐ์ฝค(Intercom)์€ ์ด ๊ฟˆ์„ ํ˜„์‹ค๋กœ ๋งŒ๋“ค์—ˆ์Šต๋‹ˆ๋‹ค. AI ๋„๊ตฌ, ํŠนํžˆ ์•ค์Šค๋กœํ”ฝ(Anthropic)์˜ ํด๋กœ๋“œ ์ฝ”๋“œ(Claude Code)๋ฅผ ๋„์ž…ํ•œ ์ง€ ๋ถˆ๊ณผ ๋ช‡ ๋‹ฌ ๋งŒ์— ์—ฐ๊ตฌ ๊ฐœ๋ฐœ(R&D) ๋ถ€์„œ์˜ ํ’€ ๋ฆฌํ€˜์ŠคํŠธ(PR, Pull Request) ์ฒ˜๋ฆฌ๋Ÿ‰์„ 2๋ฐฐ๋กœ ๋Š˜๋ฆฌ๋Š” ๋†€๋ผ์šด ์„ฑ๊ณผ๋ฅผ ๋‹ฌ์„ฑํ•œ ๊ฒƒ์ž…๋‹ˆ๋‹ค.

โ€˜How I AIโ€™ ํŒŸ์บ์ŠคํŠธ์— ์ถœ์—ฐํ•œ ์ธํ„ฐ์ฝค์˜ ์„ ์ž„ ์ˆ˜์„ ์—”์ง€๋‹ˆ์–ด ๋ธŒ๋ผ์ด์–ธ ์Šค์บ”๋Ÿฐ(Brian Scanlan)์€ AI๊ฐ€ ์–ด๋–ป๊ฒŒ ์—”์ง€๋‹ˆ์–ด๋ง์˜ ๊ทผ๋ณธ์ ์ธ ๋ณ€ํ™”๋ฅผ ์ด๋Œ๊ณ  ์žˆ์œผ๋ฉฐ, ์ธํ„ฐ์ฝค์ด ์ด๋ฅผ ํ†ตํ•ด ์–ด๋–ค ํ˜์‹ ์„ ์ด๋ฃจ์—ˆ๋Š”์ง€ ์ƒ์„ธํžˆ ๊ณต์œ ํ–ˆ์Šต๋‹ˆ๋‹ค. ์ด๋ฒˆ ๊ธฐ์‚ฌ์—์„œ๋Š” ์ธํ„ฐ์ฝค์˜ ์„ฑ๊ณต ์‚ฌ๋ก€๋ฅผ ํ†ตํ•ด AI ์‹œ๋Œ€ ์—”์ง€๋‹ˆ์–ด๋ง ์กฐ์ง์˜ ๋ฏธ๋ž˜์™€ ๊ทธ ์‹ค์งˆ์ ์ธ ๊ตฌํ˜„ ์ „๋žต์„ ์‹ฌ์ธต์ ์œผ๋กœ ํƒ๊ตฌํ•ฉ๋‹ˆ๋‹ค.


1. AI๊ฐ€ ๊ฐ€์ ธ์˜จ โ€˜์ƒ์ƒ๋ ฅ์˜ ์‹œ๋Œ€โ€™: ์—”์ง€๋‹ˆ์–ด๋ง์˜ ์ƒˆ๋กœ์šด ์ง€ํ‰

๋ธŒ๋ผ์ด์–ธ ์Šค์บ”๋Ÿฐ์€ AI๊ฐ€ ์—”์ง€๋‹ˆ์–ด๋ง์— ๊ฐ€์ ธ์˜จ ๊ฐ€์žฅ ํฐ ๋ณ€ํ™” ์ค‘ ํ•˜๋‚˜๋กœ โ€œ์ƒ์ƒ๋ ฅ์ด ์žฅ๋ฒฝ์ด์ง€, ๋„๊ตฌ๊ฐ€ ์•„๋‹ˆ๋‹ค(imagination is now the barrier, not the tool)โ€œ๋ผ๋Š” ์ ์„ ๊ผฝ์•˜์Šต๋‹ˆ๋‹ค. ๊ทธ๋Š” โ€œ๋ง ๊ทธ๋Œ€๋กœ ์ฝ”๋“œ๋ฅผ ํƒ€์ดํ•‘ํ•˜๋Š” ๋Šฅ๋ ฅ์˜ ๋ฌผ๋ฆฌ์  ํ•œ๊ณ„๋ฅผ AI๊ฐ€ ํ•ด์ œํ–ˆ๋‹คโ€๊ณ  ๊ฐ•์กฐํ•˜๋ฉฐ, ๊ฐœ๋ฐœ์ž๋“ค์ด ๋” ์ด์ƒ ๋„๊ตฌ์— ๋งž์ถฐ ์ž‘์—…ํ•˜๋Š” ๋Œ€์‹  ์•„์ด๋””์–ด ๊ตฌํ˜„์— ์ง‘์ค‘ํ•  ์ˆ˜ ์žˆ๊ฒŒ ๋˜์—ˆ๋‹ค๊ณ  ์„ค๋ช…ํ–ˆ์Šต๋‹ˆ๋‹ค.

์ธํ„ฐ์ฝค์€ ์ด๋ฏธ ์ œํ’ˆ ์ธก๋ฉด์—์„œ AI๋ฅผ ์ ๊ทน์ ์œผ๋กœ ์ˆ˜์šฉํ•˜๋ฉฐ โ€˜AI ์šฐ์„ (AI-first)โ€™ ์ „๋žต์„ ํŽผ์ณ์™”์Šต๋‹ˆ๋‹ค. ์ด๋Ÿฌํ•œ ๋ฐฐ๊ฒฝ ์†์—์„œ ์—”์ง€๋‹ˆ์–ด๋ง ํŒ€ ์—ญ์‹œ AI์˜ ์ž ์žฌ๋ ฅ์„ ๋ช…ํ™•ํžˆ ์ธ์ง€ํ•˜๊ณ  ์žˆ์—ˆ์Šต๋‹ˆ๋‹ค. ํŠนํžˆ ์ง€๋‚œ 2023๋…„ 11์›”๊ณผ 12์›”๊ฒฝ, ์•ค์Šค๋กœํ”ฝ์˜ ์˜คํ‘ธ์Šค(Opus) 46 ๋ชจ๋ธ(GPT-4U ๋ชจ๋ธ์„ ์ง€์นญํ•˜๋Š” ๊ฒƒ์œผ๋กœ ๋ณด์ž„)์˜ ์ถœ์‹œ์™€ ํ•จ๊ป˜ โ€œ๋ชจ๋“  ๊ฒƒ์ด ๋ณ€ํ–ˆ๋‹คโ€๊ณ  ์Šค์บ”๋Ÿฐ์€ ํšŒ๊ณ ํ•ฉ๋‹ˆ๋‹ค. ์ด ์‹œ์ ์„ ๊ธฐ์ ์œผ๋กœ AI ๋ชจ๋ธ์˜ ์„ฑ๋Šฅ์ด ๊ธ‰๊ฒฉํžˆ ํ–ฅ์ƒ๋˜๋ฉด์„œ ๊ฐœ๋ฐœ์ž๋“ค์€ AI๋ฅผ ๋‹จ์ˆœํ•œ ์ž๋™ ์™„์„ฑ ๋„๊ตฌ๊ฐ€ ์•„๋‹Œ, ์•„์ด๋””์–ด๋ฅผ ํ˜„์‹ค๋กœ ๋งŒ๋“œ๋Š” ๊ฐ•๋ ฅํ•œ ํ˜‘๋ ฅ์ž๋กœ ์ธ์‹ํ•˜๊ธฐ ์‹œ์ž‘ํ–ˆ์Šต๋‹ˆ๋‹ค.

ํฅ๋ฏธ๋กญ๊ฒŒ๋„, ์Šค์บ”๋Ÿฐ์€ ์—ฐ๋ง ํœด๊ฐ€ ๊ธฐ๊ฐ„ ๋™์•ˆ ๋งŽ์€ ๊ฐœ๋ฐœ์ž๋“ค์ด ํด๋กœ๋“œ ์ฝ”๋“œ๋ฅผ ํ™œ์šฉํ•ด ๊ฐœ์ธ ํ”„๋กœ์ ํŠธ๋ฅผ ์ง„ํ–‰ํ•˜๋ฉฐ ์—„์ฒญ๋‚œ ์†๋„๋กœ ์ƒˆ๋กœ์šด ๊ธฐ์ˆ ์„ ์Šต๋“ํ•˜๋Š” ํ˜„์ƒ์„ ๋ชฉ๊ฒฉํ–ˆ๋‹ค๊ณ  ์–ธ๊ธ‰ํ–ˆ์Šต๋‹ˆ๋‹ค. ์ด๋Š” ๊ธฐ์—…๋“ค์ด ์ง์›๋“ค์—๊ฒŒ ์ถฉ๋ถ„ํ•œ ์œ ๊ธ‰ ํœด๊ฐ€(PTO, Paid Time Off)๋ฅผ ์ œ๊ณตํ•จ์œผ๋กœ์จ AI ํ™œ์šฉ ๋Šฅ๋ ฅ์„ ์ž์—ฐ์Šค๋Ÿฝ๊ฒŒ ํ–ฅ์ƒ์‹œํ‚ฌ ์ˆ˜ ์žˆ๋Š” โ€˜AI ๋ฆฌํ„ฐ๋Ÿฌ์‹œ ํ•ดํ‚นโ€™ ์ „๋žต์ด ๋  ์ˆ˜ ์žˆ์Œ์„ ์‹œ์‚ฌํ•ฉ๋‹ˆ๋‹ค.

2. 2๋ฐฐ ๊ฐ€์†, ๊ทธ ์ด์ƒ์˜ ๋ชฉํ‘œ: ๋ฐ์ดํ„ฐ๋กœ ์ฆ๋ช…ํ•˜๋‹ค

์ธํ„ฐ์ฝค์˜ CTO ๋‹ค๋ผ(Darra)๋Š” R&D ์ฒ˜๋ฆฌ๋Ÿ‰ 2๋ฐฐ ์ฆ๊ฐ€๋ผ๋Š” ๋ช…ํ™•ํ•œ ๋ชฉํ‘œ๋ฅผ ์ œ์‹œํ–ˆ๊ณ , ์ธํ„ฐ์ฝค ํŒ€์€ ์ด๋ฅผ ๋‹ฌ์„ฑํ•˜๊ธฐ ์œ„ํ•ด AI ๋„์ž…์„ โ€˜์ œํ’ˆโ€™์ฒ˜๋Ÿผ ๋‹ค๋ฃจ์—ˆ์Šต๋‹ˆ๋‹ค. ์‚ฌ์šฉ์ž ํ”ผ๋“œ๋ฐฑ์„ ์ˆ˜์ง‘ํ•˜๊ณ , ๋„๊ตฌ ์‚ฌ์šฉ ๋ฐฉ์‹์„ ๋ถ„์„ํ•˜๋ฉฐ, ํ…”๋ ˆ๋ฉ”ํŠธ๋ฆฌ(telemetry) ๋ฐ์ดํ„ฐ๋ฅผ ํ™œ์šฉํ•ด ์„ฑ๊ณผ๋ฅผ ์ธก์ •ํ–ˆ์Šต๋‹ˆ๋‹ค.

ํ•ต์‹ฌ ์ธก์ • ์ง€ํ‘œ: ์ธํ„ฐ์ฝค์€ โ€˜R&D ์ธ์›๋‹น ๋ณ‘ํ•ฉ๋œ PR(merged PRs per R&D head)โ€˜์„ ์ฃผ์š” ์ง€ํ‘œ๋กœ ์‚ผ์•˜์Šต๋‹ˆ๋‹ค. ์ด๋Š” ๋‹จ์ˆœํžˆ ์†Œํ”„ํŠธ์›จ์–ด ์—”์ง€๋‹ˆ์–ด๋ฟ๋งŒ ์•„๋‹ˆ๋ผ ์ œํ’ˆ ๊ด€๋ฆฌ์ž(PM, Product Manager), ๋””์ž์ด๋„ˆ, ๊ธฐ์ˆ  ํ”„๋กœ์ ํŠธ ๊ด€๋ฆฌ์ž(TPM, Technical Project Manager) ๋“ฑ ๋ชจ๋“  R&D ๊ตฌ์„ฑ์›์˜ ๊ธฐ์—ฌ๋ฅผ ํฌํ•จํ•˜๋Š” ํฌ๊ด„์ ์ธ ์ง€ํ‘œ์ž…๋‹ˆ๋‹ค.

์„ฑ๊ณผ ๋ฐ ๋„์ „:

  • 2๋ฐฐ์˜ ์ฒ˜๋ฆฌ๋Ÿ‰: 9๊ฐœ์›” ์ „๊ณผ ๋น„๊ตํ•˜์—ฌ ์—”์ง€๋‹ˆ์–ด๋ง ํŒ€์˜ ์ฒ˜๋ฆฌ๋Ÿ‰์ด 2๋ฐฐ ์ฆ๊ฐ€ํ–ˆ์Šต๋‹ˆ๋‹ค.
  • CI ์‹œ์Šคํ…œ์˜ ๋ณ€ํ™”: PR ์ˆ˜๊ฐ€ ๊ธ‰์ฆํ•˜์ž ๊ธฐ์กด CI(Continuous Integration) ์‹œ์Šคํ…œ์ด ๊ณผ๋ถ€ํ•˜๋กœ โ€˜๋…น์•„๋‚ด๋ฆดโ€™ ์ •๋„์˜€์Šต๋‹ˆ๋‹ค. ์ธํ„ฐ์ฝค์€ ๋ณ‘๋ชฉ ํ˜„์ƒ์„ ํ•ด๊ฒฐํ•˜๊ธฐ ์œ„ํ•ด CI ์‹œ์Šคํ…œ ์„ฑ๋Šฅ์„ ๊ฐœ์„ ํ–ˆ๊ณ , ์ด์ œ๋Š” ์ฝ”๋“œ ๋ฆฌ๋ทฐ๊ฐ€ ์ƒˆ๋กœ์šด ๋ณ‘๋ชฉ ์ง€์ ์ด ๋˜์—ˆ์Šต๋‹ˆ๋‹ค.
  • โ€˜์˜ฌ์ธ(All-in)โ€™ ์ „๋žต์˜ ํšจ๊ณผ: ์Šค์บ”๋Ÿฐ์€ โ€œ๋งŒ์•ฝ ๋‹น์‹ ์ด ๋ชจ๋“  ๊ฒƒ์„ ๊ฑธ๊ณ (go all-in), ํŒ€๊ณผ ์ฝ”๋“œ๋ฒ ์ด์Šค๋ฅผ ์ค€๋น„ํ•œ๋‹ค๋ฉด, ์ „๋ฐ˜์ ์ธ ์ œํ’ˆ ํ’ˆ์งˆ๊ณผ ๊ฐœ๋ฐœ์ž ๊ฒฝํ—˜์ด ํ–ฅ์ƒ๋  ๊ฒƒโ€์ด๋ผ๊ณ  ๊ฐ•์กฐํ–ˆ์Šต๋‹ˆ๋‹ค. ๋†’์€ ์‹ ๋ขฐ ๋ฌธํ™”๋ฅผ ๋ฐ”ํƒ•์œผ๋กœ ํŒ€์ด ์‹œ์Šคํ…œ์„ ์•…์šฉํ•˜์ง€ ์•Š๊ณ  ๋ชฉํ‘œ๋ฅผ ํ–ฅํ•ด ๋‚˜์•„๊ฐ€๋„๋ก ๋…๋ คํ–ˆ์Šต๋‹ˆ๋‹ค.
  • 10๋ฐฐ์˜ ๋น„์ „: 2๋ฐฐ ์„ฑ๊ณผ์— ๋งŒ์กฑํ•˜์ง€ ์•Š๊ณ , โ€œ์™œ 10๋ฐฐ๋Š” ์•ˆ ๋˜๋Š”๊ฐ€?โ€๋ผ๋Š” ์งˆ๋ฌธ์„ ๋˜์ง€๋ฉฐ ์ง€์†์ ์ธ ๊ฐœ์„ ์„ ์ถ”๊ตฌํ•˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค.

3. ์†๋„์™€ ํ•จ๊ป˜ ํ–ฅ์ƒ๋œ ํ’ˆ์งˆ: ๊ธฐ์ˆ  ๋ถ€์ฑ„ ํ•ด๊ฒฐ์˜ ๊ธฐํšŒ

์ผ๊ฐ์—์„œ๋Š” AI๋กœ ์ฝ”๋“œ๋ฅผ ๋น ๋ฅด๊ฒŒ ์ƒ์„ฑํ•˜๋ฉด ํ’ˆ์งˆ์ด ์ €ํ•˜๋  ๊ฒƒ์ด๋ผ๋Š” ์šฐ๋ ค๋ฅผ ์ œ๊ธฐํ•ฉ๋‹ˆ๋‹ค. ํ•˜์ง€๋งŒ ์ธํ„ฐ์ฝค์˜ ๊ฒฝํ—˜์€ ์ •๋ฐ˜๋Œ€์˜€์Šต๋‹ˆ๋‹ค.

ํ’ˆ์งˆ ์ธก์ • ๋ฐ ๊ฒ€์ฆ:

  • ์ถœ์‹œ ์‹œ๊ฐ„ ๋‹จ์ถ•: ๊ธฐ๋Šฅ ๊ฐœ๋ฐœ ์ฒซ ์ฝ”๋“œ ์ž‘์„ฑ๋ถ€ํ„ฐ ๋‰ด์Šค ์ฑ„๋„์— ์—…๋ฐ์ดํŠธ๊ฐ€ ๊ฒŒ์‹œ๋˜๊ธฐ๊นŒ์ง€ ๊ฑธ๋ฆฌ๋Š” ์‹œ๊ฐ„์ด ๊พธ์ค€ํžˆ ๋‹จ์ถ•๋˜์—ˆ์Šต๋‹ˆ๋‹ค.
  • ์Šคํƒ ํฌ๋“œ ์—ฐ๊ตฌ ํ˜‘๋ ฅ: ์ธํ„ฐ์ฝค์€ ์Šคํƒ ํฌ๋“œ ์—ฐ๊ตฌ ๊ทธ๋ฃน๊ณผ ํ˜‘๋ ฅํ•˜์—ฌ ์ž์ฒด ๋ฐ์ดํ„ฐ๋ฅผ ์ œ๊ณตํ–ˆ์Šต๋‹ˆ๋‹ค. ์ด ์—ฐ๊ตฌ ๊ฒฐ๊ณผ, ์ฝ”๋“œ ํ’ˆ์งˆ์ด ์˜คํžˆ๋ ค ํ–ฅ์ƒ๋˜๊ณ  ์žˆ์Œ์ด ๋ฐํ˜€์กŒ์Šต๋‹ˆ๋‹ค.
  • ์‚ฌ๊ณ  ๊ฐ์†Œ: ์Šค์บ”๋Ÿฐ์€ ๋ชจ๋“  ์ธ์‹œ๋˜ํŠธ(incident)์— ์ฐธ์—ฌํ•˜์ง€๋งŒ, AI ๋„์ž… ์ดํ›„ ํ”„๋กœ๋•์…˜ ๊ด€๋ จ ์‚ฌ๊ณ ๋‚˜ ๊ณ ๊ฐ ๋ถˆ๋งŒ์ด ์ฆ๊ฐ€ํ•˜์ง€ ์•Š์•˜๋‹ค๊ณ  ๋ณด๊ณ ํ–ˆ์Šต๋‹ˆ๋‹ค.

๊ธฐ์ˆ  ๋ถ€์ฑ„ ํ•ด๊ฒฐ์˜ ๊ธฐํšŒ: AI๋Š” ๊ธฐ์ˆ  ๋ถ€์ฑ„(tech debt)๋ฅผ ํ•ด๊ฒฐํ•˜๊ณ  ์ฝ”๋“œ๋ฒ ์ด์Šค์˜ ์ „๋ฐ˜์ ์ธ ํ’ˆ์งˆ์„ ํ–ฅ์ƒ์‹œํ‚ฌ ์ˆ˜ ์žˆ๋Š” ์ ˆํ˜ธ์˜ ๊ธฐํšŒ๋ฅผ ์ œ๊ณตํ•ฉ๋‹ˆ๋‹ค. ๊ณผ๊ฑฐ์—๋Š” ์ฝ”๋“œ ํ’ˆ์งˆ ๊ฐœ์„ , ๋ณด์•ˆ ๊ฐ•ํ™”, ๊ฐœ๋ฐœ์ž ๊ฒฝํ—˜ ํ–ฅ์ƒ ๋“ฑ ๋‚ด๋ถ€ ํ”„๋กœ์ ํŠธ์— ํ• ๋‹นํ•  ์ˆ˜ ์žˆ๋Š” R&D ์—ญ๋Ÿ‰์ด ์ œํ•œ์ ์ด์—ˆ์Šต๋‹ˆ๋‹ค. ์ด๋Ÿฌํ•œ ํ”„๋กœ์ ํŠธ๋Š” ์ง์ ‘์ ์ธ ๋งค์ถœ(ARR, Annual Recurring Revenue)์„ ์ฐฝ์ถœํ•˜์ง€ ์•Š๊ธฐ ๋•Œ๋ฌธ์— ๋น„์ฆˆ๋‹ˆ์Šค์  ์šฐ์„ ์ˆœ์œ„์—์„œ ๋ฐ€๋ ค๋‚˜๊ธฐ ์‰ฌ์› ์Šต๋‹ˆ๋‹ค.

ํ•˜์ง€๋งŒ AI๋ฅผ ํ†ตํ•ด ์ด๋Ÿฌํ•œ ์ž‘์—…์˜ โ€˜๋น„์šฉโ€™์ด ํฌ๊ฒŒ ์••์ถ•๋˜๋ฉด์„œ, ์ด์ œ ๊ธฐ์—…์€ ๋‚ด๋ถ€ ํˆฌ์ž์— ๋” ์ ๊ทน์ ์œผ๋กœ ๋‚˜์„ค ์ˆ˜ ์žˆ๊ฒŒ ๋˜์—ˆ์Šต๋‹ˆ๋‹ค. ์Šค์บ”๋Ÿฐ์€ CTO๋‚˜ ์—”์ง€๋‹ˆ์–ด๋ง VP(Vice President)๋“ค์—๊ฒŒ โ€œ์ฝ”๋“œ๋ฒ ์ด์Šค์—์„œ ์‹ซ์–ดํ•˜๋Š” ๋ชจ๋“  ๊ฒƒ์„ ํ•œ ๋‹ฌ ๋™์•ˆ ์ˆ˜์ •ํ•˜๋Š” ๋ฐ ํˆฌ์žํ•˜๊ณ , ์–ผ๋งˆ๋‚˜ ๋น ๋ฅด๊ฒŒ ํ•ด๊ฒฐํ•  ์ˆ˜ ์žˆ๋Š”์ง€ ๋ณด๋ผโ€๊ณ  ์กฐ์–ธํ•˜๋ฉฐ, ์ด๊ฒƒ์ด ํŒ€์˜ ์‚ฌ๊ธฐ์™€ ์ƒ์‚ฐ์„ฑ ํ–ฅ์ƒ์— ํฐ ๋„์›€์ด ๋  ๊ฒƒ์ด๋ผ๊ณ  ๋ง๋ถ™์˜€์Šต๋‹ˆ๋‹ค.

4. AI ์—์ด์ „ํŠธ ์ค‘์‹ฌ์˜ ์†Œํ”„ํŠธ์›จ์–ด ๊ณต์žฅ

์ธํ„ฐ์ฝค์€ ๋ชจ๋“  ๊ธฐ์ˆ  ์ž‘์—…์ด โ€˜์—์ด์ „ํŠธ ์ค‘์‹ฌ(agent-first)โ€˜์œผ๋กœ ์ „ํ™˜๋  ๊ฒƒ์ด๋ผ๊ณ  ๋ฏฟ์Šต๋‹ˆ๋‹ค. ์ด๋Š” AI ์—์ด์ „ํŠธ๊ฐ€ ๊ธฐ๋ณธ์ ์ธ ์ž‘์—…์„ ์ˆ˜ํ–‰ํ•˜๊ณ , ์ธ๊ฐ„ ์—”์ง€๋‹ˆ์–ด๋Š” ๋” ๋†’์€ ์ˆ˜์ค€์˜ ๋ฌธ์ œ ํ•ด๊ฒฐ๊ณผ ์ฐฝ์˜์ ์ธ ์ž‘์—…์— ์ง‘์ค‘ํ•˜๋Š” ๋ฏธ๋ž˜๋ฅผ ์˜๋ฏธํ•ฉ๋‹ˆ๋‹ค.

โ€˜Create PRโ€™ ์Šคํ‚ฌ์˜ ์‚ฌ๋ก€: ์ธํ„ฐ์ฝค์€ ํด๋กœ๋“œ ์ฝ”๋“œ๋ฅผ ์‚ฌ์šฉํ•˜๋ฉด์„œ ํ•œ ๊ฐ€์ง€ ๋ฌธ์ œ์— ์ง๋ฉดํ–ˆ์Šต๋‹ˆ๋‹ค. AI๊ฐ€ ์ƒ์„ฑํ•˜๋Š” ํ’€ ๋ฆฌํ€˜์ŠคํŠธ ์„ค๋ช…(PR description)์ด ์ฝ”๋“œ ์ž์ฒด๋งŒ ์„ค๋ช…ํ•  ๋ฟ, PR์˜ โ€˜์˜๋„(intent)โ€˜๋‚˜ ๊ด€๋ จ ๋งฅ๋ฝ์„ ์ œ๋Œ€๋กœ ์ „๋‹ฌํ•˜์ง€ ๋ชปํ•ด ํ’ˆ์งˆ์ด ์ €ํ•˜๋˜๋Š” ๊ฒฝํ–ฅ์ด ์žˆ์—ˆ์Šต๋‹ˆ๋‹ค.

  • ๋ฌธ์ œ ํ•ด๊ฒฐ: ์ธํ„ฐ์ฝค์€ LLM ์‹ฌ์‚ฌ๊ด€(LLM judge)์„ ํ™œ์šฉํ•˜์—ฌ ์ข‹์€ PR ์„ค๋ช…์˜ ๊ธฐ์ค€์„ ์ •์˜ํ•˜๊ณ , ๊ธฐ์กด PR ์„ค๋ช…์˜ ํ’ˆ์งˆ ์ €ํ•˜ ์ถ”์„ธ๋ฅผ ํ™•์ธํ–ˆ์Šต๋‹ˆ๋‹ค.
  • โ€˜Create PRโ€™ ์Šคํ‚ฌ ๊ฐœ๋ฐœ: ์ด๋ฅผ ํ•ด๊ฒฐํ•˜๊ธฐ ์œ„ํ•ด โ€˜create PRโ€™์ด๋ผ๋Š” ๋‚ด๋ถ€ ์Šคํ‚ฌ(skill)์„ ๊ฐœ๋ฐœํ–ˆ์Šต๋‹ˆ๋‹ค. ์ด ์Šคํ‚ฌ์€ ์„ธ์…˜์˜ ๋งฅ๋ฝ์„ ํ™œ์šฉํ•˜์—ฌ PR์˜ ์˜๋„๋ฅผ ๋ช…ํ™•ํ•˜๊ฒŒ ์„ค๋ช…ํ•˜๋„๋ก ์—์ด์ „ํŠธ๋ฅผ ์•ˆ๋‚ดํ•ฉ๋‹ˆ๋‹ค.
  • ํ‘œ์ค€ ๊ฐ•์ œ: ์ดˆ๊ธฐ์—๋Š” ๊ฐœ๋ฐœ์ž๋“ค์ด ์ด ์Šคํ‚ฌ์„ ์ž๋ฐœ์ ์œผ๋กœ ์‚ฌ์šฉํ•˜์ง€ ์•Š์•˜๊ธฐ์—, ์ธํ„ฐ์ฝค์€ GitHub CLI๋ฅผ ํ†ตํ•œ PR ์ƒ์„ฑ์„ ์ฐจ๋‹จํ•˜๊ณ  ๋ฐ˜๋“œ์‹œ โ€˜create PRโ€™ ์Šคํ‚ฌ์„ ์‚ฌ์šฉํ•˜๋„๋ก ๊ฐ•์ œํ–ˆ์Šต๋‹ˆ๋‹ค.
  • โ€˜์†Œํ”„ํŠธ์›จ์–ด ๊ณต์žฅโ€™ ๋น„์œ : ์Šค์บ”๋Ÿฐ์€ ์ด๋ฅผ โ€˜์†Œํ”„ํŠธ์›จ์–ด ๊ณต์žฅ(software factory)โ€˜์— ๋น„์œ ํ•˜๋ฉฐ, CI/CD ํŒŒ์ดํ”„๋ผ์ธ์ฒ˜๋Ÿผ ์ฝ”๋“œ ์ž‘์„ฑ ๋‹จ๊ณ„๋ถ€ํ„ฐ ํ‘œ์ค€ํ™”๋œ ๊ฐ€๋“œ๋ ˆ์ผ(guardrails)์„ ์ ์šฉํ•  ์ˆ˜ ์žˆ๊ฒŒ ๋˜์—ˆ๋‹ค๊ณ  ์„ค๋ช…ํ–ˆ์Šต๋‹ˆ๋‹ค. ์ด๋Š” ๊ฐœ๋ณ„ ์ž‘์—…์˜ ์‚ฌ์†Œํ•ด ๋ณด์ด๋Š” ํ’ˆ์งˆ ๊ฐœ์„ ์ด ๋ชจ์—ฌ ์ „์ฒด ์‹œ์Šคํ…œ์˜ ์‹ ๋ขฐ์„ฑ๊ณผ ์˜ˆ์ธก ๊ฐ€๋Šฅ์„ฑ์„ ๋†’์ด๋Š” ๊ฒฐ๊ณผ๋ฅผ ๊ฐ€์ ธ์˜ต๋‹ˆ๋‹ค.

๊ฒฐ๊ณผ์ ์œผ๋กœ AI๋Š” ๋ฆฌ๋‹ค์ด๋ ‰ํŠธ(redirect) ์ถ”๊ฐ€์™€ ๊ฐ™์€ ๋ฐ˜๋ณต์ ์ด๊ณ  ์ง€๋ฃจํ•œ ์ž‘์—…์€ ์—์ด์ „ํŠธ์—๊ฒŒ ๋งก๊ธฐ๊ณ , ์ธ๊ฐ„ ์—”์ง€๋‹ˆ์–ด๋Š” ๊ณ ๊ฐ์„ ์œ„ํ•œ ํ˜์‹ ์ ์ธ ๋ฌธ์ œ ํ•ด๊ฒฐ์— ์ง‘์ค‘ํ•  ์ˆ˜ ์žˆ๋„๋ก ํ•ด์ค๋‹ˆ๋‹ค.

5. AI ํ™œ์šฉ์˜ ํˆฌ๋ช…์„ฑ๊ณผ ์ตœ์ ํ™”: ํ…”๋ ˆ๋ฉ”ํŠธ๋ฆฌ

AI ๋„๊ตฌ์˜ ์„ฑ๊ณต์ ์ธ ๋„์ž…๊ณผ ํ™•์‚ฐ์„ ์œ„ํ•ด์„œ๋Š” โ€˜๋ˆˆ์„ ๊ฐ€๋ฆฐ ์ฑ„ ๋น„ํ–‰ํ•˜์ง€ ์•Š๋Š” ๊ฒƒ(not flying blind)โ€˜์ด ์ค‘์š”ํ•ฉ๋‹ˆ๋‹ค. ์ธํ„ฐ์ฝค์€ AI ํ™œ์šฉ ํ˜„ํ™ฉ์„ ํŒŒ์•…ํ•˜๊ณ  ๊ฐœ์„ ํ•˜๊ธฐ ์œ„ํ•ด ๊ฐ•๋ ฅํ•œ ํ…”๋ ˆ๋ฉ”ํŠธ๋ฆฌ ์‹œ์Šคํ…œ์„ ๊ตฌ์ถ•ํ–ˆ์Šต๋‹ˆ๋‹ค.

  • ์Šคํ‚ฌ ์‚ฌ์šฉ๋Ÿ‰ ์ถ”์  (Honeycomb): ๊ฐ ๋‚ด๋ถ€ ์Šคํ‚ฌ์˜ ์‚ฌ์šฉ๋Ÿ‰ ๋ฐ์ดํ„ฐ๋ฅผ ํ—ˆ๋‹ˆ์ฝค(Honeycomb)์œผ๋กœ ์ „์†กํ•˜์—ฌ, ์–ด๋–ค ์Šคํ‚ฌ์ด ์–ผ๋งˆ๋‚˜ ์ž์ฃผ ์‚ฌ์šฉ๋˜๋Š”์ง€ ์‹ค์‹œ๊ฐ„์œผ๋กœ ๋ชจ๋‹ˆํ„ฐ๋งํ•ฉ๋‹ˆ๋‹ค. ์ด๋ฅผ ํ†ตํ•ด ๊ฐœ๋ฐœ์ž๋“ค์€ ์ž์‹ ์ด ๋งŒ๋“  ์Šคํ‚ฌ์˜ ์‹ค์ œ ์˜ํ–ฅ์„ ํŒŒ์•…ํ•˜๊ณ , ํŒ€ ์ „์ฒด์˜ AI ํ™œ์šฉ ํŒจํ„ด์„ ๋ถ„์„ํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
  • ์„ธ์…˜ ๋ฐ์ดํ„ฐ ๋ถ„์„ (S3 ๋ฐ ๋‚ด๋ถ€ ๋„๊ตฌ): ๋ชจ๋“  ํด๋กœ๋“œ ์ฝ”๋“œ ์„ธ์…˜ ๋ฐ์ดํ„ฐ(๋Œ€ํ™” ๊ธฐ๋ก)๋ฅผ S3์— ์ต๋ช…ํ™”ํ•˜์—ฌ ์ €์žฅํ•ฉ๋‹ˆ๋‹ค. ์ด ๋ฐ์ดํ„ฐ๋ฅผ ๊ธฐ๋ฐ˜์œผ๋กœ ์„ธ์…˜ ์ดํƒˆ๋ฅ , ์œ ์šฉํ•œ ๊ฒฐ๊ณผ(์˜ˆ: PR ์ƒ์„ฑ) ๋„๋‹ฌ ์‹œ๊ฐ„ ๋“ฑ์„ ๋ถ„์„ํ•˜์—ฌ AI ํ™œ์šฉ์˜ ํšจ๊ณผ์„ฑ์„ ํ‰๊ฐ€ํ•ฉ๋‹ˆ๋‹ค.
  • ๊ฐœ์ธ ๋งž์ถคํ˜• ํ”ผ๋“œ๋ฐฑ ๋„๊ตฌ: ์„ธ์…˜ ๋ฐ์ดํ„ฐ๋ฅผ ํ™œ์šฉํ•˜์—ฌ ๊ฐœ์ธ๋ณ„ AI ํ™œ์šฉ ํŒจํ„ด์— ๋Œ€ํ•œ ํ”ผ๋“œ๋ฐฑ์„ ์ œ๊ณตํ•˜๋Š” ๋‚ด๋ถ€ ๋„๊ตฌ๋ฅผ ๊ฐœ๋ฐœํ–ˆ์Šต๋‹ˆ๋‹ค. ์˜ˆ๋ฅผ ๋“ค์–ด, ํŠน์ • ์Šคํ‚ฌ ์‚ฌ์šฉ์ด ๋น„ํšจ์œจ์ ์ด์—ˆ๊ฑฐ๋‚˜ ๊ฐœ์„ ์ด ํ•„์š”ํ•œ ๋ถ€๋ถ„์„ ์‚ฌ์šฉ์ž์—๊ฒŒ ์•Œ๋ ค์ค๋‹ˆ๋‹ค. ์ด๋Š” ์ง์›๋“ค์ด AI ๋„๊ตฌ๋ฅผ ๋” ํšจ๊ณผ์ ์œผ๋กœ ํ™œ์šฉํ•˜๊ณ  ์Šค์Šค๋กœ ๊ฐœ์„ ํ•ด ๋‚˜๊ฐˆ ์ˆ˜ ์žˆ๋„๋ก ์ง€์›ํ•ฉ๋‹ˆ๋‹ค.

์ด๋Ÿฌํ•œ ํ…”๋ ˆ๋ฉ”ํŠธ๋ฆฌ ์‹œ์Šคํ…œ์€ ์–ด๋–ค ์Šคํ‚ฌ์ด ๊ฐ€์žฅ ํšจ๊ณผ์ ์ธ์ง€, ์–ด๋–ค ์Šคํ‚ฌ์ด ๊ฐœ์„ ์ด ํ•„์š”ํ•œ์ง€ ๋“ฑ ์‹คํ–‰ ๊ฐ€๋Šฅํ•œ ํ†ต์ฐฐ๋ ฅ์„ ์ œ๊ณตํ•˜์—ฌ AI ํ”Œ๋žซํผ์˜ ์ง€์†์ ์ธ ์„ฑ์žฅ์„ ๋•์Šต๋‹ˆ๋‹ค.

6. ๋‚ด๋ถ€ ์Šคํ‚ฌ ์ €์žฅ์†Œ ๊ตฌ์ถ• ๋ฐ ๋ฐฐํฌ

์ธํ„ฐ์ฝค์€ ๋‚ด๋ถ€ AI ์Šคํ‚ฌ๋“ค์„ ํšจ์œจ์ ์œผ๋กœ ๊ด€๋ฆฌํ•˜๊ณ  ๋ฐฐํฌํ•˜๊ธฐ ์œ„ํ•ด GitHub ์ €์žฅ์†Œ(repo)๋ฅผ ํ™œ์šฉํ•˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค.

  • ์ค‘์•™ ์ง‘์ค‘์‹ ์ €์žฅ์†Œ: ๋ชจ๋“  ํŒ€์ด ๊ฐ์ž์˜ ํ•„์š”์— ๋งž๋Š” ํ”Œ๋Ÿฌ๊ทธ์ธ๊ณผ ์Šคํ‚ฌ์„ ๊ฐœ๋ฐœํ•˜์—ฌ ์ด ์ค‘์•™ ์ €์žฅ์†Œ์— ๊ธฐ์—ฌํ•ฉ๋‹ˆ๋‹ค. ํ•ต์‹ฌ ํ”Œ๋Ÿฌ๊ทธ์ธ(์˜ˆ: ์•ˆ์ „ ํ›„ํฌ, PR ๋ณ‘ํ•ฉ, ํ…”๋ ˆ๋ฉ”ํŠธ๋ฆฌ ๊ด€๋ จ ๊ธฐ๋Šฅ)์€ ์—„๊ฒฉํ•œ ํ’ˆ์งˆ ๊ด€๋ฆฌ๋ฅผ ๊ฑฐ์ณ ๋ชจ๋“  ์ง์›์—๊ฒŒ ๋ฐฐํฌ๋ฉ๋‹ˆ๋‹ค.
  • IT ํŒ€๊ณผ์˜ ํ˜‘๋ ฅ: ํด๋กœ๋“œ ์ฝ”๋“œ์˜ ๊ธฐ๋ณธ ํ”Œ๋Ÿฌ๊ทธ์ธ ๋ฐฐํฌ ๋ฉ”์ปค๋‹ˆ์ฆ˜์ด ๋ถˆ์•ˆ์ •ํ•˜๋‹ค๊ณ  ํŒ๋‹จํ•˜์—ฌ, ์ธํ„ฐ์ฝค์€ ๋‚ด๋ถ€ IT ์‹œ์Šคํ…œ์„ ํ™œ์šฉํ•ด ๋ชจ๋“  ํ”Œ๋Ÿฌ๊ทธ์ธ์„ ์ง์›๋“ค์˜ ๋…ธํŠธ๋ถ ๋””์Šคํฌ์— ์ง์ ‘ ๋™๊ธฐํ™”ํ•ฉ๋‹ˆ๋‹ค. ์ด๋Š” ์•ˆ์ •์ ์ธ ๋ฐฐํฌ๋ฅผ ๋ณด์žฅํ•˜๊ณ  ๋””๋ฒ„๊น…์˜ ๋ณต์žก์„ฑ์„ ์ค„์—ฌ์ค๋‹ˆ๋‹ค.

7. ๋น„์šฉ๊ณผ ํˆฌ์ž: ์•ŒํŒŒ๋ฅผ ์œ„ํ•œ ๊ณผ๊ฐํ•œ ์„ ํƒ

AI ๋ชจ๋ธ ์‚ฌ์šฉ ๋น„์šฉ์€ ์ธํ„ฐ์ฝค์˜ ๊ฐœ๋ฐœ ์†๋„ ๊ทธ๋ž˜ํ”„์ฒ˜๋Ÿผ ๊ฐ€ํŒŒ๋ฅด๊ฒŒ ์ƒ์Šนํ•˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค. ์ด๋Š” ๋งˆ์น˜ ์ƒˆ๋กœ์šด ์‚ฌ๋ฌด์‹ค์„ ์—ฌ๋Ÿฌ ๊ฐœ ์—ฌ๋Š” ๊ฒƒ๊ณผ ๊ฐ™์€ ๋ง‰๋Œ€ํ•œ ๋น„์šฉ์ž…๋‹ˆ๋‹ค.

  • โ€˜๋น„์šฉ์€ ๋‚˜์ค‘์— ๊ณ ๋ คโ€™ ์ „๋žต: ์ธํ„ฐ์ฝค์€ ํ˜„์žฌ โ€œ๋ชจ๋“  ๊ฒƒ์— ์˜คํ‘ธ์Šค ๋ชจ๋ธ์„ ์‚ฌ์šฉํ•˜๊ณ , ๋น„์šฉ์€ ๋‚˜์ค‘์— ์‹ ๊ฒฝ ์“ฐ์ž(care about the bill later)โ€œ๋Š” ๊ณผ๊ฐํ•œ ์ „๋žต์„ ์ทจํ•˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค. ๊ทธ ์ด์œ ๋Š” ํ˜„์žฌ AI๋ฅผ ํ†ตํ•ด ์–ป๋Š” โ€˜์•ŒํŒŒ(alpha)โ€™ ๋˜๋Š” โ€˜์ด์ โ€™์ด ๋น„์šฉ์„ ํ›จ์”ฌ ์ƒํšŒํ•œ๋‹ค๊ณ  ํŒ๋‹จํ•˜๊ธฐ ๋•Œ๋ฌธ์ž…๋‹ˆ๋‹ค.
  • ์žฅ๊ธฐ์  ํˆฌ์ž ๊ด€์ : ์ธํ„ฐ์ฝค์€ AI ํ”Œ๋žซํผ์— ๋Œ€ํ•œ ํˆฌ์ž๋ฅผ ์žฅ๊ธฐ์ ์ธ ๊ด€์ ์—์„œ ์ ‘๊ทผํ•˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค. ๋‹น์žฅ์˜ ๋น„์šฉ ์ตœ์ ํ™”๋ณด๋‹ค๋Š” ์ตœ๋Œ€ํ•œ ๋น ๋ฅด๊ฒŒ AI์˜ ์ž ์žฌ๋ ฅ์„ ํ™œ์šฉํ•˜์—ฌ ํ˜์‹ ์ ์ธ ์„ฑ๊ณผ๋ฅผ ๋‚ด๋Š” ๋ฐ ์ง‘์ค‘ํ•˜๊ณ , ์ถฉ๋ถ„ํ•œ ์ด์ ์„ ํ™•๋ณดํ•œ ํ›„์— ๋น„์šฉ ํšจ์œจํ™” ๋‹จ๊ณ„๋ฅผ ๋ฐŸ์„ ๊ณ„ํš์ž…๋‹ˆ๋‹ค.

๋ฌผ๋ก  ์ด๋Ÿฌํ•œ ์ ‘๊ทผ ๋ฐฉ์‹์ด ๋ชจ๋“  ๊ธฐ์—…์— ์ ํ•ฉํ•œ ๊ฒƒ์€ ์•„๋‹™๋‹ˆ๋‹ค. ํ•˜์ง€๋งŒ ์ธํ„ฐ์ฝค์€ ์„ ๋‘ ์ฃผ์ž๋กœ์„œ AI๊ฐ€ ๊ฐ€์ ธ์˜ฌ ๋ฏธ๋ž˜ ๊ฐ€์น˜์— ๋Œ€ํ•œ ํ™•๊ณ ํ•œ ๋ฏฟ์Œ์„ ๋ฐ”ํƒ•์œผ๋กœ ๊ณผ๊ฐํ•œ ํˆฌ์ž๋ฅผ ๋‹จํ–‰ํ•˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค.


๊ฒฐ๋ก : AI๊ฐ€ ์ด๋„๋Š” ์—”์ง€๋‹ˆ์–ด๋ง์˜ ์ƒˆ๋กœ์šด ์‹œ๋Œ€

์ธํ„ฐ์ฝค์˜ ์‚ฌ๋ก€๋Š” AI๊ฐ€ ๋‹จ์ˆœํ•œ ์ƒ์‚ฐ์„ฑ ํ–ฅ์ƒ์„ ๋„˜์–ด, ์—”์ง€๋‹ˆ์–ด๋ง ์กฐ์ง์˜ ๋ณธ์งˆ๊ณผ ์ž‘๋™ ๋ฐฉ์‹์„ ๊ทผ๋ณธ์ ์œผ๋กœ ๋ณ€ํ™”์‹œํ‚ฌ ์ˆ˜ ์žˆ์Œ์„ ๋ช…ํ™•ํžˆ ๋ณด์—ฌ์ค๋‹ˆ๋‹ค. 2๋ฐฐ์˜ ๊ฐœ๋ฐœ ์†๋„ ํ–ฅ์ƒ, ์ฝ”๋“œ ํ’ˆ์งˆ ๊ฐœ์„ , ๊ธฐ์ˆ  ๋ถ€์ฑ„ ํ•ด๊ฒฐ, ๊ทธ๋ฆฌ๊ณ  โ€˜์—์ด์ „ํŠธ ์ค‘์‹ฌโ€™์˜ ์†Œํ”„ํŠธ์›จ์–ด ๊ณต์žฅ์œผ๋กœ์˜ ์ „ํ™˜์€ AI ์‹œ๋Œ€ ์—”์ง€๋‹ˆ์–ด๋ง์˜ ๋ฏธ๋ž˜๋ฅผ ์—ฟ๋ณผ ์ˆ˜ ์žˆ๋Š” ์ค‘์š”ํ•œ ์ฒญ์‚ฌ์ง„์ž…๋‹ˆ๋‹ค.

๋ธŒ๋ผ์ด์–ธ ์Šค์บ”๋Ÿฐ์€ โ€œAI์™€ ์†๋„ ์ฆ๊ฐ€๋Š” ๋ชจ๋“  ๊ฐ•์ ๊ณผ ์•ฝ์ ์„ ์ฆํญ์‹œํ‚จ๋‹คโ€๊ณ  ๋งํ•ฉ๋‹ˆ๋‹ค. ์ธํ„ฐ์ฝค์€ ๊ฒฌ๊ณ ํ•œ ์†Œํ”„ํŠธ์›จ์–ด ๊ฐœ๋ฐœ ์—ญ๋Ÿ‰๊ณผ ๋ณ€ํ™”์— ๋Œ€ํ•œ ๋†’์€ ๊ฐœ๋ฐฉ์„ฑ์„ ๋ฐ”ํƒ•์œผ๋กœ AI์˜ ๊ฐ•์ ์„ ๊ทน๋Œ€ํ™”ํ•  ์ˆ˜ ์žˆ์—ˆ์Šต๋‹ˆ๋‹ค. ์ด์ œ ๊ธฐ์—…๋“ค์€ AI๋ฅผ ํ†ตํ•ด ์ธ๊ฐ„ ์—”์ง€๋‹ˆ์–ด๊ฐ€ ๋”์šฑ ์ฐฝ์˜์ ์ด๊ณ  ์ „๋žต์ ์ธ ๋ฌธ์ œ ํ•ด๊ฒฐ์— ์ง‘์ค‘ํ•  ์ˆ˜ ์žˆ๋„๋ก ์ง€์›ํ•˜๋ฉฐ, ๊ถ๊ทน์ ์œผ๋กœ ๋” ๋†’์€ ํ’ˆ์งˆ์˜ ์ œํ’ˆ์„ ๋” ๋น ๋ฅด๊ฒŒ ์‹œ์žฅ์— ์„ ๋ณด์ผ ์ˆ˜ ์žˆ๋Š” ์ƒˆ๋กœ์šด ์‹œ๋Œ€์— ์ง๋ฉดํ•ด ์žˆ์Šต๋‹ˆ๋‹ค. ์ธํ„ฐ์ฝค์˜ ์ด์•ผ๊ธฐ๋Š” ๋ณ€ํ™”๋ฅผ ๋‘๋ ค์›Œํ•˜์ง€ ์•Š๊ณ  ๊ณผ๊ฐํ•˜๊ฒŒ ํˆฌ์žํ•˜๋ฉฐ AI๋ฅผ ์กฐ์ง์˜ ํ•ต์‹ฌ ๋™๋ ฅ์œผ๋กœ ์‚ผ๋Š” ๊ธฐ์—…๋“ค์ด ์–ป๊ฒŒ ๋  ๋†€๋ผ์šด ์ด์ ์„ ์‹œ์‚ฌํ•ฉ๋‹ˆ๋‹ค.


โ€œInside the Five Days That Remade the Supreme Courtโ€ โ€” New York Times Podcasts ๊ธฐ๋ฐ˜ ๊ธฐ์‚ฌ ์›๋ณธ ์˜์ƒ ๋ณด๊ธฐ

๋Œ€๋ฒ•์› โ€˜๊ทธ๋ฆผ์ž ์‹ฌ์˜โ€™์˜ ์–ด๋‘์šด ํƒ„์ƒ: ๋น„๋ฐ€ ๋ฌธ์„œ๊ฐ€ ๋“œ๋Ÿฌ๋‚ธ 5์ผ๊ฐ„์˜ ๊ฒฉ๋Œ

๋ฏธ๊ตญ ๋Œ€๋ฒ•์›์€ ์ง€๋‚œ 10๋…„๊ฐ„ ์ด๋ฏผ๋ถ€ํ„ฐ ๋Œ€ํ†ต๋ น ๊ถŒํ•œ์— ์ด๋ฅด๊ธฐ๊นŒ์ง€ ์ฃผ์š” ํŒ๊ฒฐ๋“ค์„ ์„œ๋‘๋ฅด๊ณ  ์€๋ฐ€ํ•˜๊ฒŒ ์ฒ˜๋ฆฌํ•˜๋Š” ์‹œ์Šคํ…œ์— ์˜์กดํ•ด์™”์Šต๋‹ˆ๋‹ค. ์ด ์‹œ์Šคํ…œ์€ โ€˜๊ทธ๋ฆผ์ž ์‹ฌ์˜(shadow docket)โ€˜๋ผ๊ณ  ๋ถˆ๋ฆฌ๋ฉฐ, ๊ณต๊ฐœ์ ์ธ ๋…ผ์˜์™€ ์‹ฌ๋„ ์žˆ๋Š” ์‹ฌ์˜ ๊ณผ์ •์„ ๊ฑด๋„ˆ๋›ฐ๊ณ  ์‹ ์†ํ•˜๊ฒŒ ๊ฒฐ์ •์„ ๋‚ด๋ฆฌ๋Š” ๋ฐฉ์‹์œผ๋กœ ์ž‘๋™ํ•ฉ๋‹ˆ๋‹ค. ๋‰ด์š•ํƒ€์ž„์Šค(New York Times)์˜ ์‹ฌ์ธต ๋ณด๋„์— ๋”ฐ๋ฅด๋ฉด, ์ด ๊ทธ๋ฆผ์ž ์‹ฌ์˜๊ฐ€ ์‹œ์ž‘๋œ ์ •ํ™•ํ•œ ์ˆœ๊ฐ„๊ณผ ๊ทธ ๋ฐฐ๊ฒฝ์ด ๊ธฐ๋ฐ€ ํ•ด์ œ๋œ ๋Œ€๋ฒ•๊ด€๋“ค์˜ ์„œ์‹ ์„ ํ†ตํ•ด ์ฒ˜์Œ์œผ๋กœ ์„ธ์ƒ์— ๋“œ๋Ÿฌ๋‚ฌ์Šต๋‹ˆ๋‹ค.

์กฐ๋”” ์บ๋„ˆ(Jodie Caner)์™€ ์• ๋ค ๋ฆฝํƒ(Adam Liptac) ๊ธฐ์ž๋Š” ์ด ์ถฉ๊ฒฉ์ ์ธ ๋ฌธ์„œ๋ฅผ ํ†ตํ•ด ๋Œ€๋ฒ•์›์˜ ์šด์˜ ๋ฐฉ์‹์„ ์†ก๋‘๋ฆฌ์งธ ๋ฐ”๊พผ 5์ผ๊ฐ„์˜ ๋‚ด๋ถ€ ๋…ผ์Ÿ์„ ํŒŒํ—ค์ณค์Šต๋‹ˆ๋‹ค. ์ด๋Š” ๋‹จ์ˆœํ•œ ์ ˆ์ฐจ์  ๋ณ€ํ™”๊ฐ€ ์•„๋‹ˆ๋ผ, ์‚ฌ๋ฒ•๋ถ€์˜ ์ •๋‹น์„ฑ๊ณผ ๊ณต์ •์„ฑ์— ๋Œ€ํ•œ ๊ทผ๋ณธ์ ์ธ ์งˆ๋ฌธ์„ ๋˜์ง€๋Š” ์‚ฌ๊ฑด์œผ๋กœ ๊ธฐ๋ก๋  ๊ฒƒ์ž…๋‹ˆ๋‹ค.

โ€˜๊ทธ๋ฆผ์ž ์‹ฌ์˜โ€™๋ž€ ๋ฌด์—‡์ธ๊ฐ€? ๋Œ€๋ฒ•์›์˜ ๋‘ ์–ผ๊ตด

๋Œ€๋ฒ•์›์˜ ํŒ๊ฒฐ์€ ๊ตญ๊ฐ€ ์ „์ฒด๋ฅผ ๊ตฌ์†ํ•˜๋Š” ๋ฒ•์ด๋ฉฐ ํ•˜๊ธ‰ ๋ฒ•์›์— ์ง€์นจ์„ ์ œ๊ณตํ•˜๊ณ  ์†Œ์†ก ๋‹น์‚ฌ์ž๋“ค์—๊ฒŒ ํ–‰๋™ ๋ฐฉ์นจ์„ ์ œ์‹œํ•ฉ๋‹ˆ๋‹ค. ์ผ๋ฐ˜์ ์œผ๋กœ ๋Œ€๋ฒ•์›์ด ์‚ฌ๊ฑด์„ ์ฒ˜๋ฆฌํ•˜๋Š” ๋ฐฉ์‹์€ โ€˜์ •์‹ ์‹ฌ์˜(merits docket)โ€˜๋ผ๊ณ  ๋ถˆ๋ฆฝ๋‹ˆ๋‹ค. ์ด๋Š” ์šฐ๋ฆฌ๊ฐ€ ํ”ํžˆ ์•„๋Š” ๋Œ€๋ฒ•์›์˜ ๋ชจ์Šต์ž…๋‹ˆ๋‹ค. ๋Œ€๋ฒ•๊ด€๋“ค์€ ์–ด๋–ค ์‚ฌ๊ฑด์„ ์‹ฌ๋ฆฌํ• ์ง€ ์‹ ์ค‘ํ•˜๊ฒŒ ๊ฒฐ์ •ํ•˜๊ณ , ์ˆ˜๋งŽ์€ ๋ฒ•๋ฅ  ์„œ๋ฅ˜(briefs)๋ฅผ ๊ฒ€ํ† ํ•ฉ๋‹ˆ๋‹ค. ์‚ฌ๊ฑด ์‹ฌ๋ฆฌ๋ฅผ ๊ฒฐ์ •ํ•˜๋ฉด ๋˜๋‹ค์‹œ ์—ฌ๋Ÿฌ ์ฐจ๋ก€ ์„œ๋ฅ˜๋ฅผ ์ œ์ถœ๋ฐ›๊ณ , ๊ตฌ๋‘ ๋ณ€๋ก (oral arguments)์„ ๋“ฃ์Šต๋‹ˆ๋‹ค. ์ดํ›„ ๋Œ€๋ฒ•๊ด€๋“ค์€ ํ•จ๊ป˜ ์•‰์•„ ํ† ๋ก ํ•˜๊ณ  ํˆฌํ‘œํ•˜๋ฉฐ, 5๋ฒˆ, 10๋ฒˆ, ์‹ฌ์ง€์–ด 15๋ฒˆ๊นŒ์ง€ ํŒ๊ฒฐ๋ฌธ ์ดˆ๊ณ ์™€ ๋ณด์ถฉ ์˜๊ฒฌ, ๋ฐ˜๋Œ€ ์˜๊ฒฌ์„ ์ฃผ๊ณ ๋ฐ›์Šต๋‹ˆ๋‹ค. ์ด ๋ชจ๋“  ๊ณผ์ •์ด 1๋…„์—ฌ์— ๊ฑธ์ณ ์ง„ํ–‰๋œ ํ›„, ์ˆ˜์‹ญ์—์„œ ๋ฐฑ ํŽ˜์ด์ง€์— ๋‹ฌํ•˜๋Š” ๋…ผ๋ฆฌ์ ์ธ ํŒ๊ฒฐ๋ฌธ์ด ๋ฐœํ‘œ๋ฉ๋‹ˆ๋‹ค. ์ด๋Š” ์ง€๋Œ€ํ•œ ์ฃผ์˜์™€ ์ˆ™๊ณ ์˜ ์‚ฐ๋ฌผ์ž…๋‹ˆ๋‹ค.

๋ฐ˜๋ฉด, โ€˜๊ทธ๋ฆผ์ž ์‹ฌ์˜โ€™๋Š” ์ด ๋ชจ๋“  ๊ณผ์ •์„ ์ƒ๋žตํ•ฉ๋‹ˆ๋‹ค. ๋งค์šฐ ์งง์€ ๊ธฐ๊ฐ„ ์•ˆ์— ๊ฐ„๋žตํ•œ ์„œ๋ฅ˜๋งŒ์œผ๋กœ ์ง„ํ–‰๋˜๋ฉฐ, ๊ตฌ๋‘ ๋ณ€๋ก ์ด๋‚˜ ๋Œ€๋ฉด ์‹ฌ์˜๋Š” ์—†์Šต๋‹ˆ๋‹ค. ๊ทธ ๊ฒฐ๊ณผ๋Š” ๊ทผ๊ฑฐ๊ฐ€ ๊ฑฐ์˜ ์—†๊ฑฐ๋‚˜ ์•„์˜ˆ ์—†๋Š” ํŒ๊ฒฐ๋กœ ์ด์–ด์ง‘๋‹ˆ๋‹ค. ์ง€๋‚œ 10๋…„๊ฐ„ ๊ทธ๋ฆผ์ž ์‹ฌ์˜๋Š” ๋Œ€๋ฒ•์› ์—…๋ฌด์˜ ์ฃผ์š” ๋ถ€๋ถ„์ด ๋˜์—ˆ์œผ๋ฉฐ, ํŠนํžˆ ํŠธ๋Ÿผํ”„ ๋Œ€ํ†ต๋ น์—๊ฒŒ ๋ง‰๋Œ€ํ•œ ์žฌ๋Ÿ‰๊ถŒ๊ณผ ๊ถŒ๋ ฅ์„ ๋ถ€์—ฌํ•œ ๊ฒฐ์ •๋“ค์ด ํฌํ•จ๋˜์–ด ์žˆ์Šต๋‹ˆ๋‹ค.

๊ทธ๋ฆผ์ž ์‹ฌ์˜์— ๋Œ€ํ•œ ๋น„ํŒ๊ณผ ์˜นํ˜ธ๋ก :

๋น„ํŒ๋ก ์ž๋“ค์€ ๊ทธ๋ฆผ์ž ์‹ฌ์˜๊ฐ€ ๋„ˆ๋ฌด ๋น ๋ฅด๊ฒŒ ์ง„ํ–‰๋˜์–ด ์ถฉ๋ถ„ํ•œ ์ •๋ณด ์—†์ด ๊ฒฐ์ •์ด ๋‚ด๋ ค์ง€๋ฉฐ, ํŒ๊ฒฐ์˜ ์งˆ์ด ๋–จ์–ด์งˆ ์ˆ˜ ์žˆ๋‹ค๊ณ  ์šฐ๋ คํ•ฉ๋‹ˆ๋‹ค. ๋˜ํ•œ, ๋Œ€์ค‘๊ณผ์˜ ์†Œํ†ต ๋ถ€์กฑ์œผ๋กœ ์ธํ•ด ๋Œ€๋ฒ•์›์ด โ€˜์ง๊ฐ, ๊ฐœ์ธ์ ์ธ ๊ฐ์ •, ๋‹นํŒŒ์  ๋ณธ๋Šฅโ€™์— ๋”ฐ๋ผ ์„ฑ๊ธ‰ํ•˜๊ฒŒ ํŒ๊ฒฐํ•˜๋Š” ๊ฒƒ์ด ์•„๋‹Œ์ง€ ์˜๋ฌธ์„ ์ œ๊ธฐํ•ฉ๋‹ˆ๋‹ค. ๋А๋ฆฌ๊ณ  ์‹ ์ค‘ํ•˜๋ฉฐ ์‚ฌ๋ ค ๊นŠ์€ ์‚ฌ๋ฒ• ์‹œ์Šคํ…œ์ด ํ”ผํ•˜๊ณ ์ž ํ–ˆ๋˜ ๋ชจ๋“  ์š”์†Œ๊ฐ€ ๊ทธ๋ฆผ์ž ์‹ฌ์˜์—์„œ๋Š” ๊ทธ๋Œ€๋กœ ๋“œ๋Ÿฌ๋‚  ์ˆ˜ ์žˆ๋‹ค๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค.

๋ฐ˜๋ฉด, ์˜นํ˜ธ๋ก ์ž๋“ค์€ ์ด๋Ÿฌํ•œ ์ข…๋ฅ˜์˜ ๋ช…๋ น์ด ์˜ค๋žœ ์—ญ์‚ฌ๋ฅผ ๊ฐ€์ง€๊ณ  ์žˆ๋‹ค๊ณ  ์ฃผ์žฅํ•ฉ๋‹ˆ๋‹ค. ์‚ฌํ˜• ์ง‘ํ–‰์ด ์ž„๋ฐ•ํ–ˆ๊ฑฐ๋‚˜ ์„ ๊ฑฐ ๋ถ„์Ÿ์ฒ˜๋Ÿผ ์‹ ์†ํ•œ ๊ฒฐ์ •์ด ํ•„์š”ํ•œ ๊ฒฝ์šฐ, ๋Œ€๋ฒ•์›์€ ์ข…์ข… ๊ธด๊ธ‰ํ•˜๊ฒŒ ๊ฐœ์ž…ํ•ด์™”๋‹ค๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค. ๋˜ํ•œ, ๊ทธ๋ฆผ์ž ์‹ฌ์˜์˜ ๊ฒฐ์ •์€ โ€˜์ผ์‹œ์ ์ธ ๊ฒƒโ€™์ด๋ฉฐ, ์‚ฌ๊ฑด์ด ํ•˜๊ธ‰ ๋ฒ•์›์„ ๊ฑฐ์ณ ๋Œ€๋ฒ•์›์œผ๋กœ ๋‹ค์‹œ ๋Œ์•„์˜ฌ ๋•Œ๊นŒ์ง€ ํ˜„์ƒ ์œ ์ง€๋ฅผ ์œ„ํ•œ ์กฐ์น˜์ผ ๋ฟ์ด๋ผ๊ณ  ๋งํ•ฉ๋‹ˆ๋‹ค.

ํ•˜์ง€๋งŒ ์‹ค์ œ๋กœ๋Š” ์ด๋Ÿฌํ•œ ๊ฒฐ์ •๋“ค์ด ์ข…์ข… โ€˜์‚ฌ์‹ค์ƒ ์ตœ์ข…์ ์ธ(conclusive)โ€™ ๊ฒฐ๊ณผ๋ฅผ ๋‚ณ์Šต๋‹ˆ๋‹ค. ์ˆ˜์‹ญ๋งŒ ๋ช…์„ ์ถ”๋ฐฉํ•˜๊ฑฐ๋‚˜ ์ˆ˜์ฒœ ๋ช…์„ ํ•ด๊ณ ํ•˜๋Š” ๊ฒฐ์ •์€ ๋‚˜์ค‘์— ๋Œ€๋ฒ•์›์ด ๋ฒˆ๋ณตํ•˜๋”๋ผ๋„ ๋˜๋Œ๋ฆฌ๊ธฐ ๋งค์šฐ ์–ด๋ ต์Šต๋‹ˆ๋‹ค. ์ผ์‹œ์ ์ด๋ผ๊ณ  ํ•˜์ง€๋งŒ ๊ทธ ์˜ํ–ฅ๋ ฅ์€ ๋ง‰๋Œ€ํ•˜๋ฉฐ, ๊ฒฐ๊ตญ ์‚ฌ๊ฑด์˜ ์‹ค์งˆ์ ์ธ ํ•ด๊ฒฐ๋กœ ์ด์–ด์ง€๋Š” ๊ฒฝ์šฐ๊ฐ€ ๋งŽ์Šต๋‹ˆ๋‹ค.

2016๋…„, ๋Œ€๋ฒ•์›์„ ๋’คํ”๋“  5์ผ: โ€˜์ฒญ์ • ์ „๋ ฅ ๊ณ„ํšโ€™ ์‚ฌ๊ฑด

๋‰ด์š•ํƒ€์ž„์Šค๊ฐ€ ์ž…์ˆ˜ํ•œ ๋Œ€๋ฒ•๊ด€๋“ค ๊ฐ„์˜ 16ํŽ˜์ด์ง€์— ๋‹ฌํ•˜๋Š” ๊ธฐ๋ฐ€ ์„œ์‹ ์€ ๊ทธ๋ฆผ์ž ์‹ฌ์˜์˜ ๊ธฐ์›์„ ์ •ํ™•ํžˆ ๋ณด์—ฌ์ค๋‹ˆ๋‹ค. ์ด ์„œ์‹ ์„ ํ†ตํ•ด ์šฐ๋ฆฌ๋Š” ๋Œ€๋ฒ•๊ด€๋“ค์ด ์‚ฌ๋ฒ• ์ ˆ์ฐจ์˜ ์˜ค๋žœ ๊ทœ๋ฒ”์„ ํฌ๊ธฐํ•˜๊ณ  ์ƒˆ๋กœ์šด ์—…๋ฌด ๋ฐฉ์‹์„ ์ฑ„ํƒํ•˜๋˜ ์ˆœ๊ฐ„์„ ์—ฟ๋ณผ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.

์‚ฌ๊ฑด์˜ ๋ฐœ๋‹จ์€ 2016๋…„ ์˜ค๋ฐ”๋งˆ ํ–‰์ •๋ถ€์˜ โ€˜์ฒญ์ • ์ „๋ ฅ ๊ณ„ํš(Clean Power Plan)โ€˜์ด์—ˆ์Šต๋‹ˆ๋‹ค. ์˜ค๋ฐ”๋งˆ ๋Œ€ํ†ต๋ น์€ ๋‘ ๋ฒˆ์งธ ์ž„๊ธฐ ๋ง์— ๊ธฐํ›„ ๋ณ€ํ™”์— ๋Œ€์‘ํ•˜๊ธฐ ์œ„ํ•ด ์˜ํšŒ๋ฅผ ํ†ตํ•œ ์ž…๋ฒ•์— ์‹คํŒจํ•˜์ž ํ™˜๊ฒฝ๋ณดํ˜ธ์ฒญ(EPA)์— ๊ทœ์ œ๋ฅผ ์ง€์‹œํ–ˆ์Šต๋‹ˆ๋‹ค. ์ด ๊ทœ์ œ๋Š” ๋ฏธ๊ตญ์˜ ์ „๋ ฅ ์‹œ์Šคํ…œ์„ ์„ํƒ„์—์„œ ๋ฒ—์–ด๋‚˜ ์ฒญ์ •์—๋„ˆ์ง€์›์œผ๋กœ ์ „ํ™˜ํ•˜๋Š” ๊ฒƒ์„ ๋ชฉํ‘œ๋กœ ํ–ˆ์Šต๋‹ˆ๋‹ค. ๋‹น์—ฐํžˆ ์‚ฐ์—…๊ณ„์™€ ๊ณตํ™”๋‹น ์„ฑํ–ฅ์˜ ์ฃผ๋“ค์€ ์ด์— ๋ฐ˜๋ฐœํ–ˆ๊ณ , ๋ฒ•์›์— ๊ณ„ํš์˜ ์ค‘๋‹จ์„ ์š”์ฒญํ–ˆ์Šต๋‹ˆ๋‹ค.

์ด๋“ค์€ DC ํ•ญ์†Œ๋ฒ•์›(DC Circuit)์— ๊ณ„ํš์˜ ์œ„๋ฒ•์„ฑ์„ ํŒ๋‹จํ•ด๋‹ฌ๋ผ๊ณ  ์š”์ฒญํ•˜๋Š” ๋™์‹œ์—, ์†Œ์†ก์ด ์ง„ํ–‰๋˜๋Š” ๋™์•ˆ ๊ณ„ํš์˜ ์‹œํ–‰์„ โ€˜์ค‘๋‹จ(halt)โ€˜ํ•ด๋‹ฌ๋ผ๊ณ  ์š”๊ตฌํ–ˆ์Šต๋‹ˆ๋‹ค. ํ•ญ์†Œ๋ฒ•์›์€ ์œ„๋ฒ•์„ฑ ์‹ฌ๋ฆฌ๋Š” ์‹ ์†ํ•˜๊ฒŒ ์ง„ํ–‰ํ•˜๊ฒ ์ง€๋งŒ, ๊ณ„ํš์˜ ์‹œํ–‰์„ ์ค‘๋‹จํ•˜์ง€๋Š” ์•Š๊ฒ ๋‹ค๊ณ  ๊ฒฐ์ •ํ–ˆ์Šต๋‹ˆ๋‹ค.

๊ทธ๋Ÿฌ์ž ์ฒญ์ • ์ „๋ ฅ ๊ณ„ํš์— ๋ฐ˜๋Œ€ํ•˜๋Š” ์ธก์€ ์ „๋ก€ ์—†๋Š” ํ–‰๋™์„ ์ทจํ–ˆ์Šต๋‹ˆ๋‹ค. ํ•˜๊ธ‰ ๋ฒ•์›์ด ์•„์ง ๊ณ„ํš์˜ ํ•ฉ๋ฒ•์„ฑ์„ ๊ฒ€ํ† ํ•˜๊ธฐ๋„ ์ „์— ๋Œ€๋ฒ•์›์— ์ง์ ‘ ๊ณ„ํš์˜ ์‹œํ–‰์„ ์ค‘๋‹จํ•ด๋‹ฌ๋ผ๊ณ  ์š”์ฒญํ•œ ๊ฒƒ์ž…๋‹ˆ๋‹ค. ์ด ์š”์ฒญ์€ ๋‹น์‹œ ๋Œ€๋ฒ•์› ๊ด€๊ณ„์ž๋“ค ์‚ฌ์ด์—์„œ ์ „๋ก€ ์—†๋Š” ์ผ๋กœ ๋ฐ›์•„๋“ค์—ฌ์กŒ์Šต๋‹ˆ๋‹ค.

๋‹น์‹œ ๋Œ€๋ฒ•์›์€ 5๋Œ€ 4๋กœ ๋‚˜๋‰˜์–ด ๊ณตํ™”๋‹น์ด ์ž„๋ช…ํ•œ ๋Œ€๋ฒ•๊ด€๋“ค์ด ๋‹ค์ˆ˜๋ฅผ ์ฐจ์ง€ํ–ˆ์ง€๋งŒ, ๋ถ„์œ„๊ธฐ๋Š” ์ง€๊ธˆ๊ณผ๋Š” ์‚ฌ๋ญ‡ ๋‹ฌ๋ž์Šต๋‹ˆ๋‹ค. ์•ค์„œ๋‹ˆ ์ผ€๋„ค๋”” ๋Œ€๋ฒ•๊ด€(Justice Anthony Kennedy)์€ ๊ณตํ™”๋‹น์ด ์ž„๋ช…ํ–ˆ์ง€๋งŒ, โ€˜์„ค๋“ ๊ฐ€๋Šฅํ•œ ์ธ๋ฌผโ€™๋กœ ๋ถˆ๋ฆฌ๋Š” ์Šค์œ™ ๋ณดํ„ฐ(swing vote)์˜€์Šต๋‹ˆ๋‹ค. ๊ทธ๋Š” ๋™์„ฑ ๊ฒฐํ˜ผ ํ•ฉ๋ฒ•ํ™” ํŒ๊ฒฐ์—์„œ ๋‹ค์ˆ˜ ์˜๊ฒฌ์— ์ฐธ์—ฌํ•˜๋Š” ๋“ฑ ์˜ˆ์ธก ๋ถˆ๊ฐ€๋Šฅํ•œ ์ธ๋ฌผ์ด์—ˆ์Šต๋‹ˆ๋‹ค. ์ด๋Ÿฐ ๋Œ€๋ฒ•์›์— ์ „๋ก€ ์—†๋Š” ๊ธด๊ธ‰ ์š”์ฒญ์ด ๋„์ฐฉํ•œ ๊ฒƒ์ž…๋‹ˆ๋‹ค.

๋น„๋ฐ€ ๋ฉ”๋ชจ๊ฐ€ ๋“œ๋Ÿฌ๋‚ธ 5์ผ๊ฐ„์˜ ๊ฒฉ๋ก 

์ด ๊ธด๊ธ‰ ์š”์ฒญ์€ DC ํ•ญ์†Œ๋ฒ•์›์„ ๊ด€ํ• ํ•˜๋Š” ์กด ๋กœ๋ฒ„์ธ  ๋Œ€๋ฒ•์›์žฅ(Chief Justice John Roberts)์˜ ์‚ฌ๋ฌด์‹ค์— ๋„์ฐฉํ–ˆ์Šต๋‹ˆ๋‹ค. ๋ณ€ํ˜ธ์‚ฌ๋“ค์€ ๋กœ๋ฒ„์ธ  ๋Œ€๋ฒ•์›์žฅ์ด ์ด ์š”์ฒญ์„ ๊ธฐ๊ฐํ•˜๊ฑฐ๋‚˜ ๋™๋ฃŒ ๋Œ€๋ฒ•๊ด€๋“ค๊ณผ ์ƒ์˜ ํ›„ ๊ธฐ๊ฐํ•  ๊ฒƒ์ด๋ผ๊ณ  ์˜ˆ์ƒํ–ˆ์Šต๋‹ˆ๋‹ค. ๊ทธ๋Ÿฌ๋‚˜ ๋กœ๋ฒ„์ธ  ๋Œ€๋ฒ•์›์žฅ์€ ๋Œ€์‹  ๋‹ค๋ฅธ ๋Œ€๋ฒ•๊ด€๋“ค์—๊ฒŒ ์ด ์š”์ฒญ์„ โ€˜๋งค์šฐ ์ง„์ง€ํ•˜๊ฒŒ ๋ฐ›์•„๋“ค์—ฌ์•ผ ํ•œ๋‹คโ€™๋Š” ๋ฉ”๋ชจ๋ฅผ ์ž‘์„ฑํ•˜๊ธฐ ์‹œ์ž‘ํ–ˆ์Šต๋‹ˆ๋‹ค. ์ด๊ฒƒ์ด ๋Œ€๋ฒ•์›์˜ ์ˆ˜ ์„ธ๊ธฐ ๊ด€ํ–‰์„ ์žฌ์ •๋น„ํ•˜๋Š” 5์ผ๊ฐ„์˜ โ€˜๋‹จ๊ฑฐ๋ฆฌ ๊ฒฝ์ฃผโ€™์˜ ์‹œ์ž‘์ด์—ˆ์Šต๋‹ˆ๋‹ค.

  1. ๋กœ๋ฒ„์ธ  ๋Œ€๋ฒ•์›์žฅ์˜ ์ฒซ ํฌ๋ฌธ (2016๋…„ 2์›” 4์ผ): ๋กœ๋ฒ„์ธ  ๋Œ€๋ฒ•์›์žฅ์€ ์„ธ ํŽ˜์ด์ง€์งœ๋ฆฌ ๋ฉ”๋ชจ์—์„œ ์˜ค๋ฐ”๋งˆ ํ–‰์ •๋ถ€์˜ ๊ณ„ํš์ด ์ค‘๋‹จ๋˜์–ด์•ผ ํ•œ๋‹ค๊ณ  ์ฃผ์žฅํ–ˆ์Šต๋‹ˆ๋‹ค. ๊ทธ๋Š” ์ด ๊ณ„ํš์ด ์ฃผ ์ •๋ถ€์™€ ์„ํƒ„ ์‚ฐ์—…์— ๋ง‰๋Œ€ํ•œ ๋ถ€๋‹ด์„ ์ค„ ๊ฒƒ์ด๋ฉฐ, ๊ทœ์ œ ์ค€์ˆ˜๋ฅผ ์œ„ํ•œ ์‹œ๊ฐ„์ด ์ด‰๋ฐ•ํ•˜๋‹ค๊ณ  ๊ฐ•์กฐํ–ˆ์Šต๋‹ˆ๋‹ค. ๋˜ํ•œ, ์˜ค๋ฐ”๋งˆ ๋Œ€ํ†ต๋ น์ด ์˜์ง€ํ•˜๋Š” โ€˜์ฒญ์ • ๋Œ€๊ธฐ๋ฒ•(Clean Air Act)โ€˜์ด ๊ทธ๋Ÿฌํ•œ ๊ถŒํ•œ์„ ๋ช…ํ™•ํ•˜๊ฒŒ ๋ถ€์—ฌํ•˜์ง€ ์•Š์œผ๋ฉฐ, ์ด๋Š” โ€˜์ฃผ์š” ์งˆ๋ฌธ ์›์น™(major questions doctrine)โ€˜์— ์œ„๋ฐฐ๋  ์ˆ˜ ์žˆ๋‹ค๊ณ  ์–ธ๊ธ‰ํ–ˆ์Šต๋‹ˆ๋‹ค. ๋กœ๋ฒ„์ธ  ๋Œ€๋ฒ•์›์žฅ์€ EPA๊ฐ€ ๋ถˆ๊ณผ ๋ช‡ ๋‹ฌ ์ „ ์ˆ˜์€ ๋ฐฐ์ถœ๋Ÿ‰ ๊ทœ์ œ์™€ ๊ด€๋ จํ•˜์—ฌ ๋Œ€๋ฒ•์›์„ โ€œ์†์˜€๋‹คโ€๊ณ  ๋А๋ผ๋ฉฐ, โ€œ์ด๋ฒˆ์—๋Š” ๋‹ค์‹œ ๊ทธ๋Ÿฐ ์ผ์ด ์ผ์–ด๋‚˜์ง€ ์•Š๋„๋ก ํ•˜๊ฒ ๋‹คโ€๋Š” ๊ฐ•ํ•œ ๋ถˆ์พŒ๊ฐ์„ ๋“œ๋Ÿฌ๋ƒˆ์Šต๋‹ˆ๋‹ค. ๊ทธ๋Š” ์ „ํ†ต์ ์ธ ๋А๋ฆฐ ๋ฒ•์  ์ ˆ์ฐจ๊ฐ€ ์˜ค๋ฐ”๋งˆ ํ–‰์ •๋ถ€์˜ โ€˜๊ตํ™œํ•œโ€™ ๊ทœ์ œ๋ฅผ ํ—ˆ์šฉํ•˜๊ณ  ์žˆ๋‹ค๊ณ  ๋ณธ ๊ฒƒ์ž…๋‹ˆ๋‹ค.

  2. ๋ธŒ๋ผ์ด์–ด ๋Œ€๋ฒ•๊ด€์˜ ๋ฐ˜๋ฐ• (2์›” 4์ผ): ๊ฐ™์€ ๋‚ , ์Šคํ‹ฐ๋ธ ๋ธŒ๋ผ์ด์–ด ๋Œ€๋ฒ•๊ด€(Justice Stephen Breyer)์€ ๋กœ๋ฒ„์ธ  ๋Œ€๋ฒ•์›์žฅ์˜ ๋ฉ”๋ชจ์— ๋ฐ˜๋ฐ•ํ•˜๋Š” ์˜๊ฒฌ์„ ์ œ์‹œํ–ˆ์Šต๋‹ˆ๋‹ค. ๊ทธ๋Š” ์ฒญ์ • ์ „๋ ฅ ๊ณ„ํš์ด ๋‹น์žฅ ์‚ฐ์—…๊ณ„์— ์•„๋ฌด๊ฒƒ๋„ ์š”๊ตฌํ•˜์ง€ ์•Š์œผ๋ฉฐ, ์ „๋ฉด์ ์ธ ๊ทœ์ œ ์ค€์ˆ˜๋Š” 2030๋…„๊นŒ์ง€ ํ•„์š” ์—†์œผ๋ฏ€๋กœ ์„œ๋‘๋ฅผ ์ด์œ ๊ฐ€ ์—†๋‹ค๊ณ  ์ง€์ ํ–ˆ์Šต๋‹ˆ๋‹ค. ๋˜ํ•œ, ํ•ญ์†Œ๋ฒ•์›์ด ์•„์ง ํ•ฉ๋ฒ•์„ฑ์„ ๊ฒ€ํ† ํ•˜๋Š” ๋™์•ˆ ๋Œ€๋ฒ•์›์ด ๊ฐœ์ž…ํ•˜๋Š” ๊ฒƒ์€ โ€œ๋งค์šฐ ์ด๋ก€์ ์ธ ์ผโ€์ด๋ผ๊ณ  ๊ฐ•์กฐํ•˜๋ฉฐ, ํ•˜๊ธ‰ ๋ฒ•์›์ด ๋จผ์ € ์ฒ˜๋ฆฌํ•˜๋„๋ก ๋‘์–ด์•ผ ํ•œ๋‹ค๊ณ  ์ฃผ์žฅํ–ˆ์Šต๋‹ˆ๋‹ค.

  3. ๋กœ๋ฒ„์ธ  ๋Œ€๋ฒ•์›์žฅ์˜ ์žฌ๋ฐ˜๋ฐ• (2์›” 5์ผ): ๋กœ๋ฒ„์ธ  ๋Œ€๋ฒ•์›์žฅ์€ ๋‹ค์Œ ๋‚  ๊ณง๋ฐ”๋กœ ์žฌ๋ฐ˜๋ฐ• ๋ฉ”๋ชจ๋ฅผ ๋ณด๋ƒˆ๊ณ , ๊ทธ์˜ ์–ด์กฐ๋Š” ๋”์šฑ ์งœ์ฆ์Šค๋Ÿฝ๊ณ  ๋‹จํ˜ธํ•ด์กŒ์Šต๋‹ˆ๋‹ค. ๊ทธ๋Š” ๋ธŒ๋ผ์ด์–ด ๋Œ€๋ฒ•๊ด€์˜ ์ ˆ์ฐจ์  ์šฐ๋ ค๋ฅผ ์ผ์ถ•ํ•˜๋ฉฐ, โ€œ์ด ๊ทœ์ œ๋Š” ์ „๋ ฅ ๋ถ€๋ฌธ์— ๋ถ€๊ณผ๋œ ๊ฐ€์žฅ ๋น„์‹ผ ๊ทœ์ œ๋กœ ๋ฌ˜์‚ฌ๋œ๋‹คโ€๋ฉฐ ์ฆ‰๊ฐ์ ์ธ ์กฐ์น˜์˜ ํ•„์š”์„ฑ์„ ์ฃผ์žฅํ–ˆ์Šต๋‹ˆ๋‹ค. ๊ทธ๋Š” ๋Œ€๋ฒ•์›์ด ๊ฒฐ๊ตญ ์ด ๊ณ„ํš์„ ๊ฒ€ํ† ํ•˜์—ฌ ๋ฌดํšจํ™”ํ•  ๊ฒƒ์ด๋ผ๋Š” ํ™•์‹ ์„ ๊ฐ€์ง€๊ณ  ์žˆ์—ˆ์œผ๋ฉฐ, โ€œ๋ฒ•์›์ด ๊ถ๊ทน์ ์œผ๋กœ ์ด ์‚ฌ๊ฑด์„ ์‹ค์งˆ์ ์œผ๋กœ ์‹ฌ๋ฆฌํ•œ๋‹ค๋ฉด, ์ด ๋ฒ•์€ ์‚ด์•„๋‚จ์„ ๊ฐ€๋Šฅ์„ฑ์ด ๋งค์šฐ ๋‚ฎ๋‹ค๊ณ  ๋ณธ๋‹คโ€๊ณ  ๋ช…์‹œํ–ˆ์Šต๋‹ˆ๋‹ค. ์ฆ‰, ๊ทธ๋Š” ๋Œ€๋ฒ•์›์˜ ์ •์ƒ์ ์ธ ์‹ฌ์˜ ์ ˆ์ฐจ๋ฅผ ๊ฑฐ์น˜์ง€ ์•Š๊ณ ๋„ ์ด๋ฏธ ๋งˆ์Œ์„ ์ •ํ•œ ์ƒํƒœ์˜€์Šต๋‹ˆ๋‹ค.

  4. ์ผ€์ด๊ฑด ๋Œ€๋ฒ•๊ด€์˜ ๊ฒฝ๊ณ  (2์›” 8์ผ): ๋ฏผ์ฃผ๋‹น์ด ์ž„๋ช…ํ•œ ์—˜๋ ˆ๋‚˜ ์ผ€์ด๊ฑด ๋Œ€๋ฒ•๊ด€(Justice Elena Kagan)์€ ๋กœ๋ฒ„์ธ  ๋Œ€๋ฒ•์›์žฅ์˜ ์ž…์žฅ์— ๋ฐ˜๋Œ€ํ•˜๋Š” ๋ฉ”๋ชจ๋ฅผ ์ œ์ถœํ•˜๋ฉฐ ๊ฒฝ๊ณ ์Œ์„ ์šธ๋ ธ์Šต๋‹ˆ๋‹ค. ๊ทธ๋…€๋Š” ๋Œ€๋ฒ•์›์žฅ์ด ํ•˜๋ ค๋Š” ์ผ์ด โ€œ์ „๋ก€ ์—†๋Š”(unprecedented)โ€ ์ผ์ด๋ผ๋ฉฐ โ€œ์ง„์ •ํ•œ ์šฐ๋ ค(real pause)โ€œ๋ฅผ ํ‘œํ–ˆ์Šต๋‹ˆ๋‹ค. ๋˜ํ•œ, ์ด ์‚ฌ๊ฑด์€ โ€œ๋ณต์žกํ•œ ๋ฒ•๊ทœ ๋ฐ ๊ทœ์ œ ์ฒด์ œ์™€ ๊ด€๋ จ๋œ ์–ด๋ ค์šด ์‚ฌ๊ฑดโ€์ด๋ฏ€๋กœ ๋” ๋งŽ์€ ์‹œ๊ฐ„์ด ํ•„์š”ํ•˜๋‹ค๊ณ  ์ฃผ์žฅํ–ˆ์Šต๋‹ˆ๋‹ค.

  5. ์•Œ๋ฆฌํ†  ๋Œ€๋ฒ•๊ด€์˜ ์ง€์ง€ (2์›” 8์ผ): ๊ณตํ™”๋‹น์ด ์ž„๋ช…ํ•œ ์ƒˆ๋ฎค์–ผ ์•Œ๋ฆฌํ†  ๋Œ€๋ฒ•๊ด€(Justice Samuel Alito)์€ ๋กœ๋ฒ„์ธ  ๋Œ€๋ฒ•์›์žฅ์„ ์ง€์ง€ํ•˜๋ฉฐ, ๋ฒ•์›์ด ํ–‰๋™ํ•˜์ง€ ์•Š์œผ๋ฉด โ€œ์‚ฌ๋ฒ•๋ถ€์˜ ์ •๋‹น์„ฑ(institutional legitimacy)์ด ๋ฌดํšจํ™”๋  ์ˆ˜ ์žˆ๋‹คโ€๊ณ  ์ฃผ์žฅํ–ˆ์Šต๋‹ˆ๋‹ค. ๊ทธ๋Š” ์˜ค๋ฐ”๋งˆ ํ–‰์ •๋ถ€๊ฐ€ ๋Œ€๋ฒ•์›์„ ๋ฌด๋ ฅํ™”์‹œํ‚ค๋ ค ํ•œ๋‹ค๊ณ  ๋А๋ผ๋Š” ๋กœ๋ฒ„์ธ  ๋Œ€๋ฒ•์›์žฅ์˜ โ€˜๋ชจ์š•๊ฐโ€™์— ๊ณต๊ฐํ–ˆ์Šต๋‹ˆ๋‹ค.

์ด๋Ÿฌํ•œ ๋ฉ”๋ชจ๋“ค์€ ๋Œ€๋ฒ•๊ด€๋“ค์ด ๋ฒ•์ ์ธ ๋…ผ๋ฆฌ๋ณด๋‹ค๋Š” ๋‹นํŒŒ์  ๋ถˆ๋งŒ, ๋Œ€๋ฒ•์›์˜ ์œ„์ƒ์— ๋Œ€ํ•œ ์šฐ๋ ค ๋“ฑ โ€˜๋ฒ•๊ณผ ๋ฌด๊ด€ํ•œโ€™ ์š”์†Œ๋“ค์— ์–ผ๋งˆ๋‚˜ ๊นŠ์ด ์˜ํ–ฅ์„ ๋ฐ›๋Š”์ง€ ๋ณด์—ฌ์ค๋‹ˆ๋‹ค. ๊ทธ๋“ค์˜ ์ผ๋ฐ˜์ ์ธ ํŒ๊ฒฐ๋ฌธ์ด ์‚ฌ์‹ค๊ณผ ๊ด€๋ จ ๋ฒ•๋ฅ  ์ž๋ฃŒ์— ๊ตญํ•œ๋˜๋Š” โ€˜๊ฐ€๋ฉดโ€™์ด๋ผ๋ฉด, ์ด ๋น„๋ฐ€ ์„œ์‹ ๋“ค์€ ๋Œ€๋ฒ•๊ด€๋“ค์˜ ์‹ค์ œ ์˜์ œ๋ฅผ ์†”์งํ•˜๊ฒŒ ๋“œ๋Ÿฌ๋‚ด๋Š” ์ฐฝ๋ฌธ์ด์—ˆ์Šต๋‹ˆ๋‹ค.

์Šค์œ™ ๋ณดํ„ฐ์˜ ๊ฒฐ์ •๊ณผ ๊ทธ ์ดํ›„

๊ฒฐ๊ตญ, ๋ชจ๋“  ๊ฒƒ์€ ์Šค์œ™ ๋ณดํ„ฐ์˜€๋˜ ์•ค์„œ๋‹ˆ ์ผ€๋„ค๋”” ๋Œ€๋ฒ•๊ด€์—๊ฒŒ ๋‹ฌ๋ ค ์žˆ์—ˆ์Šต๋‹ˆ๋‹ค. ๋…ผ์Ÿ์˜ 5์ผ์งธ์ธ 2์›” 9์ผ, ๊ทธ๋Š” ๋‹จ ์„ธ ๋ฌธ์žฅ์œผ๋กœ ๋œ ๋งค์šฐ ์งง์€ ๋ฉ”๋ชจ๋ฅผ ๋ณด๋ƒˆ์Šต๋‹ˆ๋‹ค. โ€œ๋Œ€๋ฒ•์›์žฅ์—๊ฒŒ ์ฐฌ์„ฑํ•œ๋‹ค.โ€ ๊ทธ๊ฒƒ์œผ๋กœ ๋…ผ์Ÿ์€ ๋๋‚ฌ์Šต๋‹ˆ๋‹ค. ๋Œ€๋ฒ•์›์€ 5๋Œ€ 4๋กœ ์˜ค๋ฐ”๋งˆ ๋Œ€ํ†ต๋ น์˜ ํ•ต์‹ฌ ํ™˜๊ฒฝ ์ •์ฑ…์„ ์ฐจ๋‹จํ•˜๋Š” ๋ช…๋ น์„ ๋‚ด๋ ธ์Šต๋‹ˆ๋‹ค. ์šฐ๋ฆฌ๊ฐ€ ๋ฐฉ๊ธˆ ์‚ดํŽด๋ณธ ์ด ๋ชจ๋“  ๊ฐœ์ธ์ ์ธ ๋…ผ์Ÿ๊ณผ๋Š” ๋Œ€์กฐ์ ์œผ๋กœ, ๋Œ€์ค‘์—๊ฒŒ ๋ฐœํ‘œ๋œ ๋ช…๋ น์€ ๋ฒ•์  ์ƒํˆฌ์–ด๋งŒ ๋‹ด๊ธด ํ•œ ๋‹จ๋ฝ์งœ๋ฆฌ ์„ค๋ช… ์—†๋Š” ํŒ๊ฒฐ๋ฌธ์ด์—ˆ์Šต๋‹ˆ๋‹ค.

โ€˜๊ทธ๋ฆผ์ž ์‹ฌ์˜โ€™์˜ ์œ ์‚ฐ: ์‹ ๋ขฐ ์ƒ์‹ค๊ณผ ๋‹นํŒŒ์  ์‚ฌ๋ฒ•๋ถ€

์ด ๋น„๋ฐ€ ๋ฌธ์„œ๋“ค์€ ๊ทธ๋ฆผ์ž ์‹ฌ์˜์— ๋Œ€ํ•œ ๋น„ํŒ์„ ์ •๋‹นํ™”ํ•ฉ๋‹ˆ๋‹ค. ์ด๊ฒƒ์€ ๋Œ€๋ฒ•์›์ด โ€˜์ตœ๊ณ  ์ˆ˜์ค€์˜ ์ž‘์—…(A+ work)โ€˜์„ ํ•œ ๊ฒƒ์ด ์•„๋‹ˆ์—ˆ์Šต๋‹ˆ๋‹ค. ๋Œ€๋ฒ•๊ด€๋“ค์€ ๋ถˆ๋งŒ๊ณผ ์งœ์ฆ์— ์‚ฌ๋กœ์žกํ˜€ ์•„์ด๋””์–ด๋ฅผ ๋˜์ง€๊ณ , ์„œ๋กœ์—๊ฒŒ ๋‚ ์นด๋กญ๊ฒŒ ๋ฐ˜์‘ํ•˜๋ฉฐ, ๊ตญ๊ฐ€ ์ตœ๊ณ  ๋ฒ•์›์ด ํ•ด์•ผ ํ•  ์ข…๋ฅ˜์˜ ์ผ์„ ํ•˜์ง€ ์•Š๊ณ  ์žˆ์—ˆ์Šต๋‹ˆ๋‹ค. ๊ทธ๋“ค์€ ์ถฉ๋ถ„ํ•œ ์ •๋ณด ์—†์ด ๋„ˆ๋ฌด ๋น ๋ฅด๊ฒŒ ์›€์ง์˜€์œผ๋ฉฐ, ์•„๋ฌด๋„ โ€œ์ด๊ฒƒ์ด ์–ด๋””๋กœ ์ด์–ด์งˆ๊นŒ?โ€๋ผ๊ณ  ๋ฌป์ง€ ์•Š์•˜์Šต๋‹ˆ๋‹ค.

์ด๋Ÿฌํ•œ ๋ณ€ํ™”๋Š” ํŠนํžˆ ํŠธ๋Ÿผํ”„ ํ–‰์ •๋ถ€ ์‹œ์ ˆ์— ๊ทน๋ช…ํ•˜๊ฒŒ ๋“œ๋Ÿฌ๋‚ฌ์Šต๋‹ˆ๋‹ค. ๋Œ€๋ฒ•์›์€ ์ด๋ฏผ, ๊ธฐ๊ด€ ๊ถŒํ•œ, ์ •๋ถ€ ์ง€์ถœ, ํŠธ๋žœ์Šค์  ๋” ๊ตฐ์ธ ๋“ฑ ์ฃผ์š” ์‚ฌ์•ˆ์— ๋Œ€ํ•œ ์ˆ˜๋งŽ์€ ๊ธด๊ธ‰ ์š”์ฒญ์— ์‹œ๋‹ฌ๋ ธ๊ณ , ๋Œ€๋ถ€๋ถ„ ํŠธ๋Ÿผํ”„ ํ–‰์ •๋ถ€์˜ ์†์„ ๋“ค์–ด์ฃผ์—ˆ์Šต๋‹ˆ๋‹ค. ์˜ค๋ฐ”๋งˆ ํ–‰์ •๋ถ€ ๋•Œ๋Š” ์ฃผ์š” ๋Œ€ํ†ต๋ น ์ด๋‹ˆ์…”ํ‹ฐ๋ธŒ๋ฅผ ์ค‘๋‹จ์‹œ์ผฐ๋˜ ๊ฒƒ๊ณผ ๋‹ฌ๋ฆฌ, ๊ทธ๋ฆผ์ž ์‹ฌ์˜๋Š” ํŠธ๋Ÿผํ”„ ๋Œ€ํ†ต๋ น์˜ ์ •์ฑ…์„ ์••๋„์ ์œผ๋กœ ํ—ˆ์šฉํ–ˆ์Šต๋‹ˆ๋‹ค.

์ •์น˜ํ•™์ž๋“ค์€ ๊ทธ๋ฆผ์ž ์‹ฌ์˜์—์„œ โ€˜์ •์‹ ์‹ฌ์˜โ€™๋ณด๋‹ค ํ›จ์”ฌ ๋” ๋‹นํŒŒ์ ์ธ ํˆฌํ‘œ๊ฐ€ ์ผ์–ด๋‚œ๋‹ค๊ณ  ์ง€์ ํ•ฉ๋‹ˆ๋‹ค. ์ถฉ๋ถ„ํ•œ ์ˆ™๊ณ ์™€ ์‹ฌ์˜๊ฐ€ ๋ถ€์กฑํ• ์ˆ˜๋ก ๋Œ€๋ฒ•๊ด€๋“ค์˜ ๋‹นํŒŒ์  ์ถฉ๋™์ด ๋” ๊ฐ•ํ•˜๊ฒŒ ์ž‘์šฉํ•œ๋‹ค๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค. ์‹ค์ œ๋กœ ๋ฐ”์ด๋“  ํ–‰์ •๋ถ€ ์‹œ์ ˆ, ๋Œ€๋ฒ•์›์€ ์„ธ ๊ฑด์˜ ๊ธด๊ธ‰ ์š”์ฒญ์—์„œ ๋ฐ”์ด๋“  ๋Œ€ํ†ต๋ น์—๊ฒŒ ๋ฐ˜๋Œ€ํ•˜๋Š” ๊ฒฐ์ •์„ ๋‚ด๋ ธ์ง€๋งŒ, ์ด ์‚ฌ๊ฑด๋“ค์ด ๋‚˜์ค‘์— ์ •์‹ ์‹ฌ์˜๋ฅผ ๊ฑฐ์ณค์„ ๋•Œ๋Š” ๋ฐ”์ด๋“  ํ–‰์ •๋ถ€์˜ ์†์„ ๋“ค์–ด์ฃผ์—ˆ์Šต๋‹ˆ๋‹ค. ์ฆ‰, ์ˆ™๊ณ ํ•  ์‹œ๊ฐ„์ด ์ถฉ๋ถ„ํ• ์ˆ˜๋ก ๋‹นํŒŒ์  ์ถฉ๋™์€ ์•ฝํ•ด์ง€๋Š” ๊ฒฝํ–ฅ์„ ๋ณด์ž…๋‹ˆ๋‹ค.

์ˆ™๊ณ ๋Š” ๋˜ํ•œ ๋ฒ•์›์— ๋Œ€ํ•œ โ€˜๊ตญ๋ฏผ์˜ ์‹ ๋ขฐโ€™๋ฅผ ๊ตฌ์ถ•ํ•ฉ๋‹ˆ๋‹ค. ์šฐ๋ฆฌ๋Š” ์„ ์ถœ์ง ๊ณต๋ฌด์›๋“ค์„ ์„ ํƒํ•˜์ง€๋งŒ, ๋Œ€๋ฒ•๊ด€๋“ค์€ ์„ ํƒํ•˜์ง€ ์•Š์Šต๋‹ˆ๋‹ค. ๊ทธ๋“ค์€ ํ‰์ƒ ๊ถŒ๋ ฅ์„ ์œ ์ง€ํ•ฉ๋‹ˆ๋‹ค. ๋”ฐ๋ผ์„œ ํŒ์‚ฌ๊ฐ€ ์˜๊ฒฌ์„ ์ž‘์„ฑํ•˜๋Š” ํ–‰์œ„๋Š” โ€œ์ด ๊ฒฐ๊ณผ์— ๋™์˜ํ•˜์ง€ ์•Š๋”๋ผ๋„ ๋‚ด๊ฐ€ ์–ด๋–ป๊ฒŒ ์ด๋Ÿฐ ๊ฒฐ๋ก ์— ๋„๋‹ฌํ–ˆ๋Š”์ง€ ๋ณด์—ฌ์ค„ ํ…Œ๋‹ˆ ๋‚˜๋ฅผ ์‹ ๋ขฐํ•ด์•ผ ํ•œ๋‹คโ€๊ณ  ๋งํ•˜๋Š” ํ–‰์œ„์ž…๋‹ˆ๋‹ค. ํ•˜์ง€๋งŒ ๋Œ€๋ฒ•์›์ด ๊ทธ๋ฆผ์ž ์‹ฌ์˜๋ฅผ ํ†ตํ•ด ์ด๋Ÿฌํ•œ ๊ณผ์ •์„ ๋Œ€ํญ ์ถ•์†Œํ•˜๋ฉด์„œ, ์‚ฌ๋ฒ•๋ถ€์˜ ์ •๋‹น์„ฑ์—๋Š” ์‹ฌ๊ฐํ•œ ์œ„ํ—˜์ด ๋”ฐ๋ฆ…๋‹ˆ๋‹ค. ํ˜„์žฌ ๋Œ€๋ฒ•์›์— ๋Œ€ํ•œ ๊ตญ๋ฏผ์˜ ์‹ ๋ขฐ๋Š” ์—ญ๋Œ€ ์ตœ์ € ์ˆ˜์ค€์ด๋ฉฐ, ์ฃผ์š” ๊ฒฐ์ •์„ ๋‚ด๋ฆฌ๋Š” ๋ฐ ๊ทธ๋ฆผ์ž ์‹ฌ์˜์— ๋Œ€ํ•œ ์˜์กด๋„๊ฐ€ ๋†’์•„์งˆ์ˆ˜๋ก ์ด๋Ÿฌํ•œ ๋ฌธ์ œ๋Š” ๋”์šฑ ์•…ํ™”๋  ๊ฒƒ์ž…๋‹ˆ๋‹ค.

๋Œ€๋ฒ•์›์˜ โ€˜๊ทธ๋ฆผ์ž ์‹ฌ์˜โ€™๋Š” ๋‹จ์ˆœํ•œ ์ ˆ์ฐจ์  ํŽธ์˜๋ฅผ ๋„˜์–ด, ์‚ฌ๋ฒ•๋ถ€์˜ ๋ณธ์งˆ๊ณผ ์—ญํ• ์— ๋Œ€ํ•œ ๊นŠ์€ ์„ฑ์ฐฐ์„ ์š”๊ตฌํ•˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค. ๋น„๋ฐ€ ์„œ์‹ ์ด ๋“œ๋Ÿฌ๋‚ธ 5์ผ๊ฐ„์˜ ๊ฒฉ๋Œ์€ ๋ฏธ๊ตญ ๋Œ€๋ฒ•์›์ด ์–ด๋–ค ๋ฐฉํ–ฅ์œผ๋กœ ๋‚˜์•„๊ฐ€๊ณ  ์žˆ๋Š”์ง€, ๊ทธ๋ฆฌ๊ณ  ๊ทธ ๊ณผ์ •์—์„œ ๋ฌด์—‡์„ ์žƒ์–ด๊ฐ€๊ณ  ์žˆ๋Š”์ง€์— ๋Œ€ํ•œ ์ค‘์š”ํ•œ ๋‹จ์„œ๋ฅผ ์ œ๊ณตํ•ฉ๋‹ˆ๋‹ค.