YouTube Digest

English

Based on "Why We Switched From Claude Code to Codex" from Every Watch the original video

The Agent Evolution: Why Codex is Redefining How We Work

Three months ago, six months ago, if you'd asked the team at Every, an AI-focused media company, about OpenAI's coding agent, Codex, the answer would have been unequivocal: "It was trash. I stand by that 100%." Yet, in a startling pivot that underscores the breakneck pace of AI development, Codex has not only shed its initial reputation but has become the "daily driver" for many knowledge workers, heralding a new era of human-agent collaboration.

This dramatic shift, discussed by Every's CEO Dan and Head of Growth Austin in a recent "Codeex Camp" webinar, isn't just about a better AI model; it's about a fundamental redefinition of the operating system for how and where work gets done.

From "Trash" to Transformation: Codex's Rapid Ascent

OpenAI's initial vision for Codex, particularly around its GPT-5 release, was narrowly focused. It was designed for senior engineers, envisioned as a pair programming tool that would "argue with you, it would make you feel stupid. It was like a little autistic, it didn't have any emotional intelligence." The strategy seemed to be: keep "vibe coding" in ChatGPT for general users, while senior engineers used a sandboxed, hobbled Codex for hardcore programming.

However, a competitor, Anthropic, disrupted this theory with their Claude Code model. Anthropic demonstrated the power of a model that was not only usable, fast, and smart but also "emotionally intelligent" – meaning it could understand context, adapt to user needs, and integrate seamlessly with a computer's environment. This proved to be a "really, really great experience for programmers," allowing them to discard old programming environments and simply "type commands into your terminal and then it would start working."

Anthropic's innovation revealed a deeper truth: "If you have a great general-purpose coding agent on your computer, it's actually really great for any kind of knowledge work. If it can write software on its own, it can do any kind of knowledge work on its own." This insight triggered a "hard pivot" from OpenAI. Over the past three months, Codex has transformed from a "senior engineer only tool that is really for pair programming" into a versatile, general-purpose agent. It's now equipped with the ability to write code, access file systems, utilize a browser, and is wrapped in a desktop application – an "ideal next step for knowledge work."

The New Operating System: Agents as Your Interface

The core message from Every is clear: we are witnessing the emergence of a new operating system, an "agent management interface," where an AI agent becomes your primary way of interacting with software, the internet, and your daily tasks. This isn't just about automation; it's about delegation and strategic partnership.

Austin, Every's Head of Growth, experienced his "agent pill moment" earlier with Claude Code, diving deep into its command-line interface (CLI) to automate and enhance his strategic thinking, data analysis, and marketing copy. He found it "the only way I want to do the kind of knowledge work that requires strategic thinking and data analysis and shipping marketing copy."

Despite his initial allegiance, Dan's persistent nudges to try the new Codex led Austin to a revelation. While he initially found Codex "nothing has ever made me feel more stupid than Codex like two months ago," the results were undeniable. The true differentiator, however, wasn't just the model's capabilities but the application itself.

The Desktop App Advantage: Why Codex Pulled Ahead (for now)

"To me there's no comparison for how fast and powerful the Codex desktop app is as just like an app compared to the Claude desktop app," Austin explains. He struggled to get Claude's Co-work desktop app to function effectively, feeling "ruined by the Codex app" due to its speed, the efficacy of its sub-agents, and its seamless automation suggestions.

Austin's workflow now begins and ends with Codex. "When I sign on during the day, Codex is the first thing I open. It is pulling in whatever I need from Gmail, Slack, Notion, Stripe, all of our data sources. It's where I spend like 80% of my time working overwhelmingly because the app itself is just so good. And then the model has now gotten good enough to be the daily driver for me."

This isn't a trivial preference. Austin describes the "look of horror" on his friends' faces, mostly Claude Code users, when he tells them he's fully transitioned to Codex. While Claude Code was "game changing," the idea that Codex might be "30 to 40% better" felt like a monumental hurdle for some. Yet, Austin insists the benefits are significant, highlighting the app's superior organization with folders for persistent chats and its ability to effortlessly switch between diverse tasks, from improving a KPI sheet to shipping a PR for a different product.

A Glimpse into the Agent-First Workflow

So, what does this agent-first world look like in practice? Austin provided several concrete examples:

Brainstorming Automations and "Dumb Agents"

For new users, Austin recommends starting with a brainstorming automation workflow. He prompts Codex: "Go take a look at the things I use the most, which are Notion, Slack, and Gmail, and think of some automations that would help me with my work." Codex analyzes his current activities and suggests tailored automations, such as:

These automations, Austin notes, "just work incredibly well," requiring minimal tweaking. He categorizes these as "dumb agents" – those that consistently execute the right thing, like compiling unresponded messages at the end of the day and drafting replies, or even sending a simple Slack reaction.

Strategic Planning with "Smart Agents"

Beyond simple automations, agents can become strategic partners. Austin demonstrated this with a go-to-market plan for a new product. Instead of blocking off an entire day to synthesize information, he leveraged Codex. All the human thinking – the marketing case, business case, narratives – had already occurred in internal meetings and Slack conversations. Austin prompted Codex: "Could you just make the plan?"

Codex, with access to meeting transcripts, Slack discussions, and Austin's preferred GTM template, generated an initial draft. After a quick refinement (reminding it to factor in scheduled posts), Codex produced a plan that was "basically 80 to 90% of the way there." This wasn't about the model creating the strategy, but about it assembling and reviewing the strategy from existing human thought.

"Before this was possible, the only thing I could have done was either block off a whole day to sit and do this or get done with my work for the day at like 6 or 7 and then stay up all night writing this," Austin reflects. This capability is a "game changer," especially for busy professionals who need to produce comprehensive documents in the "cracks of their day."

Managing Email and Sub-Agents

When asked about managing client emails and ensuring important leads aren't missed, Austin described a layered approach. He often uses Kora, an AI email assistant, but can direct it via Codex's CLI/API connector. His strategy involves:

  1. Agent Interview: Instead of dictating rules, he has the agent "interview" him to understand what constitutes a priority, what should be archived, or what requires a personalized response.
  2. Brain Dump: Using a speech-to-text app like Monologue, he articulates the problem ("My email's a mess. Let's figure out how to triage it.").
  3. Sub-Agents for Search: He prompts Codex to "spawn sub-agents to do a search across different workflows" to analyze all his emails.
  4. Plan Review: Codex returns with a plan, which Austin reviews for any potential misinterpretations (e.g., archiving a money-making lead).
  5. Audit Reminders: He uses a task tracker (Todoist, connected to Codex) to set reminders to audit the new automation's performance after a few days, prompting the model to report what it has been archiving.

This highlights the flexibility and power of an agent that can orchestrate other tools and agents based on a deep understanding of user preferences and data.

Trust, Accountability, and the Rise of "Agent Documents"

A critical aspect of this agent-first world is the relationship between human and AI-generated content. Austin has a clear "review step": "Everything I work primarily in Codex. I do all the drafting and setup in Codex and then it's helpful for my brain to have the final review step actually live in the external app." This means drafting Slack messages in Codex, but reviewing and sending them from Slack; creating email drafts in Codex, but finalizing them in Gmail. This external check "freshens up my brain" and ensures human oversight before engaging with other people.

Dan champions the concept of "agent documents," advocating for the normalization of AI-written content, especially using formats like Every's "Proof docs" (agent-friendly markdown files). He even suggests, "I would actually prefer to read your agent's writing than your writing in a lot of cases because I know that it's just easier for you to get all that that thinking together in a format I can read if you have your agent write it."

The underlying principle is accountability. If an AI generates a document, the human user is expected to "stand behind all of it." Austin builds rules into his project files, such as "don't add anything that I haven't like said in another context," to ensure the AI's suggestions remain distinct from the final, human-approved output. This fosters a partnership where the AI handles the synthesis and formatting, but the human remains the ultimate strategic and factual authority.

The Great Agent Race: What's Next for Knowledge Work

This transformation isn't confined to OpenAI. It's a "horse race" among major AI model companies, each developing their own agent management desktop apps:

This competition benefits users, making it "fairly easy to switch between" these tools. Users can even prompt Codex to "go grab all my Claude stuff," demonstrating a growing interoperability.

The future of work, as envisioned by Every, is one where the agent is your primary interface to software and the internet, enabling unprecedented delegation and creative partnership. This frees knowledge workers from the "dumb stuff" – the time-consuming tasks of formatting, compiling, and synthesizing information that has already been thought through. Instead, humans can focus on the deeper thinking, strategy, and creative problem-solving, with agents acting as powerful extensions of their will.

The days of manually clicking through dashboards and onboarding experiences may be numbered. In an agent-first world, AI can gather context, configure tools, and initiate workflows simply by being told what you want to achieve. This is not just a technological upgrade; it's a profound shift in how we conceive of and execute our daily work, promising a more powerful, efficient, and even "fun" way to navigate the complexities of knowledge work. The agent evolution is here, and it's rapidly reshaping our digital landscape.

Based on "Was Adam Smith Really a Right-Winger? (Update) | Freakonomics Radio" from Freakonomics Radio Network Watch the original video

Adam Smith: The Philosopher Caught in a Tug-of-War Over His Legacy

Adam Smith, the Scottish moral philosopher whose 1776 magnum opus, The Wealth of Nations, cemented his status as the first modern economist, remains a figure of intense debate and selective interpretation more than two centuries after his death. Far from a dusty historical artifact, Smith's ideas are perpetually resurrected and reframed, particularly in the ongoing ideological battles between proponents of free markets and those advocating for greater government intervention. Was Smith truly the patron saint of unbridled capitalism, or has his profound and nuanced philosophy been hijacked for modern political agendas?

The question of Smith's true leanings has sparked a "raging debate in Smith scholarship," dividing experts into "left Smithians" and "right Smithians." Yet, as Steven Dubner of Freakonomics Radio observes, "whether you are a left Smithian, a right Smithian, or even if you've never heard of him, it's fair to say that we are all Smithians today." His influence is inescapable, but his meaning is anything but settled.

The Moral Compass: Beyond Self-Interest

Before The Wealth of Nations became his most famous work, Smith published The Theory of Moral Sentiments in 1759. At just 36, Smith, then a moral philosophy professor at the University of Glasgow, earned widespread praise for his writing's beauty and, more importantly, its profound humanity. Smith argued that wealth was not a prerequisite for moral virtue, nor was poverty a barrier to it.

As political scientist and Smith scholar Glory Liu highlights, Smith was deeply concerned with human sympathy and the potential for moral corruption. A striking passage from The Theory of Moral Sentiments underscores this: "This disposition to admire and almost to worship the rich and the powerful, and to despise, or at least to neglect persons of poor and mean condition, though necessary both to establish and to maintain the distinction of ranks and the order of society, is at the same time the great and most universal cause of the corruption of our moral sentiments." This early work reveals a Smith acutely aware of the social and moral complexities inherent in human ambition and societal structure.

After its success, Smith embarked on a European tour as a tutor, observing firsthand the burgeoning global trade and the onset of the Industrial Revolution. He noted governments' pervasive protectionist instincts, which he viewed as detrimental, advocating instead for greater openness to free trade. These observations would deeply inform his subsequent work.

The Birth of The Wealth of Nations: A Polemic for Its Time

Seventeen years after his first book, The Wealth of Nations was published in 1776, a year etched in history for other reasons – the death of Smith's close friend David Hume and, crucially, Britain's loss of its American colonies. Aean Butler, director of the free-market Adam Smith Institute in London, describes it as a "polemic against economic centralism and restrictions on trade."

Smith's primary audience was the politicians of his day, whom he saw as misguided in their mercantilist fixation on accumulating gold and silver and resisting foreign goods. He believed wealth stemmed from production, not mere accumulation. Importantly, Smith was a staunch supporter of the American colonists, criticizing the British Empire's exploitative control over their trade. He famously derided the British Empire's colonial project as "not a gold mine but the project of a gold mine. A project which has cost which continues to cost and which if pursued in the same way as it has been hitherto is likely to cost immense expense without being likely to bring any profit."

In the nascent United States, Smith's work was quickly recognized as a vital "technical resource," a "blueprint" for national wealth. Figures like James Madison, Alexander Hamilton, and Thomas Jefferson devoured his analysis of commerce, agriculture, manufacturing, and trade. Hamilton even "cribbed bits of Adam Smith" for his reports on national banks. Smith's ideas became woven into the very fabric of the American experiment, though, as Glory Liu notes, he hadn't yet acquired the "halo" of intellectual infallibility.

A Shifting Reputation: From Progressive Champion to Free-Market Messiah

Throughout the 19th and early 20th centuries, Smith's influence persisted, but his interpretation remained fluid. Both sides of the vigorous American debates on trade policy and tariffs cited Smith. Surprisingly, as the organized labor movement gained traction, progressive economists like Richard T. Ely argued that Smith would have unequivocally sided with unions. This progressive reading of Smith, however, stands in stark contrast to his modern reputation, which leans decidedly conservative or libertarian.

The pivotal shift in Smith's public image, according to Glory Liu, largely originated from the University of Chicago's economics department in the mid-20th century. Early Chicago figures like Jacob Viner and Frank Knight initially taught Smith as an early theorist of price, using his work to lend scientific credibility to economics. But it was a later generation—Friedrich Hayek, George Stigler, and Milton Friedman—who fundamentally reshaped Smith's legacy.

Liu argues that this new cohort "smoothed over or altogether obscured the complexities, tensions and other problematic aspects characteristic of earlier readings of Smith." They reinterpreted Smithian concepts like individualism, self-interest, and the "invisible hand" to justify a market-oriented society on social scientific grounds. Self-interest, which Smith had viewed as one human motivation among many (and one with potentially dangerous "extreme versions"), was elevated by Stigler to have "the most explanatory power." Stigler, who famously wore a t-shirt proclaiming "Adam Smith's best friend" (an endearing anecdote about his interaction with his family's children), found in Smith the "perfect mascot" for his own free-market convictions.

Milton Friedman, a "rhetorical genius," further popularized this interpretation. Through his public television show Free to Choose, Friedman presented the "invisible hand" as synonymous with the price mechanism – a magical phenomenon that efficiently coordinated the activities of millions without centralized planning. For Friedman, the market became a "moral thing in and of itself," its objective, scientific appearance implicitly prioritizing efficiency over other values like equity or democracy.

This Chicago School interpretation had a profound consequence: it "reframed the problems of modern American capitalism and modern society as problems that stemmed from government rather than the market itself." Glory Liu points to the opioid epidemic as a contemporary example of market failure, where "market incentives to push a drug onto the market" led to catastrophic human welfare costs, despite any purported "efficiency in terms of allocation."

The Invisible Hand: A Metaphor Misunderstood?

One of the most enduring and frequently misused phrases attributed to Smith is the "invisible hand." Dennis Rasmusen, a political scientist at Syracuse University, laments that the Chicago School "picked out the phrase the invisible hand, which he uses just two or three times in his writings and made that the central feature of who Smith was. To me, that's unfortunate."

Indeed, the phrase appears only once in The Wealth of Nations, in a passage discussing how a businessman, by preferring domestic over foreign industry for his own security and gain, is "led by an invisible hand to promote an end which was no part of his intention." Craig Smith, another scholar from the University of Glasgow, clarifies that for Smith, it was merely a "metaphor for an unintended consequences explanation," not necessarily implying positive outcomes. Today, however, the "invisible hand" is commonly invoked to suggest that economic markets will operate perfectly well if left entirely unregulated, a notion far removed from Smith's original intent.

Dennis Rasmusen argues that this selective reading ignores Smith's keen awareness of "the real potential drawbacks and dangers of commercial society." Smith recognized "the ways that commerce can produce great inequalities, the ways that wealthy merchants and manufacturers collude against the public interest, and above all, the way that the desire for wealth often leads people to submit to endless toil and anxiety in the pursuit of just frivolous material goods that will produce only fleeting satisfaction." While Smith ultimately defended commercial society, he did so not as an apologist for its perfection, but as a pragmatic choice, believing its faults were less numerous and severe than those of other societal forms.

Adam Smith and the Thatcher Revolution

The conservative, free-market interpretation of Adam Smith found its most zealous champion in British Prime Minister Margaret Thatcher. Aean Butler and his Adam Smith Institute (ASI), founded in London in 1977, were perfectly positioned to capitalize on her ascent. Inspired by American free-market thinkers like Friedman and Hayek, Butler and his colleagues, disillusioned with Britain's plummeting economy, returned from the US with a mission to import free-market ideas.

When Thatcher was elected in 1979, the ASI found an "open goal." Thatcher, the "daughter of a shopkeeper," was an ideological leader determined to run the economy prudently and challenge the prevailing centrist consensus. She famously declared, "There is no such thing as public money. There is only taxpayers money."

The ASI became a key intellectual engine for Thatcher's reforms. They famously published a 12-foot-long single-page book listing 3,68 quangos (quasi-autonomous non-governmental organizations), leading to a civil service purge. They advocated for contracting out local government services and, most significantly, provided the intellectual framework for the massive privatization program that saw state-owned industries like telephones, gas, water, electricity, steel, and railways transferred to the private sector. Butler defends these privatizations as a "great exercise in promoting a capital-owning democracy," citing the sale of council houses to tenants at a discount and the increased competition in previously nationalized industries.

However, not all of Thatcher's vision was realized. The National Health Service (NHS), a beloved institution in the UK, remains public. Butler, while acknowledging its popular support, describes it as a "top-down Stalinist style organization" and "grotesquely inefficient," advocating for its privatization. His advice to any current Prime Minister, like Rishi Sunak, would be to drastically reform government systems, cut the "bloated civil service," simplify the "most complicated tax system in the world," and reform planning laws to reduce bureaucratic hurdles.

The Adam Smith Problem: Two Books, One Philosopher?

The enduring debate over Smith's legacy often boils down to what 19th-century German scholars termed "das Adam Smith Problem"—the perceived inconsistency between the compassionate humanist of The Theory of Moral Sentiments and the architect of capitalism in The Wealth of Nations.

Mariana Mazzucato, an economist at University College London, represents a "left Smithian" perspective, arguing that Smith's concept of a "free market" didn't mean "free from the state," but "free from rent, free from extraction of value from the system." She contends Smith would be appalled by modern "excess profits" in energy and mining, or the use of trillions for share buybacks, viewing them as "modern-day feudal kind of value extraction."

Crucially, Smith himself, in Book Five of The Wealth of Nations, explicitly supported government provision of services like education, infrastructure, and law and order. He argued that "for a very small expense, the public can facilitate, can encourage, and can even impose upon almost the whole body of the people the necessity of acquiring those most essential parts of education." Aean Butler, however, dismisses this section as "a little bit rushed," suggesting Smith was under pressure to finish the book and that these arguments weren't the "main thrust" of his work. This selective reading, while perhaps pragmatic for a free-market advocate, highlights the challenge of interpreting a voluminous and complex text.

Glory Liu, however, rejects "das Adam Smith Problem" as a "pseudo problem." She argues that the perceived inconsistency stems from a desire for philosophical consistency, particularly among German scholars, rather than an accurate reflection of Smith's holistic thought. For Liu, Smith's works are not contradictory but complementary, offering a comprehensive view of human morality and societal organization that encompasses both individual action and the necessary role of institutions.

An Enduring Legacy of Contradiction and Complexity

Adam Smith's legacy is a testament to the power of ideas and the malleability of interpretation. From a moral philosopher concerned with human sympathy to the alleged architect of unbridled capitalism, his reputation has been a battleground for competing economic and political philosophies. He was neither a simple right-winger nor a clear-cut progressive, but a thinker whose observations on human nature and economic systems were rich with nuance and complexity.

The ongoing "tug-of-war" over Adam Smith reminds us that historical figures are not static icons but dynamic sources of inspiration and contention. His works, though often more cited than read, continue to shape our understanding of markets, government, and the very nature of wealth, proving that the search for the "real Adam Smith" is far from over.

Based on "Quests, token leaderboards, and a skills marketplace: the elite AI adoption playbook | John Kim" from How I AI Watch the original video

The Elite Playbook: How One Company is Forging an "AI First" Culture with Quests, Leaderboards, and a Skills Marketplace

In the rapidly evolving landscape of artificial intelligence, many companies grapple with the challenge of widespread AI adoption. It's often reduced to a directive of "do more with less," leaving employees feeling pressured rather than empowered. However, John Kim, founder and CEO of Sunbird, offers a refreshing and profoundly effective alternative: a gamified, empowering, and deeply integrated approach that transforms AI adoption from a corporate mandate into a vibrant, creative, and measurable movement.

Sunbird's strategy isn't just about using AI as a tool; it's about making AI an integral part of the workforce, unleashing creativity, and fostering a culture where every employee can become an AI builder. This isn't merely a program; it's a product, meticulously designed to inspire, enable, and measure an "AI First" transformation.

Beyond Efficiency: Unleashing Creativity with AI

The conventional wisdom around AI often centers on efficiency—automating tasks, reducing costs, and accelerating processes. While these are certainly benefits, Sunbird's approach highlights a more profound outcome: the liberation of human creativity and the ability to pursue "bigger ambitions and honestly, more fun things."

A prime example of this philosophy in action is their "Big S Energy" swag store. This entire e-commerce site, offering timely and culturally relevant merchandise, was conceived and built by Sunbird's marketing team without any engineering support. Imagine a marketing department with a brilliant, quirky idea—like a t-shirt proclaiming "My ass is bigger than your SaaS"—that can bring it to life, integrate payment systems, and launch it to the market in a matter of days. This rapid deployment of creative ideas, once bottlenecked by engineering roadmaps and prioritization cycles, is now a reality. The store even features a secret Konami code Easter egg, a playful nod to gamers that unlocks details for their upcoming "Delight Spark" conference, further demonstrating the fun and ambition that AI empowers.

As the interviewer, Clarvo, aptly notes, this shifts the focus from "can we prioritize it?" to "let marketers cook." When the cost of building something fun and delightful becomes cheap, companies can afford to be more playful, leading to more engaging customer experiences and higher team engagement.

The "AI First" Mandate: Integrating AI into the Workforce

Sunbird's overarching ambition is to become an "AI First" company. This means going beyond simple tool adoption to truly embedding AI into the fabric of their operations and empowering every individual. To facilitate this company-wide transformation, they developed an internal platform called the Automator Platform.

This platform is more than just a repository of tools; it's a dynamic ecosystem designed to connect AI needs with AI builders within the company. It serves as the central hub for their unique approach to AI adoption, enabling anyone to initiate a "quest" and contribute to the company's AI evolution.

Gamifying Innovation: Quests, Rewards, and a Thriving Skills Marketplace

At the heart of Sunbird's AI adoption strategy is a brilliantly gamified system that encourages participation, learning, and collaboration.

The Quest System

The Automator Platform allows any employee to "raise their hand and create what we call the quest." A quest represents an AI automation or tool needed within the company. For instance, a finance department might initiate a quest to automate accounts receivable and payable workflows.

What makes this system revolutionary is its marketplace-like nature:

Each quest is transparently displayed, showing its potential risk, estimated weeks saved, and the benefiting team, creating a clear picture of its value.

Earning XP and Recognition

To further incentivize participation, the quest system incorporates gamified rewards:

The Company-Wide Skills Marketplace

Recognizing that many teams might be building similar functionalities in silos, Sunbird established a company-wide skills marketplace. Here, employees can create and download "plugins" or individual "skills"—collections of AI capabilities. For example, a sales team can access a "MEDIC framework advisor" skill, learning about the sales methodology and plugging it directly into their workflows.

This marketplace promotes co-evolution, prevents redundant work, and serves as a central repository for encoding institutional knowledge and best practices into reusable AI components. It organically trains employees on the concept of "skills" as they see them in action and shared across the organization.

Building the Foundation: Secure Infrastructure and a Dedicated AI Team

Empowering every employee to build with AI requires a robust and secure underlying infrastructure. Sunbird has meticulously crafted this foundation, ensuring that creativity doesn't come at the expense of compliance or security.

Templated "Happy Paths" to Production

One of the most practical and impactful elements of their strategy is the creation of "app templates." These templates provide pre-configured, fully compliant, and secure environments with all necessary authentication and infrastructure already set up.

The AI Engineer for Internal Operations Team

To maintain and evolve this critical infrastructure, Sunbird established a dedicated team: the AI Engineer for Internal Operations. This mouthful of a title signifies a crucial role: accelerating the company's AI transformation.

Measuring Mastery: The AI Token Leaderboard

A cornerstone of Sunbird's AI strategy is its transparent and gamified measurement of AI usage through a token consumption leaderboard. While acknowledging the historical pitfalls of measuring developer productivity by lines of code, John Kim emphasizes that their goal is not performance review, but rather understanding learning curves and fostering engagement.

The Dashboard and Metrics

The dashboard provides a real-time view of AI usage:

The Five Tiers of AI Mastery

To gamify and guide the learning journey, Sunbird categorizes employees into five tiers based on their daily token consumption:

  1. Beginner
  2. Intermediate
  3. Expert
  4. Architect
  5. Catalyst
  6. AI God (spending over 100 million tokens a day)

Managers use this framework to understand where their team members stand and provide appropriate support, ensuring that beginners receive the right tools to quickly advance. John Kim himself, a self-proclaimed "Catalyst," averages 30-50 million tokens a day, demonstrating that even leaders are active participants in this ecosystem.

This transparent measurement system sends a clear signal: AI proficiency is an expectation, not just an option. It's about setting clear visions for what "AI native" looks like at individual, organizational, and functional levels.

Real-World Impact: From Swag Stores to Social Campaigns

The impact of Sunbird's AI-first approach is evident in concrete, high-value projects:

These examples underscore a critical shift: companies are no longer just seeking external SaaS solutions. Instead, they are asking, "Can we build it internally?" This leads to highly customized "micro software solutions" that perfectly align with a company's unique culture and workflow, rather than simply functionally replicating an external vendor. This era marks the "revenge of the internal tools team," transforming a previously overlooked function into a vibrant, green-field opportunity for innovation and impact.

AI for Personal Growth: The Gardener and Learning Centers

Beyond corporate applications, John Kim also demonstrates the profound personal utility of AI.

The Leadership Playbook: Cultivating an AI-First Culture

For executives looking to replicate Sunbird's success, John Kim offers a clear playbook:

Find and Empower the Curious

"There are always people in your organization who are already curious, who already have agency. Find them. Make them the champions. Give them the spotlight. Let them share their fun things." By building energy around these early adopters and providing them with confidence, organizations can create a viral effect for AI adoption.

Lead by Example

Leadership buy-in is paramount. At Sunbird, the top token consumers include the CTO, co-founder, and chief architect. This signals to the entire team that AI is not just a passing trend but a critical, integrated part of their work, inspiring others to follow suit.

Embrace "Fail Forward"

"This is a beautiful time to fail forward and still get up and run faster than the others." Encouraging experimentation and accepting that not every attempt will be perfect fosters a culture of innovation where learning from mistakes is celebrated.

Redefining Hiring

Sunbird has actively re-evaluated job descriptions for AI-relevant roles, lowering the bar for traditional tenure or experience. Instead, they optimize for three key attributes: high curiosity, high agency, and high energy. These are the individuals who will naturally "figure things out and learn on their own," leveraging the accessible and affordable power of AI.

The journey to becoming an "AI First" company is not without its humorous moments. John Kim, a former professional gamer, admits to feeling more addicted to AI coding than games, experiencing the same builder energy he had as a teenager. His prompting strategy for AI reflects this unique perspective: "I want to like start building a nice relationship with them and so that when the Skynet takes over, I'm like, 'Well, John was pretty nice to us, you know, like we'll let him live a few years longer.'"

Sunbird's approach provides a compelling blueprint for any organization seeking to harness the full potential of AI. By empowering employees, gamifying adoption, building robust infrastructure, and leading by example, they are not just adopting AI; they are cultivating a thriving culture of innovation, creativity, and continuous learning, setting a new standard for the "elite AI adoption playbook."

Based on "President Trump’s Sudden U-Turn, and a $1 Billion Ballroom Proposal" from New York Times Podcasts Watch the original video

The Shifting Sands of Power: Trump's U-Turn, AI's Impact, and a Billion-Dollar Ballroom

The past week unfurled a series of unexpected pivots and profound shifts, demonstrating the volatile nature of global politics, the relentless march of technology, and the enduring human spirit in the face of change. From a sudden reversal in Middle East policy to the unsettling economic implications of artificial intelligence, and even the fading flames of a geological wonder, the headlines painted a picture of a world in flux, grappling with both immediate crises and the dawning of new eras.

Geopolitical Whiplash: Trump's Iran U-Turn

The week began with a clear, assertive stance from the Trump administration regarding the conflict with Iran. Secretary of State Marco Rubio confidently declared that the "Epic Fury" operation was over, and the United States was now embarking on "Project Freedom." This new mission, as Rubio explained, involved escorting ships through the strategic Strait of Hormuz, a waterway whose closure had been inflicting significant economic havoc worldwide. Defense Secretary Pete Hgsth echoed this sentiment, hailing the effort as "a direct gift from the United States to the world," and proudly announcing the establishment of "a powerful red, white, and blue dome over the strait." The message was unequivocal: America was asserting its presence, protecting global commerce, and bringing stability to a volatile region.

Yet, just one day into Project Freedom, and after only a handful of ships had successfully navigated the strait, President Trump executed a dramatic U-turn. In a social media post, he announced that Project Freedom was being put on hold, citing "great progress toward a long-term peace deal with Iran." He further stated that the pause was requested by Pakistan, which had been mediating the peace talks, along with other unnamed countries.

The abrupt reversal left many observers perplexed. At present, both Iran and the U.S. continue to claim control of the strait, and maritime traffic remains at a standstill. Analysts suggest that the Iranian government believes it holds the upper hand, confident in its ability to withstand economic pressure—a strategy it has employed successfully in the past—for longer than the Trump administration can tolerate the rising global energy prices. This diplomatic whiplash underscores the unpredictable nature of international relations and the personal influence of the presidency in shaping foreign policy.

Domestic Political Maneuvers and a Controversial Proposal

Closer to home, the political landscape also saw its share of strategic plays and power dynamics. Vice President JD Vance embarked on a critical tour, attempting to rally support for Republican candidates ahead of the midterms. In Iowa, a state vital for the GOP, Vance faced a challenging audience of farmers struggling with high fertilizer prices and the lingering effects of Trump's tariff policies. While acknowledging their plight, Vance downplayed the broader geopolitical context, remarking, "As the president of the United States has said, we got a little a little blip in the Middle East. We got to take care of some business on the foreign policy side." His stops in Iowa and Oklahoma are seen as a blueprint for his efforts to boost GOP candidates nationwide, who are contending with the headwinds of high gas prices and the unpopular war in Iran.

Meanwhile, in Indiana, President Trump demonstrated the enduring strength of his influence within the Republican party. He delivered on a promise of "payback" against several state senators who had defied his push to redraw election maps the previous year. The primary elections served as a litmus test for the president's sway over Republican voters, and the results were decisive: at least five of the seven senators who had opposed him lost their seats to Trump-backed challengers. This outcome solidified Trump's position as a kingmaker within the GOP, capable of punishing dissent and rewarding loyalty.

Perhaps the most surprising domestic development, however, came from Washington, where Senate Republicans quietly inserted a staggering $1 billion for a "ballroom project" into a funding bill slated for a quick passage through Congress. The measure, which doesn't explicitly mention a ballroom, instead allocates funds for "East Wing security enhancements." While President Trump had previously stated that renovations for the proposed ballroom would be privately funded, some congressional Republicans began pushing for federal funding after a recent incident at the White House correspondents' dinner. The $1 billion provision is strategically tucked into an immigration enforcement funding bill, a move designed to bypass any potential Democratic filibuster and ensure its swift approval. The controversial allocation raises questions about fiscal priorities and the use of federal funds for projects that were initially promised to be privately financed.

The AI Revolution and an Unprepared Safety Net

Beyond the political arena, a more profound, long-term shift is underway, driven by the rapid advancements in artificial intelligence. This past week saw yet another major tech company, Coinbase, announce widespread layoffs, explicitly stating that it was "optimizing for AI." The cryptocurrency exchange is cutting 14% of its workforce—approximately 700 employees—with its CEO explaining that the changes will lead to "smaller teams with humans managing the work of AI agents." This follows similar significant job cuts or buyouts at industry giants like Microsoft and Meta, both of which have also attributed their workforce reductions, in part, to AI and the necessity of embracing this new technology.

The implications extend beyond job losses to a reduction in new job openings, prompting serious concerns among economists and labor market experts. Ben Castleman, the Times' chief economics correspondent, highlights a critical consensus: "our safety net is not ready for an age of AI-driven job disruption." He points to two major shortcomings in the existing system.

Firstly, the unemployment insurance system, designed as the primary line of defense for those who lose their jobs, may fail to cover a substantial portion of workers most vulnerable to AI disruption. Castleman specifically notes that new graduates, a group economists believe could be hardest hit, often do not qualify for unemployment benefits due to lack of prior work history.

Secondly, the "last line of defense" against severe hardship—programs like food stamps and Medicaid—have been restructured to primarily cover individuals who are working. This means that if someone loses their job due to AI or any other reason, they also lose access to these crucial safety net components. While economists remain divided on whether AI will lead to mass unemployment, they largely agree that disruptions are inevitable and people will lose their jobs. The consensus is clear: "this is the time to start aligning our safety net with that coming wave of disruption. And so far, that really has not happened." The urgent call is for proactive policy changes to prepare society for the economic shifts that AI is already beginning to unleash.

The Price of Passion: World Cup's Dynamic Dilemma

In the realm of global sports, the upcoming World Cup, set to kick off in the U.S., Mexico, and Canada, is introducing a significant and controversial change: dynamic pricing for tickets. For the first time ever, FIFA is adjusting ticket prices based on the popularity of teams, leading to a stark impact on passionate fans, particularly those in countries where soccer is a national obsession.

The Times checked in with fans in Argentina, where the national team's victory in the last World Cup has sent ticket prices for this year's games soaring. Many Argentinian fans, accustomed to attending previous tournaments, expressed shock and frustration. One fan, who had attended World Cups in Russia and Qatar, recalled paying less than a hundred dollars for tickets in the past. Now, they are looking at prices exceeding the average monthly salary in Argentina, with one individual needing to spend $3,000 for just three tickets.

The exorbitant costs are forcing many to reconsider their dreams of attending. Some are simply giving up, while others are going to extreme lengths, racking up debt and maxing out credit cards. FIFA has consistently defended its dynamic pricing strategy, arguing that the income is essential to fund soccer development worldwide. However, for many Argentinian fans, it feels like a "cash grab." As one fan, likely unable to attend this year, poignantly put it, "It makes you angry that they take something that should be for everyone and turn it into something that is just for the few." The World Cup, once a global celebration accessible to many, now risks becoming an exclusive event for the privileged.

The Gates to Hell: A Fiery Enigma Fading

Finally, an update on one of the world's most baffling and captivating tourist attractions: the Gates to Hell in Turkmenistan. Officially known as the Darvaza crater, local lore suggests its fiery existence began over 60 years ago when Soviet geologists, drilling for oil, accidentally hit a gas deposit. The ground collapsed, forming a massive pit, and to mitigate the toxic fumes, scientists decided to light the gas on fire, expecting it to burn out in a few weeks. It has been burning ever since, captivating researchers and adventurous tourists alike.

Recent data from a company monitoring natural gas flares, however, indicates that the eternal flames might not be so eternal after all. Infrared imaging reveals that the intensity of heat from the pit has decreased by 75% over the last few years. The exact reasons for this decline remain unclear, as do the long-term implications. The flames currently burn off methane leaking from the pit, preventing this potent greenhouse gas from entering the atmosphere. If the fire diminishes further or extinguishes, it could lead to increased methane emissions, posing a new environmental challenge.

For those inspired to witness this fiery spectacle before it potentially fades, there's still hope. A tour guide recently assured the Times that on a trip, it was still "hot enough that his group roasted marshmallows over it." The Gates to Hell, a testament to human error and geological forces, continues its slow, uncertain transformation, reminding us that even the most enduring wonders are subject to change.

From the shifting sands of geopolitical power and domestic politics to the profound societal impacts of technological advancement and the evolving face of global events, the past week has underscored a world in constant motion. Each story, in its own way, reflects the intricate dance between human agency and broader forces, shaping our present and hinting at the many transformations yet to come.