An Interview with OpenAI CEO Sam Altman About DevDay and the AI Buildout – Stratechery by Ben Thompson

Sam Altman's Audacious Bet: Decoding OpenAI's Unified Vision, Infrastructure, and Future Blueprint

Excerpt: Amidst multi-trillion dollar ambitions and massive, seemingly disconnected bets on chips, hardware, and data centers, what is Sam Altman's grand, unifying theory for OpenAI's future, and is it a blueprint for a revolution or a high-stakes gamble?

In recent times, every move by OpenAI, from the intensive announcements at DevDay to a series of market-shaking deals in chips and infrastructure, has captivated the global tech sphere. Do these seemingly disparate, momentous actions conceal a grand, unifying theory? With this question in mind, renowned tech analyst Ben Thompson engaged in a profound dialogue with OpenAI CEO Sam Altman on his blog, Stratechery.

At the heart of this interview lies not only the revelation of OpenAI’s clear trajectory towards AGI but also a vivid illustration of how Sam Altman, a natural-born investor, is orchestrating the blueprint for this once-in-a-lifetime technological revolution with an extraordinary perspective.

Grand Unifying Theory: An AI Assistant Permeating Your Life

When questioned whether OpenAI was mirroring Windows’ history — crafting user interfaces with one hand while driving a vast hardware ecosystem with the other — Sam Altman didn't fully concur. He posited that the value of historical analogies lies in discerning differences, and OpenAI's core vision, he asserted, is unprecedented: to furnish the vast majority of people with a singular AI service that permeates every facet of their lives.

Within this blueprint, ChatGPT transcends being merely an application; it functions as the central gateway to this service, an omnipresent "AI helper." It necessitates the integration of third-party apps and services, concurrently leveraging a robust API business to ensure users receive a consistent, seamless intelligent experience across all scenarios.

However, the path to realizing this vision is "brutally difficult." Sam Altman candidly admitted that he currently dedicates most of his time to infrastructure buildout, as merely meeting the existing demands of ChatGPT presents an formidable challenge. Powering his audacious "company scale bet" is an unprecedented optimism regarding research advancements. In his view, the dawn of technological breakthroughs is imminent, making it the opportune moment to extensively lay the groundwork for products and business endeavors.

"Company-Scale Bets": An Investor's Capital and Compute Strategy

The market's fervor for OpenAI borders on frenzy; any company partnering with it witnesses a surge in market capitalization, an "Midas touch" effect that even Sam Altman himself finds "very strange." Yet, he keenly wagers that the AI market, even on the consumer side, will defy a winner-take-all dynamic. He prefers to analogize AI to a "transistor"—it will silently permeate every consumer and enterprise product, much like its electronic counterpart.

Against this backdrop, Sam Altman endorses a bold assertion: the worst decision one could make right now is likely "not spending as much money as possible."

This is precisely what OpenAI is doing—engaging in a high-stakes gamble concerning the company's destiny. From chip deals with AMD to a direct expectation for TSMC to "build more capacity," and extends to cryptic collaborations with Samsung and SK hynix in the memory sector, each maneuver is stockpiling ammunition for future exponential growth.

So, who will foot the bill for this trillion-dollar infrastructure investment? Sam Altman's response is unequivocally clear and resolute: ultimately, it will be paid for by OpenAI's own revenue.

His early fundraising experience and investor mindset played a pivotal role here. He is inherently more of an investor than an operator, which grants him a profound understanding of exponential growth, the art of capital allocation, and how to select projects and talent. He even likens OpenAI's operation to managing an internal portfolio of startups, with newly launched Sora being a standout project. This perspective elucidates why OpenAI can decisively decline numerous tantalizing product opportunities, instead concentrating resources intensely on a select few critical endeavors.

Regarding the plan to collaborate with legendary designer Jony Ive on new hardware development, Sam Altman concedes it might divert focus, yet he struggles to conceal his excitement. He believes AI introduces innovation potential to the consumer hardware sector, which he deems "stagnant." As for external speculations about his seemingly audacious series of strategic moves, he offered an enigmatic reply: "Give us a few months, and it will all make sense... We're not as crazy as we appear. There's a plan."

From Product to Platform: Trust as the Sole Currency of the ChatGPT Ecosystem

ChatGPT's ascendance has been nothing short of dramatic. Sam Altman attributes its success to two "surprises": firstly, the product's triumph significantly exceeded expectations; secondly, major tech companies, particularly Google, "misstepped" during the critical period following ChatGPT's launch, affording OpenAI a precious window to establish its supremacy.

Today, ChatGPT is evolving from a phenomenal product into a formidable platform. Within the new "Apps in ChatGPT" feature, OpenAI's strategic choices are pivotal. Sam Altman vehemently disavowed leveraging platform dominance to "dictate" to developers. While they could have pursued a model more advantageous to themselves but detrimental to partners (e.g., direct data scraping), they opted for a path fostering mutual benefit for the entire ecosystem. Users will explicitly invoke third-party applications by name, and partners can assume UI control to forge direct relationships with users. He expressed hope that "new startups could skyrocket as a result."

This deference to partners stems from Sam Altman's steadfast commitment to a core asset: trust.

He underscored that regardless of how future business models evolve—even if they incorporate affiliate marketing or advertising—it will never come at the expense of search result impartiality. "One reason people love ChatGPT is that even if it makes mistakes or hallucinates, you know it's trying to help you, and your incentives are aligned." This trust, he stressed, represents ChatGPT's most invaluable and unshakeable cornerstone.

Sora and the Future of Creation: When AI Lowers the "Activation Energy"

Sora's sudden emergence re-validated OpenAI's capacity to create blockbuster products. Sam Altman posits that Sora's success doesn't hinge on brand halo but rather on its precise capture of a profound human desire: to reduce the "activation energy" for creative expression.

Drawing from his investment experience, he recognized that any technology capable of substantially lowering the barrier to creation unleashes immense potential. Sora accomplishes this, enabling ordinary individuals to readily generate astonishing video content. A striking statistic reveals that during Sora's initial launch, "approximately over 30% of active users were active creators," a phenomenon that utterly disrupts the conventional "90/9/1 rule" in internet content creation (i.e., 90% lurkers, 9% interactors, 1% creators).

Nevertheless, Sora's exorbitant computing costs also herald a new business model. Sam Altman noted that a purely advertising-driven model struggles to sustain all use cases; consequently, "users will have to pay for generations" is highly likely to become the norm moving forward.

Navigating Complex Terrain: Copyright, Feedback, and Humility

OpenAI's viral dissemination frequently entwines with copyrighted content. Yet, Sam Altman repudiated the impetuous philosophy of "move fast and break things." He emphasized that video (e.g., Sora) evokes a fundamentally different sentiment from static images for copyright holders; video's emotional resonance is stronger, thus necessitating greater circumspection.

He revealed that most copyright holders are enthusiastic about integrating their content into AI, merely desiring clearer guidelines. He even ventured a bold prediction: while copyright owners currently fear content misuse, perhaps in less than a year, they will instead complain that "OpenAI isn't putting my content into enough videos," realizing it's an exceptional avenue for forging deeper connections with fans. As AI generates more customized content, the value of resonant public IPs (copyrighted content) will paradoxically increase rather than diminish.

Reflecting on user communication, Sam Altman conceded that they sometimes "oversell" due to excitement and ought to mitigate such instances. Simultaneously, he gleaned a valuable lesson from user feedback regarding ChatGPT-5 (presumably referring to the new UI): AI core users and general audiences (normies) engage with AI in distinctly different ways, necessitating OpenAI to more earnestly construct a single product for these two disparate user groups. He firmly adheres to a principle: "If anecdotes diverge from data, you must exert extra effort to investigate, because anecdotes are often correct."

This dialogue delineates an immensely grand yet logically coherent blueprint for OpenAI. It commences with a unified AGI vision, propels infrastructure buildout through "investor-driven, company-scale bets," cultivates an open platform founded on "trust," and relentlessly explores the future of human-AI collaboration with innovative products like Sora. Sam Altman's agenda is replete with ambition and formidable challenges, yet his final enigmatic and confident pledge perhaps serves as the ultimate annotation for this trillion-dollar venture: "Give us a few months... There's a plan."