What OpenAI And WeWork Have In Common

Ambition is not always a virtue for a start-up.

Adam Neumann, co-founder and first CEO of WeWork, and Sam Altman, co-founder and first CEO of OpenAI¹, share a loose relationship with the truth and an uncommon power of persuasion – a combination that makes them as charismatic as they are dangerous.

Another similarity between them lies in their unquenchable ambition for their companies, driven in part by ego. This ambition pushes them toward ever more reckless – because ever more brazen – diversifications.

Adam Neumann thus anchored WeWork in an expanding array of increasingly presumptuous, even fanciful, ventures beyond coworking spaces, which ultimately contributed to its downfall:

• WeLive, co-living residential spaces offering shared apartments and communal amenities.
• WeWork Labs, a startup incubation and acceleration program.
• Meetup, a social network acquired by WeWork to supposedly integrate its online community into WeWork’s physical spaces.
• Rise by We, high-end fitness clubs and wellness centers.
• WeGrow, a private school aimed at providing children with an education centered on entrepreneurship and “consciousness.”
• Wavegarden, surf parks with artificial waves².

Credit where it is due to Sam Altman: The student has surpassed the master – maybe thanks to his experience running the Y Combinator accelerator, which may have sharpened his curiosity across a wide range of technology sectors.

At the helm of OpenAI, he is pursuing a scattershot growth strategy.

The company must:

• Build data centers.
• Design its own chip.
• Invent a physical device capable of competing with the iPhone.
• Develop software and hardware for robots.
• Turn ChatGPT into a shopping tool, or even transform it into a “super app.”
• Create a web browser and a search engine.
• And invest in the private-equity firm Thrive Holdings.

The latest example of Sam Altman’s approach involves Stoke Space, a rocket company that he reportedly considered having OpenAI acquire.

This diversification strategy goes far beyond vertical integration pushed to its extreme – namely embedding GPT large language models and the ChatGPT assistant throughout their respective value chains. It might be treated more gently if OpenAI had the human and financial resources of Google or Microsoft. But the reality is that the start-up must already find a way to finance the $1.4 trillion in commitments it has made to build the compute capacity required to run ChatGPT.

Even more irresponsibly, Sam Altman proceeds with apparent indifference to revenue generation, as illustrated by OpenAI’s culpable delay in developing an advertising offering – despite the fact that, unlike Anthropic, ChatGPT’s natural habitat is the consumer market, not the enterprise one. While Google funds the development of Gemini, Nano Banana, and Veo – among others – with its advertising cash cow, OpenAI is instead building on speculative quicksand.

In this regard, the least paradoxical aspect of the red alert triggered by Sam Altman at OpenAI a few days ago, following Google’s advances in both technology and popularity, is not that the Company’s advertising roadmap has been postponed in favor of focusing on AI model improvements. This amounts to an admission that, at the very core of its business – what should be its absolute top priority -, Sam Altman has allowed OpenAI to fall behind Google, its most dangerous competitor. This is all the more damaging because Google enjoys a unique combination of AI expertise, infrastructure and data to monetize that expertise, distribution channels for related applications, and the financial means to subsidize its strategy in this domain.

In the end, OpenAI now appears to be losing the lead it once had over Google – but also, more tellingly, over Anthropic. The creator of Claude, co-founded and led by Dario Amodei, is far more disciplined than OpenAI and has so far played its hand in the corporate market³ with much greater pragmatism. Anthropic expects to become profitable by 2027, generating $3 billion in cash flow that year and up to $17 billion the following year.

By contrast, under Sam Altman’s leadership, OpenAI seems to be playing at being Google, complete with experimental “other bets,” without first having generated the revenues required to fund them.

When it comes to the gap between reality and fantasy, Adam Neumann turns out to be an amateur compared to Sam Altman.

—

¹ Sam Altman became OpenAI’s first CEO in 2019. Between 2015 and 2019, OpenAI was a nonprofit organization and had no CEO; Sam Altman and Elon Musk then served as co-chairs of its Board.

² This list may be incomplete if my research missed other WeWork diversification efforts under Adam Neumann.

³ A segment in which Anthropic does not have to subsidize a large base of non-paying users, unlike OpenAI with ChatGPT.