In what may be the most impressive magic trick since David Copperfield made the Statue of Liberty disappear many moons ago, former OpenAI CTO Mira Murati has convinced investors to pour $2 billion into her three-month-old startup, Thinking Machines Lab, valuing the company at a modest $10 billion. The remarkable achievement comes despite the company having no product, no revenue, and approximately 30 employees who appear to spend their days crafting exquisitely vague mission statements about “making AI systems more widely understood, customizable, and generally capable.”1
Industry analysts are calling it the most efficient capital-to-buzzword ratio in Silicon Valley history, with each promised feature of the non-existent product apparently worth exactly $1 billion in valuation. The fundraising round, reportedly led by Andreessen Horowitz, requires investors to commit a minimum of $50 million to participate – roughly the GDP of several African nations or approximately what OpenAI spends on electricity every three days.2
The World’s Most Expensive PowerPoint Deck
According to sources familiar with the pitch deck, Thinking Machines Lab has perfected the art of raising venture capital by combining three essential elements: impressive-sounding ex-OpenAI employees, the promise of ethical AI development, and absolutely no specific details about what they’re actually building.3
“The genius of Thinking Machines Lab’s pitch is its perfect algorithmic balance of buzzwords to substance,” explains venture capital psychologist Dr. Samantha Chen. “They’ve calibrated their language to trigger the maximum FOMO response in the VC investor silicon brain. Phrases like ‘collaborate with humans’ and ‘open science’ activate the prefrontal cortex’s ‘give them money immediately’ center, while deliberately vague promises about ‘novel scientific discoveries’ stimulate the amygdala’s ‘fear of missing out on the next OpenAI’ response.”
When pressed about what differentiates Thinking Machines Lab from existing AI companies, Murati has reportedly told investors the company is focused on “multimodal systems that work with people collaboratively,” a revolutionary approach that sounds suspiciously like what every other AI company on Earth is also claiming to do.4
The $50 Million Minimum Entry Fee: Because Exclusivity Sells
Perhaps the most brilliant aspect of Thinking Machines’ fundraising strategy is the $50 million minimum investment requirement – a sum so large it automatically filters out any investors who might ask uncomfortable questions like “What exactly are you building?” or “How is this different from ChatGPT?”5
“The $50 million minimum is actually a psychological masterstroke,” explains Dr. Chen. “It creates an artificial barrier to entry that makes getting into the deal feel like joining an exclusive club. VCs who can afford it will pay simply for the bragging rights of saying they are in the round. It’s the same principle as Veblen goods – the higher the price, the more desirable it becomes, regardless of actual utility.”
This approach has created a feeding frenzy among investors, with one anonymous VC partner reportedly selling his children’s private school to free up liquidity for the round. “My kids can learn on YouTube,” he explained. “But this Thinking Machines opportunity only comes once in a lifetime. Or at least once every 18 months when a new AI startup with former OpenAI employees launches.”
The Murati Magic: Turning Nothing into Billions
Murati has assembled an impressive team of AI researchers, including OpenAI co-founder John Schulman as chief scientist and former OpenAI leader Barrett Zoph as CTO. The team also counts Bob McGrew, previously OpenAI’s chief research officer, and Alec Radford, a former OpenAI researcher behind many transformative innovations, as advisers.6
This collection of talent has led many to wonder if the company’s true product is simply the team itself – a sort of reverse acqui-hire where investors pay billions for the privilege of eventually being acquired by Google or Microsoft.
“It’s brilliant when you think about it,” says tech industry analyst Michael Wong. “Traditional startups have to build a product, find product-market fit, scale, and then exit. Thinking Machines has created a shortcut where the exit is built into the company’s DNA from day one. It’s like Schrödinger’s startup-simultaneously a research lab, a product company, and an acquisition target, all without ever having to build anything specific.”
The Future Promises Machine
While Thinking Machines Lab’s website and public statements remain frustratingly vague, sources close to the company suggest its technology will revolutionize AI by making it “do all the things current AI does, but somehow better.”7
“We’re not just building another AI model,” Murati allegedly told investors in a closed-door session. “We’re building a thinking machine that truly understands humans, adapts to their needs, and can generate PowerPoint decks convincing enough to raise $2 billion on a $10 billion valuation with no product.”
The company’s promotional materials emphasize that unlike existing AI systems which excel primarily at programming and mathematics, Thinking Machines Lab is developing AI that can “adapt to the full spectrum of human expertise.”8 When asked what this means in practice, a spokesperson reportedly waved their hands in the air while making whooshing sounds.
The Customization Revolution: One Size Fits All, But Make It Personal
The cornerstone of Thinking Machines Lab’s pitch appears to be “customizable AI,” a revolutionary concept that somehow differs from prompt engineering, fine-tuning, RLHF, and all the other customization approaches already available in existing AI systems.9
“Current AI systems force users to interact on the AI’s terms,” explains an industry consultant who has seen Thinking Machines’ pitch. “Thinking Machines is creating AI that interacts on the users’ terms, a subtle but important distinction that absolutely justifies a $10 billion valuation despite sounding exactly like what every other AI company is trying to do.”
To achieve this groundbreaking customization, Thinking Machines is reportedly developing a technology called “personal preference neural mapping,” which sounds impressive until you realize it’s essentially just remembering what users like – something cookies have done since 1994.
The Ethics Arbitrage: Open Science, Closed Wallet
Perhaps the most ingenious aspect of Thinking Machines’ strategy is its emphasis on ethics, transparency, and open science – values that somehow don’t extend to explaining to the public what they’re actually building with $2 billion of investor money.
“Science is better when shared,” proclaims the company’s website, right before not sharing any actual science. This carefully calibrated ethical posturing allows Thinking Machines to position itself as the “good” AI company without the inconvenience of specific ethical commitments that might limit its business options.
“It’s what I call ethics arbitrage,” explains Dr. Chen. “By appearing more ethical than your competitors, you create the impression of moral superiority without the burden of actual ethical constraints. It’s like putting ‘all natural’ on a product label-it sounds good but doesn’t actually mean anything specific.”
The $10 Billion Question: What Makes This Worth $10 Billion?
When evaluating Thinking Machines’ $10 billion valuation, it’s worth comparing to other AI companies with actual products. ChatGPT reached 100 million users in two months. Claude has established itself as a thoughtful alternative. Google’s Gemini, despite a rocky start, has the backing of one of the world’s largest Monopoly. DeepSeek has demonstrated impressive capabilities.
What does Thinking Machines offer to justify a $10 billion valuation before launching a product? According to market analysis, the answer appears to be “AI vibes.”
“Valuing pre-product startups is more art than science,” explains financial analyst David Peterson. “And by ‘art,’ I mean it’s complete fiction. The $10 billion figure wasn’t derived from discounted cash flows or comparable company analysis – it’s what economists technically call a ‘made-up number’ that seemed large enough to generate headlines but not so large that people would openly laugh.”
The Thinking Machines Lab Paradox: The Less Specific, The More Valuable
In perhaps the most remarkable feat of modern venture capitalism, Thinking Machines has discovered that valuation is inversely proportional to specificity. The less they say about what they’re building, the more investors value the company.
“If they came out and said, ‘We’re building a chatbot that’s 10% better than ChatGPT,’ they’d be worth maybe $1 billion,” explains Peterson. “But by saying they’re creating AI that’s ‘more widely understood, customizable and generally capable,’ they’ve created a blank canvas onto which investors can project their wildest AI fantasies. It’s genius – the Rorschach test approach to company valuation.”
This approach has allowed Thinking Machines to avoid the pitfalls that come with specific promises. By not claiming they’ll build AGI by a certain date, create a chatbot that can pass the bar exam, or generate images of dragons wearing sunglasses, they can’t fail to deliver on those promises.
The AI Arms Race: Minimum Viable Hype
As the AI arms race heats up, new competitors are emerging with increasingly astronomical valuations and decreasingly specific products. Ilya Sutskever’s Safe Superintelligence startup is reportedly seeking similar funding levels, creating what analysts call “the great AI nothing race.”10
“We’re witnessing an evolution in startup strategy,” explains industry observer Sarah Johnson. “Traditional startups had to build a minimum viable product. AI startups now only need to create minimum viable hype. The product is almost an afterthought – a distraction from the real business of raising money at increasingly absurd valuations.”
This shift represents the final decoupling of startup valuation from traditional metrics like revenue, profit, or even user numbers. In the new paradigm, valuation is determined by a complex algorithm that factors in the prestige of former employers, the number of Stanford PhDs on staff, and how many times the pitch deck mentions “collaborative intelligence” and “customizable systems.”
Conclusion: The Emperor’s New Neural Network
As Thinking Machines Lab continues its fundraising journey, the question remains: will they deliver revolutionary AI that justifies its astronomical valuation, or is this simply the latest example of Silicon Valley’s reality distortion field?
“In the short term, it doesn’t matter,” concludes Dr. Chen. “They’ve already won by raising $2 billion at a $10 billion valuation. Success in modern tech isn’t measured by building useful products – it’s measured by convincing smart people to give you billions of dollars based on PowerPoint slides and the promise of future magic.”
Meanwhile, as Thinking Machines prepares to cash its $2 billion check, engineers at the company are reportedly working around the clock to develop something – anything – that can be demonstrated to investors before they ask for actual results.
“The pressure is enormous,” confides an anonymous source close to the company. “They essentially need to build a $10 billion AI assistant in 18 months before investors realize they could have just used ChatGPT Plus for $20 a month.”
Do you think Thinking Machines Lab can actually deliver something revolutionary, or is this just another example of AI hype gone wild? Would you invest $50 million in a company with no product based solely on the team’s pedigree? Is the AI funding bubble about to burst, or are we just getting started? Share your thoughts in the comments below – unless you’re a Thinking Machines investor, in which case, we apologize for making you question your life choices.
If you enjoyed this analysis of how to turn PowerPoint slides into billions of dollars, consider donating to TechOnion. For just a fraction of the minimum Thinking Machines investment (we'll accept anything over $12.50), you can support our ongoing mission to point out the emperor's lack of clothes in the tech industry. Remember, unlike AI startups, we actually have a product-it's this article you just read.
References
- https://thinkingmachines.ai/ ↩︎
- https://www.insider-inc.com/mira-murati-is-asking-investors-to-commit-to-at-least-50-million-2025-5 ↩︎
- https://opentools.ai/news/former-openai-cto-launches-revolutionary-ai-startup-thinking-machines-lab ↩︎
- https://pylessons.com/news/mira-murati-launch-thinking-machines-lab-ai-innovation ↩︎
- https://www.insider-inc.com/mira-murati-is-asking-investors-to-commit-to-at-least-50-million-2025-5 ↩︎
- https://siliconangle.com/2025/04/10/mira-muratis-thinking-machines-reportedly-raising-2b-funding/ ↩︎
- https://www.theinformation.com/articles/thinking-machines-lab-ceo-unusual-control-andreessen-led-deal ↩︎
- https://techcrunch.com/2025/02/18/thinking-machines-lab-is-ex-openai-cto-mira-muratis-new-startup/ ↩︎
- https://elearncollege.com/technology/mira-murati-launches-thinking-machines-lab-initiative/ ↩︎
- https://newsletter.angularventures.com/p/solo-founder-syndrome-even-if-you-re-not-alone ↩︎
