It’s no secret that artificial intelligence has fundamentally reshaped market dynamics, prompting a frenzied pursuit of transformative potential among both company founders and venture capitalists. Yet, the recent surge in valuations for nascent AI ventures often defies conventional financial logic, presenting what some in Silicon Valley candidly dub “vibe valuing”—an ability to conjure staggering valuations with little regard for traditional spreadsheet metrics. So much so, that it prompted the Economist to publish an article titled “AI valuations are verging on the unhinged“.
Consider the case of Thinking Machines Lab, the new startup helmed by Mira Murati, formerly OpenAI’s chief technologist. This venture, despite being only six months old and having no public product or discernible revenue, recently clinched a $2 billion seed round, catapulting its valuation to an astonishing $10 billion. Murati’s success is largely attributed to her firm’s impressive roster of ex-OpenAI researchers, underscoring the paramount importance of specialised talent in this rapidly evolving landscape. Another example is Ilya Sutzkever’s SSI inc, a one year old startup valued at $32 billion despite of yet having a (public) product or (known) revenue. To put that price in context, that’s approximately the market cap of Salesforce and the Goldman Sachs Group. There’s a question on whether these companies can achieve a real technological breakthrough like AGI, which would obviously be extremely valuable, or are they just pricing in the talent?
This phenomenon extends beyond individual startups. AI startups have experienced massive valuation growth, often securing high multiples even with limited revenue or profitability, driven by advancements in machine learning, automation, and enterprise adoption. In 2024, AI investments represented 42% of all venture capital raised in the US in 2024 reaching a volume of $110 billion globally. In Q1 2025 alone, $59.6 billion was invested in AI companies globally, representing 53% of all VC money invested.
Unlike traditional software companies where worth is primarily tied to revenue and user growth, AI businesses heavily depend on intellectual property, data assets, and algorithmic capabilities – elements that are far more challenging to quantify but are crucial for long-term value. Indeed, AI models often improve with data accumulation, fostering compounding competitive advantages. Investors are now focusing on metrics like algorithm performance, data quality, technical team retention, and scalability potential. But can VCs get that money back? To a large extent, the M&A and IPO market is finally thawing (as of July 2025) and the median revenue multiple for AI companies stood at an impressive 29.7x. Time will tell if buyers (or the public market) is willing to accept these valuations.
The AI talent war is heating up
This intense valuation climate is fuelling an unprecedented talent war. Mark Zuckerberg has been on a relentless quest to recruit top AI engineers and researchers, reportedly dangling $100 million pay packages to some superstars. While not every recruit secures such astronomical figures, the sums are nonetheless “astronomical”. Meta’s aggressive strategy involves a concerted effort to poach from leading research labs, exemplified by their recent $14 billion investment for a stake in Alexandr Wang’s Scale AI, making Wang one of the priciest hires ever. They’ve also successfully lured talent like Lucas Beyer and his colleagues, who left Google DeepMind and OpenAI’s Zurich office to join Meta.
The stakes are high. OpenAI, for its part, has responded to Zuckerberg’s aggressive recruitment with “impressive packages of its own” to retain its scientific breakthroughs and prevent further talent exodus. The “tribal knowledge” held by this small community of elite AI researchers, cultivated through shared experiences and close collaborations, is nearly impossible to replicate, making them incredibly valuable assets. Their motivations extend beyond mere compensation, however, as they seek the limitless computing power, vast data, robust infrastructure, and freedom to experiment that only major tech companies can provide. Compared to the tens of billions major players like Meta, Amazon, Microsoft, and Alphabet are pouring into AI infrastructure annually, humans, even highly compensated ones, appear to be a bargain.
This dynamic of soaring valuations, insatiable demand for elite talent, and a competitive environment so fierce that it fosters “experimental run rates” over stable recurring revenue, points to a market that is still very much in its formative, speculative era. While the potential for AI remains vast, the current investment trends highlight a pivotal “elephant in the room”: are these valuations truly grounded in future profitability, or are we witnessing a collective act of “vibe valuing” that, in time, might give way to cooler heads and a return to more traditional financial assessments? And there are other ‘Elephants’ like AI’s insatiable need for electricity, data and compute, which are in high demand. Time will tell.
I'm a former general partner at google ventures, head of Google for Entrepreneurs in Europe and founding head of Campus London, Google's first physical hub for startups.
I'm also the founder of Techbikers, a non-profit bringing together the startup ecosystem on cycling challenges in support of Room to Read. Since inception in 2012 we've built 11 schools and 50 libraries in the developing world.
- Weekly Firgun Newsletter – February 27, 2026 - February 27, 2026
- The Rise of the Personal AI Agent - February 26, 2026
- Pre-Seed Is Stronger than ever, but less startups are getting funded - February 25, 2026
