All eight Transformer authors have now walked out of Google
Llion Jones, the last remaining author of the groundbreaking 2017 Transformer paper 'Attention is All You Need', just confirmed he is leaving Google to start his own company. All eight researchers behind the architecture powering ChatGPT and Midjourney have departed, launching startups like Cohere and Character.ai now valued at over $4.1 billion combined.

The "T" in ChatGPT stands for Transformer — and the tech was actually dreamed up over lunch at Google's California cafeteria back in 2017.
Now, the very last author of the landmark paper behind it has walked out the door.
Llion Jones confirmed he is leaving Google to start his own company. With his departure, all eight co-authors of "Attention is All You Need" — the paper cited more than 80,000 times that laid the groundwork for ChatGPT, Midjourney, and today's generative AI boom — have officially left the search giant.
A $4.1 billion diaspora
Google gave the world the Transformer, but its creators are now cashing in on the outside.
Startups founded by the paper's eight authors are now collectively valued at around $4.1 billion, according to data from PitchBook and CoinMarketCap. The alumni roster includes AI heavyweights like Noam Shazeer, who launched Character.ai, as well as the founders behind enterprise AI platforms Cohere and Essential.ai, and biotech startup Inceptive Nucleics.
Jones admitted that the sudden recognition over the past year feels surreal. While nobody recognizes his face, he noted that it takes just five seconds to explain his claim to fame: he helped build the "T" in ChatGPT.
The curse of being too big
Why did Google let the biggest breakthrough in modern AI slip through its fingers and fuel rival OpenAI?
Insiders point directly to corporate bloat. Earlier this year, Glass.ai estimated that Google employed 7,133 AI workers out of its 140,000-strong workforce. OpenAI sparked an industry-wide arms race with just around 150 AI researchers out of roughly 375 total employees in 2023.
Former researchers revealed that getting projects greenlit at Google Brain meant navigating thick layers of bureaucracy. Many scientists became more focused on academic paper visibility and career ladders than shipping tools.
Co-author Illia Polosukhin put it bluntly: Google doesn't move unless an idea looks like a billion-dollar business. But discovering those hits requires tolerating endless dead ends and rapid iterations — something a comfortable tech giant simply couldn't stomach when its existing systems weren't broken.
Why it matters
We think the total departure of the Transformer team perfectly illustrates the classic innovator's dilemma. Google had the talent, the compute, and the early technical lead, yet structural friction and a risk-averse culture kept its own invention idling on the shelf until OpenAI ate its lunch.
For the broader tech ecosystem, however, Google's loss has been an absolute windfall. Instead of keeping foundational AI expertise locked behind mountain-high corporate approvals in Mountain View, these eight researchers have seeded over $4 billion worth of fresh competition across enterprise software, biotech, and consumer bots. It turns out that inventing the future is only half the battle; having the organizational freedom to actually ship it is what changes the world.
Sources (2)
- bloomberg.com(opens in a new tab)Originalbloomberg.com
- gadgets360.com(opens in a new tab)Reportgadgets360.com