The 8 Authors of "Attention Is All You Need" (Who Left Google)

In 2017, eight researchers at Google published a paper about translating languages. They gave it a slightly cheeky title borrowed from a Beatles song — “Attention Is All You Need” — and almost none of them understood, at the time, that they had just written the blueprint for ChatGPT, Claude, Gemini, and the entire generative-AI boom that followed. The story of the “Attention Is All You Need” authors is one of the most remarkable in modern technology: eight people who lit a fuse they didn’t know they were lighting, and who then, one by one, walked out of Google to build the industry their own paper made possible.

This is the human side of the paper behind ChatGPT. If you want the plain-English explanation of what “attention” actually is, that companion piece has you covered. Here, we’re interested in the people — who they were, what they made, and why the most valuable research team in tech history no longer works at the company that published them.

The eight authors of Attention Is All You Need, represented as eight connected points of light around the paper that founded modern AI

📄 One Paper, Eight Equal Names

The first thing to know about “Attention Is All You Need” is something most write-ups get wrong: it does not have a lead author in the usual sense. The paper carries an unusual footnote declaring all eight authors equal contributors, and even notes that the ordering was randomized. That was a deliberate rejection of the academic convention where the first name gets the glory.

So the honest way to name them is together: Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan Gomez, Łukasz Kaiser, and Illia Polosukhin. Each contributed a different, essential piece — some the core ideas, some the relentless experiments, some the code that made it all run. When you read “Google invented the Transformer,” mentally replace it with these eight names. The company provided the lab; the people provided the breakthrough.

It’s worth pausing on how ordinary the setting was. This was not a moonshot with a mandate to reinvent AI. It grew out of practical efforts to improve Google Translate. Different team members had been circling the same intuition — that a mechanism called attention was doing most of the useful work, and maybe the slow, sequential machinery everyone else relied on could simply be thrown away. They bet it could. They were right in a way that reshaped the decade.

🎸 The Title Was a Beatles Joke

Here’s a detail that humanizes the whole thing. A couple of nights before the submission deadline, the team still didn’t have a title. Llion Jones — one of the eight — noted that the paper’s whole thrust was a radical rejection of the accepted best practice, in favor of one technique: attention. The Beatles, he recalled, had a song called “All You Need Is Love.” So why not “Attention Is All You Need”?

By his own account it took about five seconds, and he didn’t expect the team to actually use it. They did. And the title became so iconic that it spawned an entire genre of imitators — you can now find hundreds of AI papers titled “X Is All You Need.”

There’s a teaching point hiding in the joke, and it’s worth keeping straight: the title means “you no longer need the old sequential machinery — attention alone is enough.” It does not mean attention is literally the only ingredient. The architecture also relies on several other essential parts to work at all. The catchy title oversells one idea, which is exactly why so many people misunderstand the paper. But as a piece of scientific branding, it’s unmatched.

Eight points of light dispersing from a single campus toward separate destinations — the Transformer authors leaving Google to found the AI industry

🚪 The Great Exodus: Where the Eight Went

Now the part that makes this a genuinely astonishing story. Over the years that followed, every single one of the eight authors left Google — and collectively they went on to found or lead a startling share of the modern AI industry. The people who wrote the paper that empowered Google’s competitors largely left to become those competitors.

Here is where they landed (roles as of 2026 — this world moves fast):

Aidan Gomez was the youngest of the group, an intern at the time of the paper. He co-founded Cohere, a company building large language models for enterprises, and serves as its CEO.
Noam Shazeer, regarded as one of the most brilliant engineers of the group, co-founded Character.AI, a wildly popular chatbot company. In 2024 he returned to Google in a high-profile deal and became a technical lead on Gemini, Google’s flagship AI.
Ashish Vaswani and Niki Parmar left together, co-founded the startup Adept AI, then departed to found Essential AI, which reached a billion-dollar valuation. Parmar has since moved on again, to a research role at Anthropic (the maker of Claude).
Jakob Uszkoreit took the most unexpected turn, co-founding Inceptive — a company applying deep learning not to language but to biology, designing RNA molecules for new medicines.
Llion Jones — the one who named the paper — co-founded Sakana AI, a research lab based in Tokyo.
Łukasz Kaiser joined OpenAI, where he became a foundational contributor to the “reasoning” models (the o1/o3 line) behind some of the most advanced AI systems available.
Illia Polosukhin left the AI mainstream entirely to co-found NEAR Protocol, a blockchain platform.

Read that list again and the scale of it sinks in. Cohere, Character.AI, Essential AI, Sakana, Inceptive, plus key people now steering Gemini, Claude, and OpenAI’s reasoning models — a huge slice of the current AI landscape traces back to eight desks at one Google office in 2017. It is one of the great talent diasporas in the history of technology.

🤔 Did They Know What They Had?

The temptation is to tell this as a story of eight visionaries who knew they were birthing the AI age. By most accounts, that’s not what happened — and the truth is more interesting.

They knew the paper was good. It won its way to a top conference and beat the state of the art in translation. But the leap from “a better translation model” to “the foundation of a technology that would write essays, code, and hold conversations” was not obvious in 2017, not even to the people who built it. That transformation took years of additional work by other researchers — new training methods, staggering increases in scale, and a completely different goal (generating open-ended text rather than translating sentence pairs). The Transformer was the seed. It took a lot more to grow the forest.

That gap between what they made and what it became is precisely why the story resonates. These weren’t prophets. They were skilled researchers who solved a concrete problem elegantly, and the elegance turned out to generalize far beyond the problem. It’s a reminder that some of the most consequential technology arrives sideways — as a translation paper, from a team that gave it a Beatles title and moved on.

🌍 Why This Story Matters

It’s easy to think of AI as the product of a few famous companies or a couple of celebrity founders. The reality this story reveals is more human and more distributed. A breakthrough came from a team — eight equal contributors, each essential — working on an unglamorous problem, and its impact scattered outward as those people scattered.

That pattern echoes the other origin stories in this series. The deep learning revolution that preceded it ran on the same fuel: a small number of stubborn researchers, an idea most of the field doubted, and years of patient work before the world caught up. The people who built the ImageNet dataset and won the 2012 contest that kicked off modern computer vision were, likewise, a handful of humans betting on an unfashionable idea.

The next time you use an AI assistant, it’s worth remembering that behind it isn’t a faceless corporation or a single genius. It’s a lineage of named people who did specific, creditable work — starting, in the language case, with eight researchers, a translation problem, and a title borrowed from the Beatles.

❓ Frequently Asked Questions

Who wrote “Attention Is All You Need”? Eight researchers at Google, credited as equal contributors: Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan Gomez, Łukasz Kaiser, and Illia Polosukhin. The paper deliberately has no single “lead” author — the ordering was randomized.

Why is it called “Attention Is All You Need”? The title, suggested by author Llion Jones, is a nod to the Beatles’ “All You Need Is Love.” It captures the paper’s central move: dropping the older sequential machinery and relying on the attention mechanism alone. It does not mean attention is literally the only component of the model.

Did all the authors really leave Google? Yes. Over the years following the paper, all eight left Google to found or join other companies — including Cohere, Character.AI, Essential AI, Sakana AI, Inceptive, OpenAI, and NEAR Protocol. (Noam Shazeer later returned to Google in 2024 to work on Gemini.)

Did the authors know they were inventing the technology behind ChatGPT? Not really. They knew the paper was a strong result in machine translation, but the path from that 2017 translation model to conversational AI like ChatGPT took years of additional research, new training methods, and enormous scaling done largely by others. ChatGPT launched in late 2022, five years later.

What companies did the Transformer authors found? Among them: Cohere (Aidan Gomez), Character.AI (Noam Shazeer), Essential AI and earlier Adept AI (Ashish Vaswani and Niki Parmar), Sakana AI (Llion Jones), Inceptive (Jakob Uszkoreit), and NEAR Protocol (Illia Polosukhin). Łukasz Kaiser joined OpenAI.

📚 Further Reading

What Is “Attention” in AI? The Paper Behind ChatGPT — the plain-English explanation of what these eight people actually built, and why it matters.
Geoffrey Hinton and the Deep Learning Revolution — the researcher whose decades of persistence set the stage for this moment.
The AlexNet Moment and Fei-Fei Li and ImageNet — the twin origin stories on the vision side of AI, told with the same eye for the people behind the breakthroughs.

📚 Recommended Resources

Disclosure: some links below are Amazon affiliate links. If you buy through them, I earn a small commission — at no extra cost to you — which helps keep this site running.