in

Desperate Google Paid $2.7 Billion to Get a Single AI Researcher Back


“Noam is clearly a great person in that space. Is he 20 times as good as other people?”

Talented Mr. Shazeer

Google reportedly spent $2.7 billion to rehire AI expert Noam Shazeer — an astonishing amount of money for what boils down to the expertise of a single computer scientist.

Shazeer left Google in 2021 after 21 years at the company to start his own startup called Character.AI, which allows users to interact with chatbots that roleplay as imaginary characters or well-known figures.

The official reason Google gave billions of dollars to the startup was to license its tech. But as the Wall Street Journal reports, “Shazeer’s return is widely viewed as the primary reason” as to why the tech giant cut the check.

The deal highlights just how much money tech companies are spending to get a potential edge in the AI arms race. In parallel, Silicon Valley investors have grown very concerned as of late that companies like Google or Microsoft are throwing away countless billions of dollars with no return in sight, inflating a massive bubble that is set to burst.

Look no further than a company spending nearly three billion on getting a single engineer to come back.

“Noam is clearly a great person in that space,” Stanford Artificial Intelligence Laboratory director Christopher Manning told the WSJ. But is he “20 times as good as other people?”

Window Shopping

While Shazeer wasn’t the sole or direct recipient of Google’s $2.7 billion, he raked in hundreds of millions of dollars from his stake in the startup as part of the deal, sources told the WSJ. That’s still a massive paycheck, especially considering Shazeer didn’t sell his company.

Shazeer was a foundational part of Google’s early efforts to develop AI. A 2017 paper he coauthored is largely seen as the early dawn of the large language models as we know them today.

And he isn’t shy about his contributions. On his LinkedIn, he claims that “I have invented much of the current revolution in large language models.”

The computer scientist left Google after the company refused to release a chatbot dubbed Meena to the public in 2021, citing safety concerns, per the WSJ. Ironically, OpenAI released ChatGPT one year later, highlighting a massive wasted opportunity for Google.

In August, news emerged that Shazeer would return to Google as part of Google’s parent company Alphabet’s deal with Character.AI.

Several other similar deals show that tech giants like Microsoft and Amazon are racing to secure AI researchers like Shazeer. In June, for instance, Amazon hired top executives from an AI startup called Adept through a technology licensing deal.

As companies continue to pour billions into the environmentally damaging and still largely unproven tech, though, investors are starting to ask some tough questions. Should they really be paying several billion dollars to get the expertise of a single hire — or is it a colossal waste of money that won’t ever see a return? All we can do is watch the situation play out.

More on Google: AI Datacenters More Than 600 Percent Worse for Environment Than Tech Companies Claimed

The Most Sophisticated AIs Are Most Likely to Lie, Worrying Research Finds

The Most Sophisticated AIs Are Most Likely to Lie, Worrying Research Finds

An ‘iPhone of AI’ Makes No Sense. What Is Jony Ive Really Building?

An ‘iPhone of AI’ Makes No Sense. What Is Jony Ive Really Building?