It turns out that today’s AI landscape can be traced back to—what do you know—fear, jealousy, and intense capitalist ambition. First emails revealed in Justice Department antitrust case against Google reported from Business Insider, show Microsoft executives expressing alarm and envy over Google’s leadership in AI. That created an urgency that led to the Windows maker’s initial billion-dollar investment in its now-indispensable partner, OpenAI.

In a heavily redacted 2019 email thread titled “Thoughts on OpenAI,” Microsoft CEO Satya Nadella forwarded a lengthy message from CTO Kevin Scott to CFO Amy Hood. “Very good email explaining why I want us to do this … and also why we will then ensure our infra staff perform,” Nadella wrote.

Scott wrote that he was “very, very worried” about Google’s rapidly growing AI capabilities. He says he initially dismissed the company’s “gaming stunts,” possibly referring to Google’s AlphaGo models. One of them beat the 2017 Go Ke Jie World Champion, a remarkable achievement at the time. (Later Google models surpassed this one, eliminating the need for human training entirely.)

But Scott says dismissing Google’s progress in the game “is a mistake.” “When they took all the infrastructure they had built to build [natural language] patterns that we could not easily reproduce, I began to take things more seriously,” Scott wrote. “And as I dug in to try to figure out where all the capability gaps were between Google and us for model training, I was very, very worried.”

Microsoft CTO Kevin Scott presents on stage in front of a blue wall with the Microsoft logo on it.  Blurred heads of audience in foreground.

Microsoft CTO Kevin Scott (Microsoft)

Scott recounts how Microsoft struggled to copy Google’s BERT-large, an AI model that deciphers the meaning and context of words in a sentence. Scott blamed the infrastructure leaps his rival had made — and Microsoft hadn’t.

“It turned out that simply copying BERT-large was not easy for us. Although we had the template for the model, it took us ~6 months to train the model because our infrastructure was not up to the task,” the Microsoft CTO wrote. “Google had BERT at least six months before that, so in the time it took us to hack together the ability to train a model with 340M parameters, they had a year to figure out how to get it into production and move forward to larger, more interesting models.”

He also admired and envied Google’s Gmail autocomplete capabilities, saying it was “getting pretty good.” He commented that Microsoft is “many years behind the competition in terms of [machine learning] scale.” He commented on the “interesting” growth of OpenAI, DeepMind and Google Brain.

Scott touted Microsoft’s “very smart” people on the machine learning teams, but said their ambitions were limited. “But the core deep learning teams in each of these larger teams are very small and their ambitions are also limited, which means that even when we start feeding them resources, they still have to go through a training process to to expand,” Scott wrote. “And we’re several years behind the competition in terms of the scale of ML.”

After prompting Hood that Scott’s concerns were “why do I want to do this,” meaning investing in OpenAI, the company granted its CEO’s wishes. Microsoft invested $1 billion in the Sam Altman-led startup in 2019, and the rest is rapidly changing history. ($13 billion is now invested.) It’s a technology that’s doing some amazing things, but it threatens to gut the job market and give propagandists their most powerful tools yet in what was already an age of rampant disinformation.