Leaked Documents Reveal Microsoft’s Influence as OpenAI Became a Revenue-Crazy Behemoth
Back in March of 2019, a strange thing happened when an obscure non-profit technology company called OpenAI became a for-profit “final” company—whatever that is. Last month, OpenAI had announced the development of an abstract, powerful language model called GPT-2 that was deemed too risky to release. Then in November, OpenAI seemed to change its mind and GPT-2 was finally released.
OpenAI said in a blog post about the release that it had seen, “there is no strong evidence of misuse so far,” but added that it is impossible to “know all threats.” Most people have never used GPT-2, because OpenAI has never injected it into an infected chatbot.
As someone who wrote about this at the time, it was confusing to watch it all play out. OpenAI looked like small potatoes, but it was also building awesome AI technology, and it was changing the public’s image from being a run-of-the-mill computer lab that advertised its horrors to the detriment of anyone’s hair in a business that needed to ship something quickly because it was clearly promising someone, somewhere, that they were going to get rich.
The discovery of a document in Elon Musk’s lawsuit against OpenAI and Microsoft gave a small window into what was actually happening inside Microsoft during this strange time for this strange company, and how the revolution could turn OpenAI into the money-hungry monster it is today, with revenues growing tenfold between 2023 and 2025.
GeekWire’s Todd Bishop dug through a trove of emails, memos, documents and the like from Microsoft and OpenAI, and what he found was revealing. Microsoft, and CEO Satya Nadella in particular, had invested heavily in OpenAI at the time, and were not silent on OpenAI’s transition to a for-profit status. And they weren’t shy about the need to make money as quickly as possible. Of course none of this should surprise us, but it makes for interesting reading nonetheless.
During that gap where GPT-2 sat unreleased and OpenAI had just become a defunct non-profit organization, Microsoft’s chief financial officer, Amy Hood, weighed in on the company’s concerns about that “integrated” component. He wrote in a July 14 email to a group that includes Nadella, “Since the cap is actually larger than 90% of public companies, I’m not sure if it’s too oppressive or too unkind but that’s Sam’s call in his head.”
GPT-3, which was much more exciting than GPT-2 was released in 2020, and the first version of OpenAI’s language model, Dall-E was released in January 2021. The following month, Microsoft and OpenAI were negotiating for an additional injection of money from Microsoft, and Sam Altman wrote an email to Microsoft, saying “We want to do everything we can to make it more successful on the sheet and more successful” We are making a lot of money for you as quickly as possible and you will be motivated to make this additional investment soon.”
In November of 2022, ChatGPT was released, and as you know, all hell broke loose. In January of 2023, Nadella sent a text message to Altman, saying “when do you think you’ll use up your paid ChatGPT subscription?”
Altman said he “hopes we’ll be ready by the end of jan, but we can be more flexible than that. The real reason for the rush is that we’re out of power and delivering a bad user experience,” and asked “any choice as to when we do it?”
“Let me think about it and think about it. All in all getting this sooner is better,” Nadella replied. Two weeks later, he followed up and asked “how many subscribers have you added to ChatGPT?”
Three days later, the paid version of ChatGPT was launched.
