The email landed in my inbox: “GPT-2 is too dangerous to release.” I remember thinking, “Too dangerous? What could a language model possibly do?” Fast forward to today, and those early whispers seem almost quaint, overshadowed by the roaring debates about AI ethics and corporate greed. How did a humble non-profit morph into a tech titan obsessed with revenue?
From Nonprofit to “Capped” For-Profit: The OpenAI Origin Story
Back in March 2019, something strange happened. OpenAI, a tech nonprofit that barely registered on the radar, became a “capped” for-profit company—a designation that still feels a little vague. Just a month prior, they’d announced GPT-2, a language model so powerful it was deemed too risky for public consumption. Then, in November, the script flipped, and GPT-2 was released. What changed?
OpenAI stated in their release blog that they’d seen “no strong evidence of misuse so far,” while admitting that total awareness of potential threats was impossible. It’s safe to say that GPT-2 never became a household name, largely because OpenAI didn’t integrate it into any viral chatbots.
I watched this unfold, writing about it at the time. OpenAI seemed insignificant, yet it was developing unsettling AI technology. The company’s public image did a 180: it went from a benevolent computer lab, cautious about causing harm, to an enterprise scrambling to ship products, seemingly promising riches to someone behind the scenes. It was a disturbing narrative to observe.
What impact did Microsoft have on OpenAI?
Now, thanks to document disclosures related to Elon Musk’s lawsuit against OpenAI and Microsoft, we’re getting a glimpse into what was happening within Microsoft during OpenAI’s transformation. These documents shed light on how that evolution might have turned OpenAI into the revenue-focused entity it is today, with revenue increasing tenfold between 2023 and 2025.
Todd Bishop at GeekWire sifted through emails, memos, texts, and other materials from Microsoft and OpenAI. His findings are eye-opening. Microsoft, and CEO Satya Nadella, had invested heavily in OpenAI and weren’t silent during its rocky shift to for-profit status. Nor were they timid about the need to generate income quickly. None of this is shocking, but it’s still fascinating to read.
During the period when GPT-2 was shelved and OpenAI had recently become a capped nonprofit, Microsoft’s CFO, Amy Hood, voiced concerns about the “capped” aspect. In a July 14 email to a group including Nadella, she wrote, “Given the cap is actually larger than 90% of public companies, I am not sure it is terribly constraining nor terribly altruistic but that is Sam’s call on his cap.” This was the first tendril of the beast. It was as if someone was slowly turning up the heat, testing the limits of OpenAI’s altruism.
The Pressure to Monetize: A Shift in Priorities
Consider the release of GPT-3 in 2020, an even bigger deal than GPT-2. Then came Dall-E in January 2021. The following month, Microsoft and OpenAI were discussing further investment. Sam Altman emailed Microsoft, saying, “We want to do everything we can to make you all commercially successful and are happy to move significantly from the term sheet,” adding that he wanted “to make you all a bunch of money as quickly as we can and for you to be enthusiastic about making this additional investment soon.”
When did ChatGPT subscriptions launch?
Fast forward to November 2022, when ChatGPT was unleashed, and the world changed. In January 2023, Nadella texted Altman, “when do you think you will activate your paid subscription for ChatGPT?”
Altman replied that he was “hoping to be ready by end of jan, but we can be flexible beyond that. the only real reason for rushing it is we are just so out of capacity and delivering a bad user experience,” and asked “any preference on when we do it?”
“Let me think about it and weigh in. Overall getting this in place sooner is best,” Nadella responded. Two weeks later, he followed up and asked “how many subs have you guys added to ChatGPT?”
Three days later, the paid version of ChatGPT launched, completing OpenAI’s transformation into a revenue-crazed behemoth.
What are the implications of OpenAI’s focus on revenue?
The implications of all this are massive. The initial promise of AI as a tool for good now seems obscured by the relentless pursuit of profit. The company had become a roaring furnace of innovation, fueled by investment and burning with the need for monetization.
Document disclosures offer a sobering look at the forces shaping the AI landscape. Is this relentless drive for revenue ultimately beneficial for innovation, or does it risk compromising the ethical considerations that should guide AI development?