I’ve heard rumors that OpenAI is moving closer to a fully for-profit model, away from its original non-profit structure. A lot of people dislike this direction. I understand why, but I have a potentially unpopular opinion: I support the move.

Sustainability and the bitter lesson

The most obvious reason is that OpenAI needs to sustain itself. Their initial shift to a capped-profit model came from the realization that scaling matters enormously in deep learning, a point often called the “bitter lesson.” In modern machine learning, bigger models and more compute keep winning.

That kind of scaling takes a lot of capital. As a non-profit, OpenAI would have to rely more heavily on donors, and donor incentives can also compromise independence.

Generating their own revenue gives them more autonomy and a more sustainable way to fund research. At least from OpenAI’s perspective, that part seems straightforward.

Alignment with OpenAI’s charter

I also think this move fits OpenAI’s long-term charter better than it might seem. Their stated goal is to benefit humanity by creating Artificial General Intelligence (AGI). That is a bold ambition, but such a system won’t emerge in a vacuum.

Progress toward AGI will probably look more like a gradual slope than a sudden quantum leap. In the meantime, if OpenAI develops useful systems, they should make them available, even if that means selling them.

The current language models and voice models offered by OpenAI through subscriptions are not just gimmicks. They provide real value. Charging users who benefit from them is a reasonable way to turn that value into capital for more research.

The openness debate

There is also the criticism that OpenAI isn’t truly “open.” I agree with that criticism, but I think they have limited choices. I don’t buy the idea that they withhold models only for safety reasons. I think competitiveness is a major reason too.

As mentioned earlier, they need capital to fund research. Open-sourcing their models without restrictions would weaken their competitive position and make it easier for others to catch up. In my view, that is the main reason behind their closed-source approach.

Uncertainties

Of course, the worry is real. A for-profit company’s primary objective is to generate revenue and increase capital. Whether OpenAI can keep its original mission under that structure remains to be seen.

Still, from OpenAI’s perspective, I think this is a logical step. It gives them a way to fund research and keep building systems people actually use.