Tech company OpenAI changes course, chief technology officer resigns

The technology company OpenAI plans to review its corporate structure. Chatbot maker ChatGPT wants to permanently convert from a foundation to a for-profit company, Reuters news agency reports.

Around the same time that this news broke, the director of technology OpenAI, Mira Murati announced known on X to leave the artificial intelligence company. She says she is doing this to “make time and space for my own research”.

Murati worked at OpenAI for 6.5 years, including a brief stint as CEO. That happened last November, when CEO Sam Altman was fired by the board of directors. It was one of many developments at the time several weeks of crisis within the company. Altman eventually returned as CEO within a week, under pressure from investor Microsoft.

The unrest last year was the result of a leadership battle within the company. Part of the OpenAI leadership wanted a faster pace of technological development and a greater focus on profitability, while another part argued for safety and care.

In addition to Murati, two high-profile researchers are now leaving OpenAI. They are not the first to do so: their co-founders Ilya Sutskever and John Schulman left a few months ago. Co-founder Greg Brockman announced last month that he would soon be taking a sabbatical.

It’s hard to separate the recent exodus from the company’s new AI direction. OpenAI started as a foundation in 2015, with the goal of “protecting society against uncontrolled AI systems”. Now it wants to be a for-profit company, in which the foundation is still there, but which has only a minority.

According to Reuters, the company is going down new investment roundwith which he hopes to raise 6.5 billion dollars (about 5.8 billion euros). This values ​​the AI ​​company at around $150 billion. For comparison, OpenAI was still worth $14 billion in 2021. Tech companies such as Apple and chip maker Nvidia are said to be interested in investing in the company.

In the new investment round, CEO Sam Altman would receive 7 percent of the shares (value equal to 9.3 billion euros).

According to Jelle Zuidema, associate professor of natural language processing at the UvA, not much will change in the company’s working methods despite the new developments. “OpenAI was already operating as a for-profit company The high market value of AI technologies puts a lot of pressure on companies to push the boundaries. “

A few weeks ago the company launched from 1, the first self-proclaimed AI model that can reason for itself”. Zuidema: “OpenAI really needed a new attention grabber to make a new funding round possible. They succeeded: with o1 they delivered something that surprised people. But they didn’t solve anything fundamental with it. o1 is a a much more expensive model than previous models, as it requires a lot of calculation time during use It produces good reasoning examples, but also sometimes big errors.”

As a scientist, he says he is very concerned about the new developments. “When ChatGPT and image generators appeared, we as scientists immediately took notice. With artificially created images, texts and sound, it is difficult to distinguish between fake and real. You need watermarks for that, but unfortunately there are commercial reasons not to. carry out.”