LinkedIn founder and Greylock partner Reid Hoffman is leaving the nonprofit board of OpenAI, he announced in a post on Friday.
Hoffman cited a desire to invest in companies using OpenAI’s software, which could cause a conflict of interest, he said in the post.
related investing news
“I started to wonder: Will my position as a 501c3 board member of OpenAI potentially look like it’s leading to differential economic advancement? To be clear, since the start, OpenAI and its board has been very careful to monitor and avoid any conflicts to date,” Hoffman wrote. “But by stepping off the board, I can proactively put to rest any downstream potential issues for both OpenAI and all Greylock portfolio companies I’ve backed.”
OpenAI is one of the leaders in developing and training large language models, a relatively new approach in artificial intelligence that uses lots of computing power and data to produce software that can return blocks of text or answers that look like a human wrote them.
Hoffman said companies that use large language models could become incredibly valuable.
“There are future trillion-dollar companies being built and invested in right now, which will not only change markets, but launch new ones,” Hoffman wrote.
OpenAI gives other companies access to its AI models through a programming interface, or API. Earlier this week, it said it would charge companies using its AI about one-fifth of one cent to produce about 750 words.
OpenAI has an unusual corporate structure: It was originally founded as a nonprofit in 2015, before shifting its structure in 2019 to become a for-profit company, although it says its profit is “capped” as part of its bylaws and that excess profits would flow to the company’s nonprofit wing.
The unusual structure factored into Microsoft’s recent $10 billion investment in OpenAI, in which the tech giant would reportedly get a share of OpenAI’s profits until it recouped its investment. Hoffman is on Microsoft’s board.
Previously, Tesla CEO Elon Musk left OpenAI’s board in 2018. OpenAI said at the time that Musk was avoiding conflicts with Tesla’s work in artificial intelligence.
An OpenAI representative didn’t return a request for comment.