If you think that there's nothing more to learn about the OpenAI drama, think again. First, it was thought that the former board fired Altman over an AI breakthrough. Now, it seems that Altman may have made the first move.
Altman Wanted a Board Member Out First
The entire narrative pointed to the former board members being the bad guys in the story, but there could be more to the story than we know. The initial rumors only stated that the higher-ups no longer trusted Altman since he wasn't entirely forthcoming with OpenAI projects.
After the rumors about Altman discovering an AI breakthrough called Q* have been disproven, new reports have emerged brought by an insider in OpenAI. It was stated that the chief executive tried to manipulate the board into firing a board member, Helen Toner.
Due to this, the former board members believed that Altman was "manipulative and conniving," as reported by Ars Technica. Altman reportedly took issue over Toner co-writing a paper that focuses on the safety measures that can be implemented for AI.
In addition to that, she also mentioned Anthropic's chatbot Claude as an example of what a responsible release looks like, as they decided "not to productize its technology in order to avoid stoking the flames of AI hype."
Since Claude was delayed on purpose until another company released a similarly capable product, Anthropic managed to prove that it was capable of avoiding "frantic corner-cutting," which ChatGPT has been accused of even by third parties.
Since the paper painted OpenAI in a bad light, Toner apologized to the board, but Altman used the opportunity to goad the board into ousting Toner. The chief executive was accused of misinterpreting reactions from the board to achieve the outcome he wanted.
If this narrative is true, then perhaps the former board's decision to oust Altman was part of its efforts to create AI products that will benefit AI humanity, which is the opposite of a chatbot that lacks the necessary guardrails.
OpenAI COO Brad Lightcap did say that Altman's ouster had nothing to do with OpenAI's financial, business, safety, or security/privacy practices. Instead, it was a "breakdown in communication between Sam and the board."
Microsoft Backs Sam Altman
Microsoft already had plans in place in the event that OpenAI does not rehire Altman. The software giant was ready to add the chief executive to its AI research team, along with the employees who were willing to follow the CEO.
In fact, the company said that the decision to fire Altman was "mind-bogglingly stupid." What's worse is that the board kept the decision from Microsoft so that Sam Altman could not be warned about the impending layoff, as per The Verge.
Microsoft President Brad Smith also tried clearing the allegations about Altman creating an unsafe AGI, saying that we are years, if not decades, away from developing something as advanced as the rumored Q*.
The mentioned AGI in the argument was the same AI breakthrough that OpenAI researchers were worried about. The group sent a letter to the board expressing their concerns, which is why it was thought to be the reason why Altman was removed as CEO.
