A majority of the OpenAI board members were heavily sympathetic to the AI doomsday cult mostly known through MIRI. A novel discovery spooked them, made them believe that The End Is Near, and that continuing on the path led by Altman would lead to imminent disaster. OpenAI is designed so that the board do pretty much whatever they deem necessary if this happens, and the board interpreted the necessary action as unilaterally halting AI development -- a staple objective of the AI doomsday cult.

So the board did this. They jumped the gun - this extreme move might have become justified at a later point in time, but definitely not yet. In a pedestrian understanding of game theory typical of AI doomsday cult members, the board failed to anticipate that other actors may not agree with their interpretation of reality, and perform their own actions in response. This response largely boiled down to scattering the institutional knowledge of OpenAI to the winds, in their eyes greatly exacerbating the safety concerns that led to firing Altman in the first place.

Altman, meanwhile, as a strong believer in OpenAI's mission, attempted to salvage the situation by brokering a deal to transfer OpenAI's institutional knowledge to Microsoft.

After having their shitty options explained to them - exacerbate a potential safety risk by seeding 750 new AI companies, leave control of the situation to Microsoft's board or step down and retain control under the OpenAI non-profit, they folded like a wet paper bag. The whole saga was a waste of their board seats and influence on humanity's first deployment of AGI. They all got the outcome that their competence deserved.