The company behind headline-grabbing AI bot ChatGPT has sacked and then reinstated its CEO, all in the space of just five days.

Sam Altman was one of the original funders of OpenAI (alongside Elon Musk, Microsoft and Amazon Web Services – among others), and in 2019 was installed as its CEO. In the intervening years he became more than just the company chief, however, but was increasingly being seen as something of a godfather to the industry – giving talks at high profile conferences, appearing before US legislators, and generally being viewed as a trusted insider regarding the direction in which AI is headed.

Despite this growing profile, the board of OpenAI decided to part ways with Altman, giving him the sack before going on record to say he had not been “consistently candid” in his communications.

Somewhat ironically, though, the AI business appeared to overlook one very important factor: people power.

With Altman being a popular individual both within and outside of the company, the board was delivered a letter within days of his sacking in which almost the entire workforce threatened to quit were he not reinstated. Signatories included the interim CEO and even one board member who changed their mind on the original verdict.

Of course, the bottom line was also thought to be a major factor in the board’s U-turn, with investors displeased with the news of Altman’s sacking. Although the company was recently valued at $80bn (£63bn) it has enormous overheads. As one insider put it: so data intensive is the process to respond to user queries that even the super-rich Google would go out of business if it had the same outgoings for each search it completed.

There was also the issue of potentially lucrative secrets going elsewhere, with Altman having agreed terms to work with Microsoft in his vanishingly short period of unemployment.

All this meant that within just five days of being sacked, Sam Altman was back in his old job. It also puts him back in position to launch the much-touted follow-up to ChatGPT, thought to be so advanced that it would make its paradigm-shifting predecessor look like “a quaint relative”.