They realized they could use the government to protect their interests. After years of praising open-source, they also realized that they don't really like competition. Queue up Peter Thiel's Zero to One and his praising of monopolies.
Cf. with the posture that OpenAI is taking towards government and government regulation.
When Sam Altman went to congress, his message was that AI is dangerous, and therefore the government should regulate it. When asked how smart AI had to be before it should be regulated, he said that AI which is sufficiently advanced to be able to persuade people is dangerous--we don't want bad players and their machines influencing elections, do we?
This is beautiful. It looks like a very nice, bright line the government can use to help it decide what the regulations should be. And how very public spirited--statesmanlike, even--of Sam Altman to ask for government regulation of his own company.
But it achieves two very important goals for OpenAI:
1. Every time you use ChatGPT you want it to persuade you! If you didn't want it to change your opinions about something, you wouldn't have asked it a question. Therefore, the government regulations will apply to any OpenAI competitor.
Anybody competing with OpenAI already faces huge barriers to entry. But he just made it even harder. Competitors, by law, will have to hire a phalanx of lawyers and compliance officers, whose full time job will be to make programmers jobs harder.
2. The more we are talking about regulating dangerous AI, the LESS we are talking about the fact that OpenAI just stole the IP of virtually everybody who has ever copywritten anything!
To the point that we don't talk about it any more at all. Sam Altman has accumulated all of the IP of the information age, for free, and is now in a position to sell it back to us.
It's a paradigm example of how corporations should lobby governments. Business schools and law schools will be doing case studies of this until the end of time.
Competition Is for Losers