The worrying inflection point that US-based AI platforms will embrace

Elon Musk's surprise move of merging xAI with X (ex-Twitter) is a process that should have required at least four independent parties: each company and their respective investment banks. Here, the deal only involved Elon agreeing with Elon and Morgan Stanley agreeing with Morgan Stanley. This ensures that the valuation of the end result is purely what Elon wants it to be.

Elon Musk said his xAI artificial intelligence startup has acquired the X platform, which he also controls, at a valuation of $33 billion, marking a surprise twist for the social network formerly known as Twitter. - Bloomberg, March 29, 2025

From a business and market point of view, this means a few things.

If Musk intends (among many other things implied in this move) to catch up in the AI game by using X/Twitter users' data, OpenAI or Anthropic will have to do the same with... TikTok. Meta/Facebook having its very own consanguine relationship with its users' data, and Microsoft having clear access to millions of companies' data, this leaves Apple pretty much cut for any serious AI race from now on.

This is an enormous inflection point from AI platforms that will scale and try to be dominant for the next decades. Not as expanded Wikipedias but rather as idiot savants raised on the worst type of data possible: us shouting at each other over disinformation and flat-earthism. It's a serious bias now acted on in how these large models will reason, exchange, and give back information. The worst of the net from these last 20 years.

Not to mention that these models still have only one way to monetize: becoming Netflix and catering to consumers' superficial gimmicks like Studio Ghibli filters for picture making (in exchange for a heavily subsidized monthly fee) or... ads (which is even worse).

No wonder Sam Altman is terrified by China's Open Model approach.

As we speak, only Microsoft is becoming the white knight with an alternate approach to this whole mess. Oh, the irony.

What about solving hard problems with AI? Well, we still managed to do this well before ChatGPT became all the craze (and it feels like it went downhill from there):

Note: Of much less importance at this point, the merger is also just another signal that the rule of law in the US is becoming increasingly a vague afterthought. It is certainly nothing that big players will trust to be enforceable in the next few years, not to mention that the FTC's Bureau of Competition won't take care of any antitrust enforcement.