On Twitter/X, while the purchase of it still appears on paper to have been a terrible decision monetarily, primarily due to the mass advertiser exodus for controversial reasons, the product is doing quite well now.

Estimated number of X users has grown +20-35% in 3 years since purchase, and that growth % should not be understated - that's +110 to +160 million new users. And X is operating more smoothly with more capabilities than it did before purchase.

Depends on what you mean by "quite well". According to the SpaceX S-1, Twitter/xAI xAI posted an operating loss of $2.47 billion on $818 million in revenue for Q1. In 2025 it lost it lost $6.36 billion on $3.2 billion in revenue. SpaceX said it expects a "multi-year investment horizon" for these investments to pay off.

I have no problem with a multi-year investment horizon, but it is far from clear the Twitter purchase was a good decision. xAI is unique in that it has a robust (but not money making as far as we know) social media component. But hasn't had the explosive user growth that its competitors like Anthropic, ChatGPT, or Gemini have had. Anthropic is likely to be profitable soon, and Gemini likely already is.

I don't know how AI will shake out, but I believe due to the enormous capital costs the weak hands will be forced to fold at some point. Right now, xAI is the weak hand.