It's a losing record, yes.

Only someone who went to Harvard and was a presidential speechwriter could prepare notes for a convoluted, delusional address that was that stupid. Whereas my buddy in grad school said of Israel/Palestine: "I'm never gonna be on the side with the fighter jets." He nailed it!

I'm extremely tired of the way advocates for Palestinian human rights are dismissed as having only recently formed their opinions because of social media.

No. These opinions come from Millennials and Gen Z being more diverse and better educated than any generation before them.

American apartment living past the age of 30 is overwhelmingly either associated with despair and personal failure (when living in an apartment in 99.9% of the country) or unattainable glamorous living (when living in like two neighborhoods in NYC)

bsky.app/profile/prch...

But private equity is not interested in a sustainable business model either. PE doesn't care if genAI ever covers its costs.

PE wants a 10X return on investment. They want to sell their stake to the next big buyer and then bail out.

If it weren't for a giant, sloshing pool of private equity keeping it afloat, genAI wouldn't have lasted this long.

This is why predicting when the bubble will pop is so tricky. It's hard to say how much money is "out there" to be sunk into this.

There may be some future where academics or day traders run bespoke LLMs on private datasets at hefty costs. That's not the present business model. Google, FB, OpenAI, or Anthropic would not pay hundreds of billions to keep that afloat.

1) The current model (lying chatbot) does not have enough paying users.

2) The improvement in quality necessary to entice paying users has no economy of scale.

3) The few useful edge cases *today* are not enough to sustain the product at its present size.