logoalt Hacker News

thraway3837today at 12:36 AM3 repliesview on HN

Good analysis. But who cares? It takes a long time for companies to figure out how to become profitable. And I honestly believe that OpenAI/Anthropic etc. have done humanity a huge favor. The money they're burning is not yours or mine. They're institutional investor money. So, again, who cares?

It will become profitable. Local models and local on-laptop inference will get good enough. This argument has been made for decades. It's not like everyone is walking around hosting email and photos on their personal machines. Sometimes it takes a large investment to make servers and clouds for this stuff possible.

We need to get away from this idea that in order for one thing to succeed, the other must fail. We also need to stop thinking in binary and accept that all these things (profitability, local models, powerful laptops, etc.) can all happily coexist.


Replies

joshuastudentoday at 12:45 AM

They haven't done humanity a favor at all. The innovation that these LLMs have produced has been small. A few fun math theorems where the answer was gleaned from a pattern in the training data. Great,.but it doesn't change the world one bit.

That latest drug for pancreatic cancer? Yeah, all human. After the trillions already spent, AI hasn't come up with any new medications, no new inventions to save lives... Nothing

show 1 reply
layer8today at 12:41 AM

Buyers of consumer storage, RAM, and GPUs care. People affected by the data center buildouts care. Workers losing jobs due to underpriced tokens care. People on the receiving end of AI slop care.

show 1 reply
maximinus_thraxtoday at 12:51 AM

> The money they're burning is not yours or mine. They're institutional investor money. So, again, who cares?

This is not happening in a vacuum. A lot of index funds and retirement accounts have bought into AI and AI adjacent companies, many with stakes in OpenAI. If OpenAI keels over, even when private, it will affect a lot of americans. If they IPO, it's even worse.

show 1 reply