← sources

Why are investors still dumping money into AI?

source: reddit (reddit:cscareerquestions)published 2026-10-21 05:45ingested 2026-10-22 02:37

retained text

A few things have made headlines this year that even non tech folks should have heard about 1) (Chinese) open weight models are only few months behind Anthropic / OpenAI 2) Massive price war between frontier labs cutting api costs significantly 3) Almost any lab that has invested a couple billion into frontier models have been able to make it - Meta, Google, kimi, Alibaba

All this added together means that first there is absolutely nothing proprietary about Anthropic/ OpenAI. They should only be worth about the investment needed to create this “frontier” team, ie a couple billion. Secondly it’s quite evident open weights will be very big, or even if it’s closed it will be very low margins because of open weight options. Meaning only the AI infrastructure will be worth something, the models/ technology it self is worth very little. 

If  whatever “AGI”  (i hate this word too) is achieved to reduce costs then the large companies for instance an investment bank would just deploy their own AGI. Or pay a very low price for someone to deploy for them.

I thought investors would at least look at the basic economics of this/ the trend of token prices instead of just randomly giving Anthropic a 1T valuation… Can someone explain why?

document fetched 2026-10-22 00:24 · canonical https://reddit.com/r/cscareerquestions/comments/1xbexvj/why_are_investors_still_dumping_money_into_ai