Egregoros

Signal feed

Timeline

Post

Remote status

Context

14
It's not like I never gave AI a try at all but I did NOT realize they were still THIS BAD in 2026. I really gotta be more careful.

This is grok output with thinking.

>$186 million tokens
>$tokens
????

>$400k / $2 = 186
????


I repeated this prompt a couple other ways and it gave figures between 1 and 2 billion tokens for the cost of a strix halo setup capable of running GLM 5.1, for those wondering what I was doing.

I like actually knowing what I am talking about and have no interest in vague platitudes. This is a large reason why I am dissatisfied when I ask an LLM to give me a cost-benefit analysis or similar (it puts "vastly more" in a "gain" category when I ask it for a table with hard numbers) and also a large reason why I don't feel like talking AI with you if you say shit like this. Coalpost old friend, sorry

I could do it myself for cheaper. I already know what's good. I just try to take the busywork off myself with the LLM. Keeping an eye on these things. I think they could be useful with a big ass harness but I will not build such a harness myself. Either the bubble corpos do it themselves or I will continue to criticize the niggertarded output!

Replies

0

Fetching replies…