Caught a Google engineer saying their new AI model hallucinates less but costs 3x more to run

I was sitting at a coffee shop last week and overheard two Google engineers talking about their latest LLM release. One said they cut hallucination rates by 40% but the compute cost per query went from $0.02 to $0.07. Makes me wonder if that trade off is actually worth it for most users, or if it's just a flex for enterprise clients. Has anyone else noticed pricing creeping up across the big AI companies this year?

0 comments

0 Comments

No comments yet

Be the first to share your thoughts on this discussion.