The AI Race Just Got Awkward

Re:The AI Race Just Got Awkward by senordevnyc on Wednesday September 30, 2026 @12:49PM (#49911350)

Thank you!

I’m incredibly skeptical that OpenAI is spinning up custom ASICs for improved inference performance, but they never thought of optimizing KV cache until a tiny Chinese lab did it? Give me a break.

[ Parent | on HN ]

Re:The AI Race Just Got Awkward by bel8 on Wednesday September 30, 2026 @12:56PM (#49911470)

They certainly thought. But were they able to do it now without DeepSeek papers?

timeline suggests not.

[ Parent | on HN ]

Re:The AI Race Just Got Awkward by senordevnyc on Wednesday September 30, 2026 @01:13PM (#49911700)

So DS comes up with 437x improvement, and the evidence that O/A copied them is that they dropped caching prices by like 50%? Really?

[ Parent | on HN ]

Re:The AI Race Just Got Awkward by abcthingx on Wednesday September 30, 2026 @01:19PM (#49911774)

What evidence evidence would actually change your stance on this? a formal admission from OpenAI and Anthropic. Yet that's unlikely

[ Parent | on HN ]

Re:The AI Race Just Got Awkward by senordevnyc on Wednesday September 30, 2026 @04:02PM (#49913542)

Any actual evidence would be worth considering. This is a pretty weak correlation and an assumption, nothing more.

[ Parent | on HN ]

Re:The AI Race Just Got Awkward by croon on Thursday October 01, 2026 @04:03AM (#49919016)

80%. You can't do that unless you significantly cut costs somewhere. The timing is certainly indicative.

[ Parent | on HN ]