Thread

1. amelius · 1789928450 · focus · HN ↗

   Will that be enough for AI's hunger?

   1. GoToRO · 1789929801 · focus · HN ↗

      It will be just in time for when AI will run very well on consumer hardware and the need for data centers will collapse.

      1. IshKebab · 1789930756 · focus · HN ↗

         That's never going to happen. By the time you can run current frontier models on your $10k desktop the frontier will have massively advanced and people will want those models instead.

         1. hypfer · 1789931551 · focus · HN ↗

            I'm not sure if this prediction will hold true.

            We're not seeing the progress in those "frontier models" that we have previously seen. There's certainly still gas left in tank tank, but we're way into the diminishing returns by now.

            Cloud inference still beats hardware investments by orders of magnitude of course, but that's only if your data doesn't really matter to you.

            1. airspresso · 1789931811 · focus · HN ↗

               We are certainly not in the diminishing returns phase for LLM progress. No sign of that yet.

               1. SideQuark · 1789959234 · focus · HN ↗

                  Google scholar has a flood of papers showing LLM diminishing returns on pretty much every facet

                  <a href="https:&#x2F;&#x2F;scholar.google.com&#x2F;scholar?hl=en&amp;as_sdt=0%2C23&amp;q=llm+diminishing+returns+&amp;btnG=" rel="nofollow">https:&#x2F;&#x2F;scholar.google.com&#x2F;scholar?hl=en&amp;as_sdt=0%2C23&amp;q=llm...

                  1. T-A · 1789975332 · focus · HN ↗

                     The first title I see there is &quot;The Illusion of Diminishing Returns: Measuring Long Horizon Execution in LLMs&quot;:

                     <a href="https:&#x2F;&#x2F;proceedings.iclr.cc&#x2F;paper_files&#x2F;paper&#x2F;2026&#x2F;hash&#x2F;3b4e1336f775c3dba16ebbb8d2afd258-Abstract-Conference.html" rel="nofollow">https:&#x2F;&#x2F;proceedings.iclr.cc&#x2F;paper_files&#x2F;paper&#x2F;2026&#x2F;hash&#x2F;3b4e...

                     1. SideQuark · 1790024360 · focus · HN ↗

                        And you apparently only read the title and ignored the content.

                        It’s also poor reasoning to you only pick one thing you think supports your view, and ignore vastly more things not supporting it, all from the same useful criteria.

                        Now if only you’d carefully read the report you chose, and spend equal time looking at ample presented evidence, you’d develop a more accurate understanding.

Unofficial Hacker News client; not affiliated with Y Combinator.