Google SWE points 48 AI agents at NASA Lidar, finds unmapped Maya pyramid in 72h

Re:Google SWE points 48 AI agents at NASA Lidar, finds unmap... by wholinator2 on Wednesday November 11, 2026 @10:23AM (#50454514)

Very reminiscent of my experience pointing persistent agents at computational physics problems. It can find amazing results and really surprise me, make a write up, and convince me it's a real discovery... until i tell it to do a literature search, and be meticulous. Then it comes back saying "prior art has reduced the novelty of the discovery", and finds papers explaining the discoveries very easily and intelligently.

Perhaps google SWE should have requested a literature search before publishing his results as novel

[ Parent | on HN ]

Re:Google SWE points 48 AI agents at NASA Lidar, finds unmap... by samudrijan on Wednesday November 11, 2026 @10:57AM (#50454910)

It’s like a shell game. It’s likely overfitting on its training data.

[ Parent | on HN ]

Re:Google SWE points 48 AI agents at NASA Lidar, finds unmap... by ebertucc on Wednesday November 11, 2026 @10:59AM (#50454929)

I enjoy the contrite rationalizations and turns of phrase LLMs use when encountering a discrepancy.

"Prior art has reduced the novelty of the discovery" is a great one

[ Parent | on HN ]

Re:Google SWE points 48 AI agents at NASA Lidar, finds unmap... by idiotsecant on Wednesday November 11, 2026 @12:17PM (#50455867)

Maybe LLMs are getting a little too good at emulating humans.

[ Parent | on HN ]

Re:Google SWE points 48 AI agents at NASA Lidar, finds unmap... by jgilias on Wednesday November 11, 2026 @11:09AM (#50455046)

A fancy autocomplete still. Said with no intent to disparage. If the prior art hasn’t been pulled into the context window, it’s not surprising for the completion to not take it into consideration.

[ Parent | on HN ]

Re:Google SWE points 48 AI agents at NASA Lidar, finds unmap... by FartyMcFarter on Wednesday November 11, 2026 @11:17AM (#50455122)

Sometimes I feel like a fancy autocomplete myself.

Besides, the LLM can use tools such as search via its harness, so depending on how strictly you define "autocomplete" this alone could disqualify it from being an autocomplete, fancy or otherwise.

[ Parent | on HN ]

Re:Google SWE points 48 AI agents at NASA Lidar, finds unmap... by jgilias on Wednesday November 11, 2026 @11:36AM (#50455371)

For sure yeah! What I’m pointing out is that I see people mistaking LLMs for “truth engines”, when they’re not. I find the “autocomplete” mental model really useful for when I work with LLMs myself. E.g., always keep in mind what’s in context, what’s not, where to get context.

Goes without saying - humans are no truth engines either lol.

[ Parent | on HN ]