frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

Open in hackernews

Advising is the next step for LLMs, a change in how they answer questions

2•Taikhoom10•4h ago
I am not some sort of AI expert but this is what I personally think is going to be the next big step for llms. For simplicity I will be using ChatGPT throughout rather than llm. (First Mover Advantage perk) The next big leap isn't agentic capabilities or search through AI. Those are more of mediums of through which AI will progress into AGI one day. I think. What I want to talk about is the way llm's answer questions. I wont talk about if there right or wrong, more of how they just answer. Rather lets say I ask ChatGPT to help me write a title for a post on a forum. It will help me do that, now one can argue it might not do a good job. But rather say it responds by saying, don't post on this said forum right now because there isn't a lot of viership at this time. Peak post times are in the early morning. And of course there can still be a option to go through with the title brainstorm thing, but that advice is crucial. Llm's are a bunch of yes men by default, they need to be advisors. And good ones that too. A king needs a advisor that actually critiques the mistakes and provides better advice, not agree with whatever the king has to say. Otherwise the king would not need a advisor. The same way the llms today need to provide advice sometimes, rather than saying yes. I know this might be counterintuitive but humans make mistakes. Like ordering overly expensive coffee, llms should help us reduce those mistakes as well.

The Influence of Topography on the Global Terrestrial Water Cycle

https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2023RG000810
1•bryanrasmussen•13s ago•0 comments

Common Expression Language – Fast, safe expression language

https://cel.dev/
1•thunderbong•1m ago•0 comments

Does your brand appear in AI? (Simple Tool)

https://pop32.com/
1•onescales•3m ago•0 comments

I just decompiled a decompiler so I could use it to decompile itself (2008)

https://dougmccune.com/blog/2008/09/21/i-just-decompiled-a-decompiler-so-i-could-use-it-to-decompile-itself/
1•userbinator•6m ago•0 comments

Optimists are alike, every pessimist has their own way

https://www.kobe-u.ac.jp/en/news/article/20250722-66828/
1•XzetaU8•6m ago•0 comments

AMD MI300 compute and memory partitioning

https://rocm.blogs.amd.com/software-tools-optimization/compute-memory-modes/README.html
1•ankitg12•14m ago•0 comments

A break from tech: a writing piece about Ghost Towns

https://helenaeverley.substack.com/p/ghost-towns
1•hez2000•15m ago•0 comments

Show HN: A P2P Contribution-Based Protocol for Civic Trust

https://github.com/contribution-protocol/contribution-protocol-project
1•mzk_pi•16m ago•0 comments

The Incomm Scientology Keyboard / Cherry G80-1165HAU

https://old.reddit.com/r/MechanicalKeyboards/comments/1m1f0p6/the_incomm_scientology_keyboard_cherry_g801165hau/
1•decimalenough•23m ago•0 comments

A parser and interpreter for a small language

https://blog.ploeh.dk/2025/07/07/a-parser-and-interpreter-for-a-very-small-language/
2•Bogdanp•26m ago•0 comments

Show HN: AI That's You

https://www.ren.social
19•rennysocially•26m ago•4 comments

Complete silence is always hallucinated as "ترجمة نانسي قنقر" in Arabic

https://github.com/openai/whisper/discussions/2608
17•edent•28m ago•3 comments

Ask HN: Are OpenAI models IP-range restricted?

1•lynx97•37m ago•0 comments

Conversational image segmentation with Gemini 2.5

https://developers.googleblog.com/en/conversational-image-segmentation-gemini-2-5/
3•fofoz•40m ago•0 comments

Startup Lessons (2013)

https://defmacro.org/2013/07/23/startup-lessons.html
1•fogzen•42m ago•0 comments

Rethinking Syntax: Binding by Adjacency

https://github.com/manifold-systems/manifold/blob/master/docs/articles/binding_exprs.md
1•owlstuffing•43m ago•1 comments

Video of Earth's surface lurching sideways in an earthquake

https://theconversation.com/the-first-video-of-earths-surface-lurching-sideways-in-an-earthquake-offers-new-insights-into-this-force-of-nature-261004
3•tigerlily•44m ago•0 comments

Vitalik Buterin's response to AI 2027

https://vitalik.eth.limo/general/2025/07/10/2027.html
1•fbrusch•45m ago•0 comments

Trump Aides Discussed Ending Some SpaceX Contracts, but Found Most Were Vital

https://www.wsj.com/business/trump-aides-discussed-ending-some-spacex-contracts-but-found-most-were-vital-d1cf9ab5
1•doener•46m ago•0 comments

Flight heads back to Paris after being denied landing clearance in Chicago

https://local12.com/news/nation-world/flight-af136-heads-back-to-paris-after-being-denied-landing-clearance-in-chicago-air-france-o-hare-airport-cincinnati-travel-headaches-nowhere-operation-reasoning-midflight-revert-course-charles-de-gaulle-radar-runway-26r-cruising-altitude-route-origin
4•georgecmu•49m ago•0 comments

PEP 798 – Unpacking in Comprehensions

https://peps.python.org/pep-0798/
2•asicsp•50m ago•1 comments

Oklahoma Pre-K teacher allegedly calls being left-handed 'evil' and 'sinister'

https://kfor.com/news/oklahoma-pre-k-teacher-allegedly-calls-being-left-handed-evil-and-sinister/
1•JumpCrisscross•53m ago•1 comments

A Friendly Introduction to SVG

https://www.joshwcomeau.com/svg/friendly-introduction-to-svg/
1•OuterVale•56m ago•0 comments

Build functional Nostr apps in one prompt

https://shakespeare.diy/
1•sak5sk•56m ago•0 comments

Measuring heart rate with consumer ultra-wideband radar

https://research.google/blog/measuring-heart-rate-with-consumer-ultra-wideband-radar/
1•phreeza•59m ago•0 comments

Find domains using regex-like AI prompt

https://www.appealing.ai/
2•yonibulkin•1h ago•2 comments

My Two Lives as a Programmer

https://www.maxmynter.com/pages/blog/two-lives
1•vedhsaka•1h ago•0 comments

Show HN: Created an "Indian Dowry Calculator", Calculate How Much You're Worth?

https://dahejcalc.in
1•airobus•1h ago•0 comments

Building Faster AMD64 Memset Routines

https://msrc.microsoft.com/blog/2021/01/building-faster-amd64-memset-routines/
1•90s_dev•1h ago•0 comments

Machine Bullshit: Characterizing the Emergent Disregard for Truth in LLMs

https://arxiv.org/abs/2507.07484
1•ilt•1h ago•0 comments