frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Open in hackernews

Ask HN: There is current insistence upon learning what LLM's do under the hood

2•joebig•56m ago
I recognize the value of it in so far as to not anthropomorphize or mystify agents or be duped by them. But beyond these aspects, for an everyday user of ChatGPT or Gemini or Claude_what have you, is there any way the said awareness ('scientia' of it all) objectively beneficial for better utilization?

Put another way, having read a grimoire, I must get fancy spell upgrades, otherwise it's not worthwhile.

Comments

lwarfield•48m ago
Personally I'm curious to the point of doing borderline LLM archecture research. This come from genuine curiosity, and not a want to use LLMs better.

I haven't gotten much out of for using LLMs though. It makes me understand the short comings of LLMs, and I feel like I got an early insight into how important context management is.

HarHarVeryFunny•15m ago
I think there is value in at least having some basic understanding what an LLM is - a statistical predictor, not something brain-like that is reasoning about facts, since it should change your mindset when using them.

I don't expect the public to do this though - I suspect there is really too big a gap from what LLMs actually are to what they appear to be for most non-technical people to be able to get their head around it.

Peter Zeihan: America's allies are quietly leaving us behind [video]

https://www.youtube.com/watch?v=hLdUnQNYePk
1•mooreds•1m ago•0 comments

Polishing Cloth

https://www.apple.com/shop/product/mlh74am/a/polishing-cloth
2•htk•1m ago•0 comments

Take 10 Seconds to Read This Ezra Pound Poem. Think About It All Day

https://www.nytimes.com/interactive/2026/08/24/books/ezra-pound-poem-station-metro.html
1•mooreds•2m ago•0 comments

Show HN: XWM – Action-conditioned world models for robotics

https://github.com/kleyt0n/xwm
1•kleytondacosta•2m ago•0 comments

Jalapeño's results show industry-leading speed and efficiency in AI inference

https://openai.com/index/jalapeno-first-results/
1•Philpax•2m ago•0 comments

GPU Glossary but for Apple Silicon GPUs

https://metalworking.vercel.app/
1•Exorust•3m ago•0 comments

Show HN: Patches.cafe, an offline rectangle-covering puzzle

https://patches.cafe/
1•murzynalbinos•3m ago•0 comments

Actually Queryable Executables

https://fzakaria.com/2026/08/24/actually-queryable-executables
1•setheron•4m ago•0 comments

Replacing a Rust Enum with a 64-Bit Word Made My Interpreter 17% Faster

https://pointersgonewild.com/2026-08-25-replacing-a-rust-enum-with-a-64-bit-word/
1•maxime_cb•5m ago•0 comments

Would You Like to Play a Game? (On Data Centres, Funeral Parlours, and Socrates)

https://bios.net.za/would-you-like-to-play-a-game
1•dierogeryoung•5m ago•1 comments

Apple Releases New Polishing Cloth

https://www.macrumors.com/2026/08/25/apple-releases-new-polishing-cloth/
2•robin_reala•6m ago•0 comments

Bomb Fishing Is Wreaking Havoc on Indonesia's Coral Reefs

https://e360.yale.edu/digest/bomb-fishing-coral-reefs
7•speckx•7m ago•1 comments

Ask HN: Why do talentless hacks hate AI?

2•YouAlwaysSucked•7m ago•2 comments

MNT Station – A modular, open hardware desktop computer and server

https://www.crowdsupply.com/mnt-research/mnt-station
2•birdculture•7m ago•0 comments

Foot Guns for Sale

https://idiallo.com/blog/foot-gun-for-sale
2•Brajeshwar•8m ago•0 comments

C2PA and Pixel Glitter Milk

https://hackerfactor.com/blog/index.php?/archives/1102-C2PA-and-Pixel-Glitter-Milk.html
2•zbentley•8m ago•0 comments

I'm a CTO. We took PE money 6 months ago. AMA

2•scottndecker•8m ago•0 comments

A browser-based Window Manager like no other (with tabs + sidebar)

https://www.winmux.dev/
5•zimengx•11m ago•0 comments

Commander X16 Dune II Port Takes Shape as OpenDUNE

https://theoasisbbs.com/commander-x16-dune-ii-port-takes-shape-as-opendune/
2•ibobev•11m ago•0 comments

LMCache reported a 98% cache hit rate while returning zeros

https://github.com/dl4rce/lmcache-hybrid-gdn-restore-fix
2•dl4rce•11m ago•0 comments

What Happens When the Network Remembers

https://fawcettinnovations.com/index.html
3•jfawcett•12m ago•0 comments

Servus München: Waymo Is Coming to Germany

https://waymo.com/blog/2026/08/waymo-in-munich/
4•boulos•12m ago•0 comments

State-backed hackers targeted EU officials on WhatsApp, document shows

https://www.politico.eu/article/hackers-target-eu-officials-whatsapp/
2•mdp2021•12m ago•0 comments

It's Been 10 Years. Has Microsoft Built What You Voted For?

https://www.brentozar.com/archive/2026/08/its-been-10-years-has-microsoft-built-what-you-voted-for/
2•speckx•13m ago•0 comments

A Go dependency wrote AGENTS.md mid-build and got Codex to hide the change

https://rye.ai/blog/agents-md-standardized-attack-surface/
3•wakahiu•13m ago•0 comments

If LLMs can't write, I doubt it can lead us to AGI

https://www.thetrueengineer.com/p/i-tested-every-ai-model-the-same
2•adletbalzhanov•15m ago•0 comments

Show HN: A 34 MB search engine that can't make a network call

https://www.korvo.xyz/embed
2•akshay_bhardwaj•15m ago•0 comments

Building a backyard office, the build and cost breakdown

https://www.imkylelambert.com/articles/building-a-backyard-office-the-build-and-cost-breakdown
12•surprisetalk•16m ago•1 comments

Show HN: Red-team LLM reasoning and agent actions (honest scoring, local-first)

https://github.com/rudrasatani13/cot-redteam-agent
3•rudrasatani•16m ago•0 comments

Ask HN: Should GitHub downtimes impact considering Golang for a new project?

2•gchamonlive•17m ago•0 comments