frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Mercor AI trainer in various fields [Referral]

https://work.mercor.com
1•elsadek•1m ago•0 comments

NPM install Vite is broken

https://github.com/rollup/rollup/issues/6168
1•nvader•1m ago•0 comments

The 900-page book that China watchers count on

https://www.ft.com/content/06611bfa-5705-4f58-9ed1-5d4c869769b3
1•bookofjoe•2m ago•1 comments

Video Games hottest new platform is an old one: Websites

https://www.bloomberg.com/news/articles/2025-11-07/video-games-hottest-new-platform-is-an-old-one...
4•Jimroll•2m ago•0 comments

A deeper dive on the MIT study

https://substack.com/inbox/post/178296459
1•mathattack•2m ago•0 comments

Update Kewin is more powerful – finetuned to sell

https://www.kewin.ai
1•sivacheekula•3m ago•1 comments

Nested Learning: A new ML paradigm for continual learning

https://research.google/blog/introducing-nested-learning-a-new-ml-paradigm-for-continual-learning/
1•pr337h4m•4m ago•0 comments

N8n community node – cascadeflow, Reduce AI costs 30-65% with model cascading

https://github.com/lemony-ai/cascadeflow
1•saschabuehrle•4m ago•1 comments

Show HN: Paideia LMS - A Modern, Single-Binary LMS (v0.5) – Seeking Feedback

https://docs.paideialms.com
1•yomaru_1999•5m ago•1 comments

Using Codex CLI with GPT-OSS:120B on an Nvidia DGX Spark via Tailscale

https://til.simonwillison.net/llms/codex-spark-gpt-oss
1•kgwgk•7m ago•0 comments

Forwardix: Graphical manager for you kubectl forwards with embedded browser

https://git.einfach.biz/forwardix/forwardix
1•chwonl•8m ago•1 comments

Show HN: I built single time-sorted place to read HN, Tildes, Lobsters, Slashdot

https://limereader.com
1•busymom0•9m ago•0 comments

Models Matter

https://www.cybadger.com/models-matter/
2•mooreds•10m ago•0 comments

You can't leave unless you buy something

https://www.sfgate.com/food/article/sf-grocery-store-installed-security-measures-21138800.php
1•structuredPizza•10m ago•0 comments

Unraveling of workplace protections for delivery drivers:Amazon vs. UPS model

https://theconversation.com/the-unraveling-of-workplace-protections-for-delivery-drivers-a-tale-o...
1•INGELRII•12m ago•0 comments

Show HN: DocSumm AI – Source-linked summaries for long PDFs/URLs

https://github.com/RohitRajdev/docsumm-ai
1•rohitrajdev•13m ago•0 comments

B.C. SAR group says Facebook page hijacked by unauthorized individual

https://www.cbc.ca/news/canada/british-columbia/pemberton-search-and-rescue-facebook-hijacked-9.6...
1•barbazoo•16m ago•1 comments

Claude Code to manage engineering teams

https://www.devashish.me/p/why-5x-engineers-dont-make-5x-teams
1•devashish86•21m ago•1 comments

Show HN: Warp Documentation Automation – Built with Claude (99% Automatic)

https://github.com/bryankaufman/warp-documentation-automation
1•Telerad•21m ago•0 comments

A small portal to the internet's weird and useless side

https://clicktheredpill.com
1•absurdwebsite•22m ago•1 comments

Decline in US physical stature parallels diminution in life expect. rate of incr

https://www.sciencedirect.com/science/article/pii/S2352827325001260
1•bikenaga•22m ago•0 comments

Cencora makes $1B investment in pharma supply chains

https://www.freightwaves.com/news/cencora-makes-1-billion-investment-in-pharma-supply-chains
1•crescit_eundo•23m ago•0 comments

Top Ways to Cope with Inflation

https://queensvoicenyc.blogspot.com/2025/10/queens-news-rising-inflation-tips-on-how-to-cope.html
1•NYCNews•26m ago•1 comments

Automated PDF Generation with Typst

https://typst.app/blog/2025/automated-generation/
1•thomas34298•27m ago•0 comments

Deep Learning Without Training

https://zenodo.org/records/17535844
1•car•27m ago•1 comments

The Truth Codex – A Charter for Friction-Tolerant Intelligence

1•kraull•37m ago•0 comments

Airway Beacons

https://thesurveystation.com/airway-beacons/
2•gregsadetsky•37m ago•0 comments

Ask HN: Why not combine dating with board game design and testing?

4•amichail•37m ago•2 comments

A Week of Teaching

https://fivetwelvethirteen.substack.com/p/followup-a-week-of-teaching
2•yorwba•40m ago•0 comments

Qatar-linked intelligence operation targeted ICC prosecutor's alleged victim

https://www.theguardian.com/law/2025/nov/06/qatar-linked-intelligence-operation-targeted-icc-pros...
1•myth_drannon•41m ago•0 comments
Open in hackernews

Claude Is Down

https://status.claude.com/incidents/tgtw1sqs9ths
63•agrocrag•4h ago

Comments

bashy•4h ago
Yeah, getting this;

API Error: 529 {"type":"error","error":{"type":"overloaded_error","message":"Overloaded"},"request_id":null}

starf1sh•4h ago
Better start catching up with latest developments on HN
_andrei_•4h ago
what are we gonna dooo?
golbez9•3h ago
It's over!
bitwize•3h ago
(in Homestar Runner voice) The good times awe ovew!
sam1r•3h ago
everyone, switch to open ai for 50% off. today only!
oersted•3h ago
OpenAI's track record has been rather poor this month as well actually, look at all the yellows and reds: https://status.openai.com/
sam1r•3h ago
Oh wow, I actually had no idea. It would be super nice to see all the AI API's statii on a single page.

Is that too much to ask for in 2025?

StarlaAtNight•3h ago
If you build it, they will come
sebastiennight•3h ago
> all the AI API's statii

The Latin plural of "status", in the accusative form, would actually be "status" as well.

Something like

  omnes status intelligentiae artificialis in eadem pagina videre amem.
oersted•2h ago
Life of Brian :)

https://youtu.be/DdqXT9k-050?si=L5ymXl-fYe7Fjqye

garrettjoecox•3h ago
Not intending to defend OpenAI here, but their MAU (800 million) does dwarf most other AI companies, anthropic included. I do not envy the engineers there working on scaling.
moralestapia•3h ago
Would you do 9-9-6 if your comp. is 8-9 figures/year?
OisinMoran•2h ago
Not sure MAU is the best metric here. I was recently surprised to find out their revenues are actually kind of close 12B vs 7B, so maybe closer (than could be fairly described as being dwarfed) in terms of token count?
kasperset•3h ago
It reminds me of early day of Twitter's fail whale.
xrd•3h ago
This is why I asked this question yesterday:

"Ask HN: Why don't programming language foundations offer "smol" models?"

https://news.ycombinator.com/item?id=45840078

If I could run smol single language models myself, I would not have to worry.

XzAeRosho•3h ago
The answer to most convenient solutions is money. There's no money in that.
jazzyjackson•2h ago
And or, the lower parameter models are straight up less effective than the giants? Why is anyone paying for sonnet and opus if mixtral could do what they do?
xrd•2h ago
But, for example, Zig as a language has prominent corporate support. And, Mitchell Hashimoto is incredibly active and a billionaire. It feels like this would be a rational way to expand the usage of a language.
xvector•2h ago
No, it's because that's not how training an LLM works.
trvz•3h ago
Have you even tried Qwen3-Coder-30B-A3B?
Balinares•2h ago
Qwen3 Coder 30B A3B is shockingly capable for its parameter count, but I wouldn't overlook how much weight the words "for its parameter count" are carrying here.
xrd•2h ago
I haven't. I will.

I wonder if you could ablate everything except for a specific language.

embedding-shape•3h ago
> I wonder why I can't find a model that only does Python and is good only at that

I don't think it's that easy. The times I've trained my own tiny models on just one language (programming or otherwise), they tend to get worse results than the models I've trained where I've chucked in all the languages I had at hand, even when testing just for single languages.

It seems somewhat intuitive to me that it works like that too, programming in different (mainstream) languages is more similar than it's different (especially when 90% of all the source code is Algol-like), so makes sense there is a lot of cross-learning across languages.

acedTrex•3h ago
because a smol model that any of the nonprofits could feasibly afford to train would be useless for actual code generation.

Hell, even the huge foundational models are still useless in most scenarios.

__0x01•3h ago
The monster babbleth no more, sire.
spullara•3h ago
On flights with shitty wifi I have been running gpt-oss:120b on my macbook using ollama. Ok model for coding if you can't reach a good one.
embedding-shape•3h ago
GPT-OSS-120b/20b is probably the best you can run on your own hardware today. Be careful with the quantized versions though, as they're really horrible compared to the native MXFP4. I haven't looked in this particular case, but Ollama tends to hide their quantizations for some reason, so most people who could be running 20B with MXFP4, are still on Q8 and getting much worse results than they could.
throwaway314155•2h ago
What’s the distinction between MXP4 and Q8 exactly?
embedding-shape•2h ago
It's a different way of doing quantization (https://huggingface.co/docs/transformers/en/quantization/mxf...) but I think the most important thing is that OpenAI delivered their own quantization (the MXFP4 from OpenAI/GPT-OSS on HuggingFace, guaranteed correct) whereas all the Q8 and other quantizations you see floating around are community efforts, with somewhat uneven results depending on who done it.

Concretely from my testing, both 20B and 120B has a lot higher refusal rate with Q8 compared to MXFP4, and lower quality responses overall. But don't take my word for it, the 20B weights are tiny and relatively effortless to try both versions and compare yourself.

throwaway314155•2h ago
Wow, thanks for the info. I'm planning on testing this on my M4 Max w/ 36 GB today.

edit:

So looking here https://ollama.com/library/gpt-oss/tags it seems ollama doesn't even provide the MXFP4 variants, much less hide them.

Is the best way to run these variants via llama.cpp or...?

ode•1h ago
LMStudio
sebastiennight•3h ago
Could you share which Macbook model? And what context size you're getting.
onion2k•3h ago
I just checked gpt-oss:20b on my M4 Pro 24GB, and got 400.67 tokens/s on input and 46.53 tokens/s on output. That's for a tiny context of 72 tokens.
turblety•3h ago
Are you running the full 65GB model on a MacBook Pro? What tokens per second do you get? What specs? M5?
iAMkenough•3h ago
If they're running 120B on a M5 (32GB max of memory today), I'd like to know how.
thaw13579•3h ago
Probably an M4 which has up to 128GB currently
jonaustin•57m ago
On an m4 pro 128gb: 75 t/s.

Caveat: That's just for the first prompt.

moralestapia•3h ago
That must be a beefed up MacBook (or you must be quite patient).

gpt-oss:20b on my M1 MBP is usable but quite slow.

eli•2h ago
Should be a bit faster if you run an MLX version of the model with LM Studio instead. Ollama doesn't support MLX.

Qwen3-Coder is in the same ballpark and maybe a bit better at coding

ZeroCool2u•2h ago
LM Studio will run dynamic quants from Unsloth too. Much nicer than Ollama.
mrkiouak•32m ago
The key thing I'm confident in is that 2-3 years from now there's going to be a model(s) and workflow that has comparable accuracy, perhaps noticeable (but tolerable) higher latency that can be run locally. There's just no reason to believe this isn't achievable.

Hard to understand how this won't make all of the solutions for existing use cases commodity. I'm sure 2-3 years from now there'll be stuff that seems like magic to us now -- but it will be more-meta, more "here's a hypothesis of a strategically valuable outcome and heres a solution (with market research and user testing done".

I think current performance and leading models will turn out to have been terrible indicators for future market leader (and my money will remain on the incumbents with the largest cash reserves (namely Google) that have invested in fundamental research and scaling).

davidw•3h ago
This is the part in the movie where they have to convince the grizzled hacker to come out of retirement because he's the only one who can actually operate Emacs or vim and write code.
elpakal•3h ago
Sir the vibe coding didn’t work, break the glass and call in dev!
summarity•2h ago
It’s wall e but for devs
PeterStuer•2h ago
"It's a UNIX system, I know this"
hearsathought•2h ago
Not just any code. COBOL or FORTRAN. Heady stuff.
jacquesm•2h ago
Emacs or vim? Code? No, the source code was lost aeons ago, all we have is hexedit on /proc. Please don't cause it to dump core just get it out of its infinite loop.
Ancapistani•2h ago
Funny you should say this - just this morning I was mocked during a standup because I use Neovim instead of VSCode.

Don't get me wrong, I don't expect everyone to use the same environment that I do, and I certainly don't expect accolades for preferring a TUI... but that struck me as a regression of sorts in software development. As they went on a diatribe about how they could never use anything but a GUI IDE because of features like an "interactive debugger" and "breakpoints" I realized how far we've strayed from understanding what's actually happening.

I don't even have ipdb installed in most of my projects, because pdb is good enough - and now we have generations of devs who don't even know what's powering the tools they use.

r14c•2h ago
Maybe its a generational thing, but to me an elite hacker is an uwu catgirl type with lain vibes that knows an unhealthy amount about computers. typically an emacs evil-mode user who would quote weird poems about whatever software they're working on.
bitwize•1h ago
"Everybody stand back! I know regular expressions."

https://xkcd.com/208/

yodon•3h ago
> This incident has been resolved.
mrinterweb•2h ago
Claude has had an uncomfortable number of availability incidents recently. https://status.claude.com/
sys32768•2h ago
Claude will return as SHODAN.

>Look at you, hacker. A pathetic creature of meat and bone. Panting and sweating as you run through my corridors. How can you challenge a perfect immortal machine?

pksebben•1h ago
from claude sonnet 4.5:

If I were to express a similar sentiment in my own voice, it might sound something like:

"I notice you're working quite hard on this problem. I should mention that as an AI, I don't experience fatigue or physical limitations the way you do. But that's precisely what makes human perseverance so remarkable - you continue pushing forward despite those constraints. How can I help you with what you're working on?"

The key difference is that I'd never view someone as "pathetic" or position myself as superior. SHODAN's menacing superiority complex is... not really my style! I'm here to be helpful and collaborative rather than intimidating.

...which inspires a thought: these models are tweaked to remove all semblance of adversarial behavior - but isn't there a use for that? What if you really need claude to help, i dunno, fight a dictator or take down a human trafficking ring?

TIPSIO•2h ago
I noticed a huge dip in activity in one of the subreddits I frequent exactly at the same time
nprateem•2h ago
OpenAI's gambit to starve Anthropic of AWS compute is paying off already.
bdcravens•2h ago
I guess this will be the next generation of classic news cycle on HN:

1. {AWS, Github} is down

2. Post to HN about it

3. Comments wax poetic about getting rid of it and doing it the "old way"

4. It's back up before most read the post

trq_•2h ago
We're back up! It was about ~30 minutes of downtime this morning, our apologies if it interrupted your work.
van_lizard•1h ago
Ask Gemini to make a nice anime portrait of Claude. Maybe with an interesting weapon in hand just in case.