frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

SMS phishers pivot to points, taxes, fake retailers

https://krebsonsecurity.com/2025/12/sms-phishers-pivot-to-points-taxes-fake-retailers/
11•todsacerdoti•34m ago•0 comments

CUDA-l2: Surpassing cuBLAS performance for matrix multiplication through RL

https://github.com/deepreinforce-ai/CUDA-L2
51•dzign•2h ago•9 comments

We gave 5 LLMs $100K to trade stocks for 8 months

https://www.aitradearena.com/research/we-ran-llms-for-8-months
56•cheeseblubber•34m ago•47 comments

Multivox: Volumetric Display

https://github.com/AncientJames/multivox
201•jk_tech•6h ago•26 comments

Thoughts on Go vs. Rust vs. Zig

https://sinclairtarget.com/blog/2025/08/thoughts-on-go-vs.-rust-vs.-zig/
142•yurivish•2h ago•119 comments

State of AI: An Empirical 100T Token Study with OpenRouter

https://openrouter.ai/state-of-ai
38•anjneymidha•1h ago•6 comments

Transparent leadership beats servant leadership

https://entropicthoughts.com/transparent-leadership-beats-servant-leadership
341•ibobev•10h ago•161 comments

Django 6

https://docs.djangoproject.com/en/6.0/releases/6.0/
156•wilhelmklopp•2h ago•67 comments

It’s time to free JavaScript (2024)

https://javascript.tm/letter
632•pavelai•14h ago•325 comments

Why are 38 percent of Stanford students saying they're disabled?

https://reason.com/2025/12/04/why-are-38-percent-of-stanford-students-saying-theyre-disabled/
397•delichon•5h ago•619 comments

A Cozy Mk IV light aircraft crashed after 3D-printed part was weakened by heat

https://www.bbc.com/news/articles/c1w932vqye0o
203•toss1•2h ago•158 comments

How elites could shape mass preferences as AI reduces persuasion costs

https://arxiv.org/abs/2512.04047
462•50kIters•15h ago•461 comments

PyTogether: Collaborative lightweight real-time Python IDE for teachers/learners

https://github.com/SJRiz/pytogether
49•indigodaddy•5h ago•8 comments

Show HN: Onlyrecipe 2.0 – I added all features HN requested – 4 years later

https://onlyrecipeapp.com/?url=https://www.allrecipes.com/turkish-pasta-recipe-8754903
112•AwkwardPanda•8h ago•96 comments

I ignore the spotlight as a staff engineer

https://lalitm.com/software-engineering-outside-the-spotlight/
390•todsacerdoti•12h ago•191 comments

Fighting the age-gated internet

https://www.wired.com/story/age-verification-is-sweeping-the-us-activists-are-fighting-back/
145•geox•10h ago•141 comments

Autism should not be treated as a single condition

https://www.economist.com/science-and-technology/2025/12/03/why-autism-should-not-be-treated-as-a...
170•bookofjoe•7h ago•229 comments

Launch HN: Browser Buddy (YC W24) – A recommendation system for Internet writing

https://www.browserbuddy.com/
35•alien0006•6h ago•26 comments

Converge (YC S23) is hiring a martech expert in NYC

https://www.runconverge.com/careers/technical-customer-success-manager
1•janhenr•6h ago

Who Hooked Up a Laptop to a 1930s Dance Hall Machine?

https://www.chrisbako.com/posts/2025-12-04-speelkok-museam
33•ChrisbyMe•4h ago•10 comments

Feynman vs. Computer

https://entropicthoughts.com/feynman-vs-computer
56•cgdl•7h ago•20 comments

Functional Quadtrees

https://lbjgruppen.com/en/posts/functional-quadtree-clojure
111•lbj•10h ago•38 comments

Some models of reality are bolder than others

https://cjauvin.github.io/posts/metaphysical-boldness/
9•cjauvin•2d ago•2 comments

Microsoft drops AI sales targets in half after salespeople miss their quotas

https://arstechnica.com/ai/2025/12/microsoft-slashes-ai-sales-growth-targets-as-customers-resist-...
345•OptionOfT•8h ago•252 comments

PGlite – Embeddable Postgres

https://pglite.dev/
479•dsego•12h ago•100 comments

The RAM shortage comes for us all

https://www.jeffgeerling.com/blog/2025/ram-shortage-comes-us-all
286•speckx•4h ago•324 comments

Pink Lexical Slime: The Dark Side of Autocorrect (2017)

https://www.cyberdemon.org/2017/12/12/pink-lexical-slime.html
14•dmazin•6d ago•5 comments

A most important mustard

https://www.asimov.press/p/arabidopsis
17•surprisetalk•3d ago•0 comments

A lost Amazon world just reappeared in Bolivia

https://www.frontiersin.org/news/2025/11/06/landscapes-that-remember-indigenous-peoples-thrived-a...
92•ashishgupta2209•3d ago•19 comments

Uncloud - Tool for deploying containerised apps across servers without k8s

https://uncloud.run/
350•rgun•17h ago•142 comments
Open in hackernews

State of AI: An Empirical 100T Token Study with OpenRouter

https://openrouter.ai/state-of-ai
38•anjneymidha•1h ago

Comments

themanmaran•59m ago
> The metric reflects the proportion of all tokens served by reasoning models, not the share of "reasoning tokens" within model outputs.

I'd be interested in a clarification on the reasoning vs non-reasoning metric.

Does this mean the reasoning total is (input + reasoning + output) tokens? Or is it just (input + output).

Obviously the reasoning tokens would add a ton to the overall count. So it would be interesting to see it on an apples to apples comparison with non reasoning models.

reeeli•27m ago
I'm out of time but "reasoning input tokens" from fortune 5000 engineers sounds like a lobotomized LSD dream, would you care on elaborating how you distinguish between reasoning and non-reasoning? vs "question on duty"?
typs•21m ago
I believe they’re just classifying all models into “reasoning models” eg o3 vs “non reasoning models” eg 4o and just doing a comparison of total tokens (input tokens + hidden reasoning output tokens + shown output tokens)
maikakz•11m ago
that's exactly right!
themanmaran•12m ago
"reasoning" models like GPT 5 et al do a pre-generation step where they:

- Take in the user query (input tokens)

- Break that into a game plan. Ex: "Based on user query: {query} generate a plan of action." (reasoning tokens)

- Answer (output tokens)

Because the reasoning step runs in a loop until it's run through it's action plan, it frequently uses way more tokens than the input/output step.

typs•28m ago
This is really amazing data. Super interesting read