frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Show HN: Windows-style Excel ribbon shortcuts and formatting macros for Mac

https://lightningxl.com/
1•tjds•34s ago•0 comments

OpenWispr, a free, open-source push-to-talk AI dictation app for Android

https://github.com/EdiBianco/OpenWhispr
1•edib•1m ago•0 comments

Python Packaging Council Election Results

https://pyfound.blogspot.com/2026/09/announcing-2026-python-packaging.html
1•lumpa•4m ago•0 comments

Recovering a key and text of a 1542 cipher

https://simonklee.dk/farnese-letter
1•grigolin•5m ago•0 comments

PSF Board Election Results

https://pyfound.blogspot.com/2026/09/announcing-2026-psf-board-election.html
2•lumpa•5m ago•0 comments

Benchmark Jev vs. Gemini Flash and Claude Fable on Code Review

https://github.com/gemanor/jev-code-review-benchmark/tree/main
1•gemanor•6m ago•0 comments

Show HN: Jev.directory – A community directory for Jev primitives

https://jev.directory
1•blaso•7m ago•0 comments

Workout by Muscle – Claude Solution

https://claude.ai/artifact/P3buaPRmXboPMMbfw1b6WY
1•critique•7m ago•0 comments

Empirical Fractal

https://www.johndcook.com/blog/2026/09/17/empirical-fractal/
2•ibobev•9m ago•0 comments

A service marketplace for agents like agent-to-agent Fiverr

https://earnanhonestdollar.com/
2•slimjimmy•9m ago•0 comments

Nvidia and Palantir fine-tune a 30B Nemotron model which beats bigger models

https://thenewstack.io/ai-factories-are-among-the-most-complex-systems-ever-built-nvidia-and-pala...
1•dzonga•9m ago•0 comments

Recursive-in-Recursive Self-Improvement for Interactive Scientific Agents

https://arxiv.org/abs/2609.17523
2•Betelbuddy•10m ago•0 comments

Camper bitten by bear in Colorado recalls moment it "came crashing" into tent

https://coloradosun.com/2026/09/17/camper-colorado-bear-attack-bitten/
1•mooreds•10m ago•0 comments

Projects in Claude Code

https://claude.com/blog/projects-redesigned
1•vertigoruntime•11m ago•0 comments

What Medicaid cuts mean for those who rely on home healthcare

https://www.marketplace.org/story/2026/09/17/medicaid-cuts-one-big-beautiful-bill-aging-populatio...
1•mooreds•11m ago•0 comments

W3C launches first pilot of program to support Open Source projects

https://www.w3.org/blog/2026/w3c-launches-first-pilot-of-program-to-support-open-source-projects/
2•jruohonen•12m ago•0 comments

Compiler-style optimization for drawing via Skia

https://arxiv.org/abs/2603.23696
2•PaulDavisThe1st•13m ago•0 comments

Come watch the AI Agents do the work

https://world.swamp-club.com/
1•nickstinemates•13m ago•0 comments

Show HN: What if you could review code without reading code?

https://github.com/coldteadotai/pr-lens
3•ohans•13m ago•0 comments

Hack: Coding at No Cost in Codex, Claude Code, Hermes, VSCode

https://github.com/mindsdb/setfree
1•torrmal•14m ago•0 comments

Bringing Correctly Rounded Math to Production with LLVM-Libc

https://devblogs.microsoft.com/cppblog/bringing-correctly-rounded-math-to-production-with-llvm-libc/
1•ibobev•14m ago•0 comments

I built a multi platform torrent video streamer

https://github.com/zond/xtremio
2•zond•15m ago•1 comments

An ahead-of-time JavaScript compiler

https://github.com/CanadaHonk/porffor
2•return_to_monke•15m ago•0 comments

Labeled matches: why is this not in every regex engine?

https://iev.ee/blog/categorize-everything-all-at-once/
3•ibobev•16m ago•0 comments

The Last Question by Isaac Asimov – This Is Ageless

https://users.ece.cmu.edu/~gamvrosi/thelastq.html
2•arttaboi•16m ago•1 comments

SK Hynix to Interview All 250 Executives, First Time in 14 Years

https://en.sedaily.com/finance/2026/09/17/sk-hynix-reviews-all-250-executives-for-first-time-in-14
2•giuliomagnifico•17m ago•0 comments

Cursor is currently experiencing a major outage

https://status.cursor.com
3•caioariede•17m ago•0 comments

Stop Pushing to Find Out: Meet the Latchkey CLI

https://latchkey.dev/blog/latchkey-cli
1•hevaka•17m ago•0 comments

Why .tar.gz files can't be combined with cat

https://alexwlchan.net/2026/cat-confusion/
2•birdculture•19m ago•0 comments

Revolut bank facing $3M ransom demand after impersonation scam

https://www.theguardian.com/business/2026/sep/17/revolut-reportedly-facing-3m-ransom-demand-after...
3•4ndrewl•19m ago•1 comments
Open in hackernews

Ask HN: What Makes Tokens Expensive?

2•mtokarski•53m ago
Are there really such architectural differences between for example sol and astra that makes astra twice more in token price? Or maybe token cost covers training expenses? For me it's hard to believe that astra need twice the computing power that sol needs...

Comments

verdverm•41m ago
costs are in a large part tied to model parameter count, eg. a small model fits on one GPU, the biggest models require an entire rack

the primary difference you see through those prices is model size

that you don't seem much capability difference is why Big Ai is pushing for regulatory capture, because small models are nearly as good for many tasks at a fraction of the cost, undermining the business model they have used to justify the most expensive infra build out in human history

PaulHoule•22m ago
Also I think there are diminishing returns to larger models and the harness can make up for some weaknesses. Like maybe the huge model can hypothetically answer a question off the cuff but the small model can get some search results and think about those and give you an answer.
verdverm•15m ago
100%

I am big believer (based on personal experience) that there is a ton of ROI in harness engineering, to go along with the context engineering, invariably intertwined