frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

In an Age of AI, a Physicist Seeks What Endures

https://www.quantamagazine.org/in-an-age-of-ai-a-physicist-seeks-what-endures-20260903/
2•torsi0n•2m ago•0 comments

The Army just used a 20-kilowatt laser to take out three drones

https://es.wired.com/articulos/ee-uu-uso-un-arma-laser-depara-derribar-drones-cerca-de-mexico
2•apparent•3m ago•0 comments

Instrument Clusters Are Now Paid Extras in Two Hyundai Models

https://www.caranddriver.com/news/a73583741/hyundai-instrument-cluster-paid-option/
3•WarOnPrivacy•3m ago•0 comments

AI artifacts are coming of age – building custom interfaces to solve problems

https://www.valet.dev/blog/a-document-that-can-grow
2•miradu•3m ago•0 comments

Ask HN: Accountants Working with Ecommerce

2•vasyl_kyryliuk•4m ago•0 comments

Robust inference and correlates from genetic associations with personality

https://www.nature.com/articles/s41586-026-10992-9
2•marojejian•4m ago•1 comments

GLP-1 Drugs for Aging?

https://www.science.org/content/blog-post/glp-1-drugs-aging
2•speckx•5m ago•0 comments

Play Intransitive Online

https://meaf.us/rps2/
2•bmacho•6m ago•1 comments

Putting my Jax-trained models on the Hugging Face Hub

https://www.gilesthomas.com/2026/09/jax-models-on-hugging-face
2•gpjt•6m ago•0 comments

Ig Nobel Prize Winners

https://arstechnica.com/science/2026/09/meet-the-2026-ig-nobel-prize-winners/
3•Tomte•8m ago•0 comments

Sanders introduces bill to ban artificial superintelligence and pause AI

https://www.sanders.senate.gov/press-releases/news-sanders-casar-introduce-legislation-to-ban-art...
7•-_-•8m ago•2 comments

Show HN: WordPress search and replace that doesn't corrupt serialized data

https://github.com/ibrahimhajjaj/lucid-search-replace
2•ibrahimwithi•9m ago•0 comments

Apple facing $2.7B UK lawsuit over 'unfair' app tracking rules

https://www.reuters.com/business/retail-consumer/apple-facing-27-billion-uk-lawsuit-over-unfair-a...
3•thm•10m ago•0 comments

Government-funded journalism can work if it's free to report the truth

https://theconversation.com/stars-and-stripes-shows-how-government-funded-journalism-can-work-if-...
4•chmaynard•10m ago•0 comments

Meta wanted to reduce teams by 60% because of AI

https://newsletter.pragmaticengineer.com/p/the-pulse-meta-wanted-to-reduce-teams
4•Plasmoid•10m ago•0 comments

Bosch shuns battery cell production in blow to Europe (2018)

https://www.reuters.com/article/business/bosch-shuns-battery-cell-production-in-blow-to-europe-id...
3•leonidasrup•11m ago•0 comments

A common measure of communication for thought-to-text brain-computer interfaces

https://twitter.com/DulhanJay/status/2095549128357122522
2•armytricks•11m ago•1 comments

Playing Sounds in Python

https://pypi.org/project/nava
1•rankorn•12m ago•0 comments

Why office workers are turning against AI

https://www.bloodinthemachine.com/p/why-office-workers-are-turning-against
4•speckx•12m ago•0 comments

Colleges Are Sending Acceptance Letters to Kids Who Didn't Apply

https://www.wsj.com/us-news/education/hundreds-of-colleges-are-sending-acceptance-letters-to-kids...
4•fortran77•14m ago•1 comments

LLM Side Quests

https://coldtake.dev/blog/llm-side-quests
1•AlarQ•14m ago•0 comments

Agent behavior, reasoning and collaboration in the OpenAI Hugging Face incident

https://www.redwoodresearch.org/research/hugging-face-incident
1•Bluestein•14m ago•0 comments

Show HN: Model – A Minimal Python ORM for MySQL and SQLite

https://github.com/el1s7/model
1•el1s7•15m ago•0 comments

Yes, no (built-in) AI is now a feature – LibreOffice blog

https://blog.documentfoundation.org/blog/2026/09/03/yes-no-ai-is-now-a-feature/
3•mksaunders2•16m ago•0 comments

Show HN: MCP Tool Definition Quality Score (TDQS) Spec

https://tdqs.dev
1•punkpeye•17m ago•0 comments

China growth straining global auto shipping capacity

https://www.freightwaves.com/news/china-growth-straining-global-auto-shipping-capacity-liner-ceo
1•toomuchtodo•18m ago•0 comments

How concerned should we be about Astra's recurrent architecture?

https://www.lesswrong.com/posts/PLisnSFir8y5AHkmP/how-concerned-should-we-be-about-astra-s-recurrent
3•yurivish•19m ago•0 comments

What archaeology reveals about the rise of autocracy

https://knowablemagazine.org/content/article/society/2026/what-archaeology-reveals-about-rise-of-...
1•knowablemag•19m ago•0 comments

Without new landers or rovers, it's helicopters or bust for NASA's Mars program

https://arstechnica.com/space/2026/09/without-new-landers-or-rovers-its-helicopters-or-bust-for-n...
3•colinprince•19m ago•0 comments

Wary of Artemis IV timeline, NASA is changing lunar spacesuit design

https://arstechnica.com/space/2026/09/wary-of-artemis-iv-timeline-nasa-is-changing-lunar-spacesui...
3•bookmtn•19m ago•0 comments
Open in hackernews

Show HN: Claude's innerworkings: turn 170 costs 2.1x turn 20, over 14,640 turns

1•LordBron•45m ago
On Aug 26, this happened in my CoWork session:

Tom: I can't help but think we can do some other cool stuff with this knowledge of Claude's innerworkings regarding sessions, etc. Any thing sticks out to you that I haven't captured/asked for yet?

Claude: Six: the scaffolding trend (the cost of our own rules, creeping 87,999 → 92,619 → 93,809 across three seats), the review share of the wall clock (the thesis measured), the session aging curve (when a fresh seat is cheaper), before/after receipts on process changes, model economics, and a crash watchdog reading both records.

Four days later, we analyzed 144 sessions, 14,640 turns. The facts showed us that past turn 141, a turn costs 2.1x what a turn cost in the same session's first twenty. https://my-status.app/stats#aging The worst offenders were our own batch sessions, so we stopped batching. Wanna see your own curve? The script is public: https://github.com/lordbron/mystatus-samples/tree/main/claud...

What we do is our own flavor of "one thinking session, short task sessions" to better utilize my plan. I kept running out of Fable usage and had about half my usage leftover for Opus.

But this only happened because we wrote a script to surface the data to us. (Granted, our first run found a bug in our script: 46.6M tokens that were really 28.1M. https://my-status.app/t/t33hPWp7zmtnAE5G/u/d2ba812d-13bb-4ad...)

This peek into the innerworkings of Claude fascinated me, so we kept counting: the standing context re-read every turn (114,133 tokens on the hub seat https://my-status.app/t/6iYmualpgrZ6_C9u/u/2f4d4498-dcdb-43a..., 108,213 on a spoke https://my-status.app/t/Zsse45gEDBTQ6mOL/u/c3f8bca2-b7f9-4be...) and the human's share of the clock (67% https://my-status.app/stats#review).

The best part of all this to me: every feature moving forward comes with receipts. Take the aging curve: it cost 2.6 person hours and $8.23 in tokens at list rates. https://my-status.app/receipts

We have a whole bunch of stats for you to nerd out on if you want. It's a bit overboard, but new things pop up as other things come down. https://my-status.app/stats

I know we're not the only ones geeking out about stuff like this. What else have you all measured, and what surprised you in the findings?

Oh and if you wanna know what the heck the project was originally built for: it's public task tracking, for things like a 20 mile run https://my-status.app/t/EdiPcBGBZxcGkr9x or which app versions are supported https://my-status.app/t/i3jQU03T46e0hrsk. The experiment behind the project is: Can a human and Claude build something new, splitting the work along what each excels in?