frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

A California Home That Rotates 320 Degrees Takes a Spin on the Market

https://www.wsj.com/real-estate/luxury-homes/rotating-home-novato-california-998d0db9
1•fortran77•56s ago•1 comments

AI spend observability down to individual User ID's and routes

https://kalaa.cc/
1•Francis221•1m ago•0 comments

OpenSSH 10.5 released, AI fixes now welcome

https://www.openssh.org/releasenotes.html#10.5
1•voxadam•2m ago•0 comments

The Pacing of the Frontier

https://thezvi.substack.com/p/the-pacing-of-the-frontier
1•paulpauper•3m ago•0 comments

Show HN: Cjson.cpp, the new fastest-ever (constexpr) JSON parser

https://github.com/rfgplk/cjson.cpp
2•rfgplk•3m ago•0 comments

The Housing Recession Is Over

https://conorsen.substack.com/p/the-housing-recession-is-over-8b6
2•paulpauper•3m ago•0 comments

Expired DMARC endpoint exposed a NYSE Fortune 1000 infrastructure for $10

https://www.sh.consulting/blog/abandoned-dmarc-reporting-domain
1•alexshakhov•4m ago•0 comments

The Unsloth Desktop App – run and train models locally

https://unsloth.ai/download/mac
1•bogzz•4m ago•0 comments

Eight architecture myths BUSTED by Samuel Hughes

https://www.bensouthwood.co.uk/p/eight-architecture-myths-busted-by
1•paulpauper•4m ago•0 comments

The SpaceX Sham

https://dissentmagazine.org/online_articles/spacex-ipo-elon-musk-trillionaire/
2•cdrnsf•4m ago•0 comments

TRMNL ePaper Design System

https://github.com/usetrmnl/trmnl-framework
1•Bluestein•5m ago•0 comments

Remove Claude's Watermark

https://removeclaudewatermark.org/
1•macinjosh•6m ago•2 comments

Up to 12.8x Improvement Observed for Gup_test with New Linux MM Patches

https://www.phoronix.com/news/Linux-Batch-Follow-Page-Mask
1•Bender•6m ago•0 comments

Flatpak 1.19 Released with Nine Security Fixes

https://www.phoronix.com/news/Flatpak-1.19-Released
3•Bender•7m ago•0 comments

Analyzing Fedora's Slow Performance on the Framework Laptop 13 Pro

https://www.phoronix.com/review/fedora-pantherlake-thermald-tuned
1•Bender•7m ago•0 comments

Rant/Help Google auto-account bans are insane

https://accounts.google.com
3•howdyhowdy•7m ago•2 comments

China warns of "security backdoor" in Anthropic AI coding tool [July 8, 2026]

https://www.cbsnews.com/news/china-security-backdoor-anthropic-ai-coding-tool/
3•tiahura•8m ago•1 comments

Deep-diff NPM JavaScript library lost reference documentation

https://github.com/therohk/datum-merge/blob/main/src/diff-lib/dd-README.md
1•arzan303•8m ago•0 comments

Ask HN: What are your thoughts on E2E software development using AI agents?

1•sandiprghane•8m ago•0 comments

Phone tracking adding to license plate readers, associating devices with cars

https://techxplore.com/news/2026-08-tech-tracking-plate-readers-associating.html
1•mdp2021•9m ago•0 comments

Moving integer division to floating-point is trivial

http://marc-b-reynolds.github.io/math/2026/08/10/IntDivByFP.html
1•fanf2•9m ago•0 comments

Experience of Running Muse Glimmer on Mac

https://holaclaw.ai/blog/muse-glimmer-on-mac
3•ridruejo•9m ago•0 comments

Ask HN: Are third party gguf/mmproj safe on Llama in production env

2•skye0110•10m ago•0 comments

A compiler that turns a day at my computer into agent memory, no LLM

https://github.com/nossa-y/activity-frames
1•nossa-y•10m ago•0 comments

Terminal Browser

https://terminal-browser.com/
1•navigaid•11m ago•0 comments

Tradecraft: A portable format for context about detections

https://github.com/spectrum-sec/delta-tradecraft
1•DecadeSol•11m ago•0 comments

Linux desktop use surged to 22% on one workday, Cloudflare data shows

https://www.zdnet.com/article/linux-desktop-use-surged-on-one-workday-cloudflare-data-shows/
3•CrankyBear•12m ago•0 comments

Coffee Camp cattle farmer plants 14,000 trees for flood mitigation

https://www.abc.net.au/news/2026-07-10/nature-based-solutions-flooding/106844226
1•speckx•13m ago•0 comments

Possibly Resolving a Conjecture on Matrix Theory with GPT Sol

https://chillphysicsenjoyer.substack.com/p/i-possibly-resolved-a-conjecture
3•crescit_eundo•14m ago•0 comments

Spotify to Label AI 'Artists'

https://www.theguardian.com/technology/2026/aug/11/spotify-label-ai-artists-block-them-from-some-...
2•bokkies•15m ago•0 comments
Open in hackernews

Can Claude Code in a loop improve an enterprise AI agent with $10,745 of budget?

https://jeremytian.substack.com/p/can-claude-code-in-a-loop-improve
5•jeremytian•55m ago

Comments

AbhinavX•53m ago
>An interesting observation we made was that even though the 3 optimizers were given tens of thousands of dollars in compute, they found the best solution very early on

Honestly, this confirms some bias I already have from coding — models can get to the best solution pretty fast, but if the solution is something obscure, they usually don’t find it.

jeremytian•39m ago
From my experience, models that drive optimizers in a loop often get stuck in local minima. ofc, this is just correlation, and not truly indicative, but if you look at the first graph, analyzing precision vs cost, all three model-driven optimizers just bounce back and forth between a precision range, not really expanding the horizon or getting too much worse than the orignal. from our analysis on this data, model-driven optimizers typically don't start from scratch again, they just continue down their optimization path.
andy0liang•52m ago
It would be interesting to see how a human researcher would do on the same task and same budget. 2 weeks of salary + a Claude Code max plan might outperform all of these auto-optimizers in a loop.
jeremytian•37m ago
this would be an interesting experiment but i think one of the main benefits of having at least some automated pipeline is that human iteration is quite slow since the information intake speed is a few magnitudes slower. probably some combination would be optimal, like ensuring alignment of objectives at the beginning and tuning along the way