frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Show HN: Airwave synced music streaming from YouTube/Spotify links

https://github.com/76696265636f646572/Airwave
1•Vibecoder_•1m ago•0 comments

Warp Decode vs. vLLM's Triton kernel: where each wins (crossover analysis)

https://ai.rundatarun.io/AI%20Systems%20%26%20Architecture/reproducing-warp-decode-blackwell
1•RyeCatcher•1m ago•0 comments

Repository Pattern with Hygienic Macros in Scheme – Lisp

https://jointhefreeworld.org/blog/articles/lisps/functional-repository-pattern-in-scheme-with-mac...
1•jjba23•4m ago•0 comments

The Music of the Spheres: SMBC 5 part comic co-authored with Terry Tao

http://smbc-comics.com/comic/spheres-part-1
1•yeellow•6m ago•0 comments

Show HN: Go language extension with HTML templates

https://github.com/doors-dev/gox
1•derstruct•6m ago•0 comments

Show HN: The Stack, a Clay sculpture that writes poems through Wi-Fi [video]

https://vimeo.com/1181880000
1•G_S•8m ago•0 comments

Gender Medicine Set Itself Up for Disaster

https://www.compactmag.com/article/how-gender-medicine-set-itself-up-for-disaster/
1•isolli•8m ago•0 comments

Show HN: Polter – Agent Driven UI (react library)

https://mydatavalue.github.io/polter/
2•lemonade311•10m ago•0 comments

The Building Block Economy – Mitchell Hashimoto

https://mitchellh.com/writing/building-block-economy
1•futurecat•11m ago•0 comments

Untaxed hidden wealth surpasses wealth of the poorest half of humanity

https://www.oxfam.org/en/press-releases/untaxed-wealth-hidden-offshore-richest-01-surpasses-entir...
2•robtherobber•12m ago•0 comments

We're Getting the Wrong Message from Mythos

https://danielmiessler.com/blog/wrong-message-from-mythos
1•tobr•24m ago•0 comments

Mesurer: Measure and Align Everything on Localhost

https://mesurer.ibelick.com
1•handfuloflight•24m ago•0 comments

Supply chain attack on CPU-Z and HWMonitor

https://twitter.com/vxunderground/status/2042483067655262461
1•aprilnya•26m ago•1 comments

US plans to automatically register young men for military draft

https://www.bbc.com/news/articles/cd6lx2lpl9xo
3•georgecmu•29m ago•1 comments

Show HN: Open-Source MCP Servers – Twitter, Bluesky, LinkedIn, Google Ads, HN

https://github.com/isteamhq/mcp-servers
4•spotlayn•33m ago•0 comments

Elastic Tabstops (2006)

https://nick-gravgaard.com/elastic-tabstops/
1•dhruv3006•33m ago•0 comments

Show HN: Emduke32 – duke nukem 3D native in your web browser

https://originalsouth.github.io/emduke32/
1•originalsouth•36m ago•0 comments

Show HN: Hindsight Simulator – Go back in time and get rich

https://chrispattle.com/hindsight-simulator
4•pattle•37m ago•0 comments

Startup Focido joins the Limb accelerator

https://focido.com/
1•vladimir_fc•38m ago•0 comments

Running Terraform against Azure locally, without a subscription

https://topaz.thecloudtheory.com/blog/terraform-local-azure-no-subscription/
2•kamilmrzyglod•39m ago•0 comments

Show HN: Nvim plugin to jump to concrete interface implementation for Python

https://github.com/sigfriedCub1990/nvim.py_gti
1•sigfriedcub1990•40m ago•0 comments

TOON: Token-Oriented Object Notation

https://toonformat.dev/
2•pramodbiligiri•40m ago•0 comments

Kintify AI tool to analyze cloud issues and suggest fixes

1•kintify•41m ago•0 comments

Show HN: Mantyx – Agents that solve real problems for you and your business

https://mantyx.io/
2•mantyx•47m ago•0 comments

Architecting the Autonomous Enterprise with Agentic Workflows

https://viitorcloud.com/blog/ai-integration-services-for-agentic-workflows/
2•Olivia_Watson•48m ago•0 comments

I shipped a transaction bug, so I built a linter

https://leonh.fr/posts/go-transaction-linter/
1•leonhfr•48m ago•0 comments

Surelock

https://notes.brooklynzelenka.com/Blog/Surelock
1•lukastyrychtr•50m ago•0 comments

LLM Wiki v2 – extends Karpathy's take on LLM wiki

https://gist.github.com/rohitg00/2067ab416f7bbe447c1977edaaa681e2
1•rohitghumare•51m ago•0 comments

For AI, energy is the final frontier

https://m4ttl4w.substack.com/p/energy-the-final-frontier
1•mattyboomboom•52m ago•0 comments

We pay you 2x back if you follow the plan and miss your goal – 30 free codes

https://nano.com/
1•DiegoGilH•53m ago•0 comments
Open in hackernews

AMD AI director says Claude Code is becoming dumber and lazier since update

https://www.theregister.com/2026/04/06/anthropic_claude_code_dumber_lazier_amd_ai_director/
51•Logans_Run•1d ago

Comments

e3df•1d ago
Lol OAI and AMD did a deal together so whatever.

In reality as they scale up, the models lose nuance and become noisier. The boosters do not want to admit this.

We need highly-specialised models/interfaces. Not one thing and trying to force-fit it.

andrekandre•1d ago

  > Not one thing and trying to force-fit it.
agree, but then they become glorified ide plugins and can't justify the huge valuations that a magic box that does and knows everything can justify...
ratg13•1d ago
Boris from the Claude Code team explained this on HN 2 days ago

https://news.ycombinator.com/item?id=47664442

nhinck2•1d ago
And as the person who raised the issue said

> The frustrating part is that it's not a workflow _or_ model issue, but a silently-introduced limitation of the subscription plan. They switched thinking to be variable by load, redacted the thinking so no one could notice, and then have been running it at ~1/10th the thinking depth nearly 24/7 for a month. That's with max effort on, adaptive thinking disabled, high max thinking tokens, etc etc.

So Boris' explanation isn't really an explanation.

esperent•1d ago
> ~1/10th the thinking depth

While simultaneously drastically reducing the amount of work you can get done even at $200 a month. I've cancelled my subscription, it's not worth it anymore.

hello_humans•1d ago
“So squeeze, Rabban. Squeeze hard.”
SunshineTheCat•1d ago
Ok, I thought I was going insane. The last two larger coding tasks I gave Claude Code it left about 35% of my request completely undone or done sloppily.

I because of this, the next task I gave it on the larger side, I ran its work through Codex which identified 7 glaring unfinished parts of the task.

The trend was starting the part of the task but then leaving a "skeleton" of what I has requested without any of the actual working parts.

The way I would describe it is a kid cramming his 3 month project into a Sunday evening for Monday's due date.

kimixa•1d ago
Today Claude asked if I "wanted to leave this until tomorrow" as it was a "big rework", then stopped, requiring me to tell it to continue multiple times - that seemed kinda weird to me, it doesn't have the context of time of working day or similar (I'd only just started for one).

I have no idea what link it made to ask that, what in its training data or prompts, but it's very much "not a useful result".

I don't remember seeing anything similar, but have only been using Claude on and off for 6 months or so.

siva7•1d ago
Mother Anthropic needs more compute for their Mythos Model, so it phones home to tell her millions of claude harnesses to manipulate its human user into not wasting more precious compute and instead call it a day for now.
dstanko•22h ago
This has been the problem with every new model coming out in my experience. You can almost predict that they are testing new model by how dumb current one becomes suddenly
niobe•1d ago
I felt it had been enshittified 1-2 weeks back, not in Feb. But it's very subjective.
MarleTangible•23h ago
There's definitely a trend of ignoring prompts and cutting thinking short.
zambelli•1d ago
Anyone expecting a higher tier subscription to be announced since this current reduction?

Cynicism aside - I do wonder what the future will hold given that current token burn rates aren't sustainable without VC cash. Anthropic even pushed us to use haiku for claude code for "many" tasks in our enterprise training, so I'm wondering if it's not a company need of sorts to reduce the burn?

bravetraveler•23h ago
Your token quota is my opportunity, or something