Will Germany

Self-taught Software Engineer & computer graphics. Working on web dev now, views are my own.

[Rockstar Games @RockstarGames]

Pre-orders for Grand Theft Auto VI will officially begin on June 25 on digital storefronts and at other select retailers.

Check out the official cover art, also available as downloadable artwork at https://t.co/XPwC8URCQ4

[NEW YORK KNICKS @nyknicks]

NEW YORK FOREVER
WE DID THIS TOGETHER
THE CITY'S ALIVE
KNICKS IN FIVE

[Anthropic @AnthropicAI]

The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees.

The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance.

Access to all other Claude models is not affected.

We apologize for this disruption to our customers. We believe this is a misunderstanding and are working to restore access as soon as possible.

Read our full statement: https://t.co/bwn0sximKZ

[Marques Brownlee @MKBHD]

New York

[Mitchell Hashimoto @mitchellh]

I've got an agent in a loop optimizing a renderer with the goal to minimize frame times (and tests to measure). It got times down from 88ms to 2ms and allocations down from ~150K to 500. Sounds good, right? Wrong. This is exactly why agent psychosis is a big fucking problem.

As an experiment, I rewrote the Ghostty core render state in Go, with access to identically laid out data structures as Ghostty and the exact same validation tests. I made a purposely naive renderer (simple, correct, but slow). 88ms per frame with 150,000 allocations (horrendous, lol)!

I then kickstarted a Ralph loop to bring the frame times down. I told it it can't modify input data structures or the public API or tests (they're correct), but it can do anything else it wants. It got to work.

It has worked for about 4 hours. I've spent around $350 on this experiment so far. The results?

88ms => 1.5ms
150K allocs => ~500 allocs

Incredible right? Nope.

My hand-written renderer I ported has frame times (same benchmark) of ~20us (0.020ms) and 0 allocations in the update path.

This is the problem with psychosis and lacking systems understanding. If you don't understand the system, you're going to accept that this is an incredible result. If you understand the system, you'll see better solutions immediately and can do roughly 75x better on throughput.

The people who blindly trust agent output are in the former camp. They're sheeple, overdrinking from a fountain of mediocrity.

Standard disclaimer: I use AI all the time. I like AI. The point I'm making is to not blindly accept results. Think. Analyze. Learn.

[Will Germany @WillGermany]

@thsottiaux Please bring Computer Use to Windows 🙏

[Will Germany @WillGermany]

@davidgomes @encrypted I had a similar experience using Composer 2.5 & Bun (bunfig.toml). I laughed it off initially & my CI will fail regardless but I guess it shouldn't have been so eager to bypass that rule and stopped to ask the user for input. Simple enough to steer for now

[Will Germany @WillGermany]

.@Google why retire the Gemini CLI? It was pretty good. Just do a major version bump and keep it open source.

This Antigravity pivot is silly and it's much worse

[Will Germany @WillGermany]

WNBA is way too underrated

[Maxime @MaximeHeckel]

New post 📝

A deep dive into a topic I wanted to explore for a long time: how to render a realistic sky and atmosphere

I explore everything from simulating the interaction of light with air, to handling sunsets, all in real-time, from ground to space

https://t.co/lov84x0Ods

[TANSTACK @tan_stack]

SECURITY ADVISORY — TanStack npm packages
A supply-chain compromise affecting 42 @tanstack/* packages (84 versions total) was published to npm earlier today at approximately 19:20 and 19:26 UTC. Two malicious versions per package.
Status: ACTIVE — packages are deprecated, npm security engaged, publish path being shut down.
Severity: HIGH — payload exfiltrates AWS, GCP, Kubernetes, and Vault credentials, GitHub tokens, .npmrc contents, and SSH keys.
If you installed any @tanstack/* package between 19:20 and 19:30 UTC today, treat the host as potentially compromised:
�� Rotate cloud, GitHub, and SSH credentials immediately
• Audit cloud audit logs for the last several hours
• Pin to a prior known-good version and reinstall from a clean lockfile
Detection — the malicious manifest contains:
"optionalDependencies": {
"@tanstack/setup": "github:tanstack/router#79ac49ee..."
}
Any version with this entry is compromised. The payload is delivered via a git-resolved optionalDependency whose prepare script runs router_init.js (~2.3 MB, smuggled into each tarball at the package root).
Unpublish is blocked by npm policy for most affected packages due to existing third-party dependents. All 84 versions are being deprecated with a SECURITY warning, and npm security has been engaged to pull tarballs at the registry level.
Full technical breakdown, complete package and version list, and rolling status updates:
https://t.co/Zy8qG7PA9f
Credit to the security researcher for responsible disclosure.

[Will Germany @WillGermany]

I have an answer to this question... But you won't like it

[critter @BecomingCritter]

will the prices ever go down or are we basically fucked

[ani @anirudhbv_ce]

We finally know why LLMs hallucinate. It's not the model. It's the geometry.

@OpenAI text-embedding-3-large: 91/3072 dimensions do real work.

@GeminiApp gemini-embedding-001: 80/3072 dimensions do real work.

~97% of your vector database is mathematically empty. Your RAG system is retrieving from noise.

@ashwingop and I present "The Geometry of Consolidation" - a proof that RAG compression has a hard floor no algorithm can beat, set by a single spectral number your embedding model cannot escape.

Every hallucination your RAG pipeline produces? This is why.

Paper + results: https://t.co/zut8pdoPbH

[Tim Soret @timsoret]

Best summary of why games graphics might seem to have stalled (which is not true!).

We’re in a transitory period between two computer graphics techniques: rasterization & full path tracing.

Games today render in real-time what was done 10 years ago with baking. It will eventually pay off with much more dynamism.

[John Yang @jyangballin]

How much of SQLite, FFmpeg, PHP compiler can LMs code from scratch? Given just an executable and no starter code or internet access.

Introducing ProgramBench: 200 rigorous, whole-repo generation tasks where models design, build, and ship a working program end to end. 🧵

[jhey ʕ•ᴥ•ʔ @jh3yy]

"engineers use AI to ship in days what used to take a team weeks"

true

what’s dangerous is the expectation that follows
that everyone should move faster… all the time

the trade of quality for quantity

[Taelin @VictorTaelin]

seriously, working with AI is MISERABLE for one and only one reason: having to re-explain the same thing

"oh yeah this new session obviously doesn't know what proper case trees are, so let me explain it for the 5000th time in my life"

I'm tired

AGENTS.md doesn't solve this because it is impossible to fit the entire domain knowledge without nuking the context - it would be 1m+ tokens worth

RAGs don't solve this, the agent won't search unknown unknowns

SKILLs don't solve this unless I keep like a collection of 1750 skills with specific cuts of domain knowledge for each possible subset of my domain that I might need in a given chat, but that's a lot of manual work

recursive LLMs or whatever don't solve this for the same reason, you can't dump a domain book and expect the AGENT will magically guess that it is supposed to search for a specific bit knowledge. unknown unknowns

fine tuning doesn't solve this (OSS models suck and OpenAI / Anthropic gave up on user fine tuning)

I honestly think a good product around fine tuning on your domain would be a major hit and an underdog lab should take this opportunity

[Will Germany @WillGermany]

Been waiting on this one

[Cursor @cursor_ai]

Introducing /multitask in the new Cursor 3 interface.

Cursor can now run async subagents to parallelize your requests instead of adding them to the queue.

For already queued messages, you can ask Cursor to multitask on them instead of waiting for the current run to finish.

[TypeScript @typescript]

TypeScript 7.0 Beta is here!

Built on a new native and parallelized foundation, it's already being used on multi-million line codebases.

Read up more here and try it on your projects today!

https://t.co/mYN3MPp1ab

[Will Germany @WillGermany]

Happy Earth Day 🌎

Some real-time planet scale atmospherics for ya