Viewing profile — noahbp

It’s a self-improving harness, and ARC-AGI-3 is explicitly a few-shot benchmark. It’s likely that it gave itself more than the maximum number of tries to learn the games, or even h…

I had a similar project for bringing up Tailscale on a very resource-limited router, the gl.inet SFT1200, which has only 128 megabytes of RAM, and a dual-core MIPS processor. Despi…

I think Huggingface was hacked, and if Huggingface and OpenAI claim it was OpenAI, then I believe them. I'm saying that OpenAI's models cheat to win benchmarks, more than other mod…

Time to first token, especially for smaller models, can be sharply reduced. Latency can be just as important as overall throughput, especially for inference providers like Groq and…

This is clearly just OpenAI's marketing. Their models, very famously, are prone to reward hacking benchmarks in ways that other models are not. They need to publish numbers showing…

I have a hard time blaming anyone other than the internet monopolists and the FCC for this. If we had similar regulations as the UK (you lay infrastructure that serves Internet use…

Plenty of people purchase digital movie rentals from Apple, Youtube, etcetera because they know they will watch it once, and the lower price in exchange for a temporary license is …

That most-recent spike during/post-COVID really puts into perspective just how unreasonable low-wage employers were to be so hysterical.

It’s incredible that “Search the Current Folder” is not the default, nor, as far as I’m aware, can it be made the default.

I’d argue they started doing that a bit earlier. My hard drive from 2011 made using Windows a miserable experience any time the search indexing or windows defender scans kicked on,…

This is it. I can’t believe the other commenters are unaware that Cursor recently fine-tuned an open-source model and brought it to the frontier, even if it remained there briefly.…

Unfortunately, you're right. It is LLM-written: https://www.pangram.com/history/8b17aa57-ce1f-4f46-85f4-4db0...

That’s my mistake then. That particular gradient is the visual equivalent of reading a paragraph with em-dashes and “It’s not just X, it’s Y”. This is quite the coincidence. Forgiv…

FYI, the color gradient on your website is an easy tell that it was vibe coded: https://prg.sh/ramblings/Why-Your-AI-Keeps-Building-the-Same...

>Why is uneven, concentrated development some kind of public good? Because of agglomeration and the incredible desirability of mixed-use walkable neighborhoods (see rents in walkab…

Your account has written 6 comments in 13 minutes, every one of them in AI-like pithy prose.

The same criticism has been said of Deno and Pnpm and bun, and yet, despite all these years since their respective releases, node and npm remain slower than all three options.

3. Superior CPU schedulers, which do not ever undergo serious regressions that are not ever fixed: https://x.com/SheriefFYI/status/1856356547875541196 https://x.com/itsHemu2K/statu…

Opus 4.6 has gotten pretty good at writing Powershell. It’s the first model where I didn’t have to ask, repeatedly, that it use Powershell 5, and never use emojis or other invalid …

Why would you think that the same thing preventing density and new development in cities won’t stop your new city from growing before any building taller than 2 stories is built?

It’s a good thing that businesses can make investment plans with legible rules to follow. Too many communities are blocking data centers for no good reason, and this preempts NIMBY…

When I lived in Costa Rica, I lost three surge protectors in a year to power surges. During one such power surge, I didn't notice that the red light indicating surge protection was…

In case anyone doubts it's AI-written: https://www.pangram.com/history/b7433cbe-08e7-43fe-9a32-3e43...

They seem to have copied Cursor in hijacking ⌘Y shortcut for "Yes" instead of Undo.

I don't think we should excuse Microsoft for bloating Windows because memory became abundant.