lxe

That is definitely not the case. It transcribes everything very accurately, especially all the "ahs" and "uhms".

What's the best local TTS model right now? I'm running parakeet on a mac which transcribes all my uh's and aahs. I'm running whisper on linux/cuda and I by far prefer that one over parakeet.

It's kind of ironic how this doesn't work in Firefox.

Scientists are the ones with real agency and leverage. The change has to come from the grassroots.

Careful with this philosophy. It does work well for the short term. At some point of constant following of 'disagree and commit' mantra, you'll end up in a world where you have zero agency and zero energy to constantly do the work you hate.

The system is fine. The culture is broken. Scientific publishing isn't forced on the community by regulation or necessity. You can publish papers in infinite number of ways online. Unlike something like healthcare or housing, where there are no alternatives, there are plenty of alternatives when it comes to media publishing.

Genuinely interesting, thank you.

Have you tried setting up a bot for something like WhatsApp? I'd much rather use a KGB spy platform with actual user ergonomics.

This is pretty awesome. I work with editors and monaco-like things a ton, and I review (look at) very large PRs very often. Having this speedy optimized interface is a delight. Check out their trees lib as well.

Did they even ask their customer base before approving the design? I don't care about Ferrari, but people who do care about Ferrari will not like this.

Bold of you to assume that lawmakers have any common sense when it comes to technology legislation. It could have taken 3 interns 3 hours at each browser company to implement a cookie consent standard 15 years ago, yet here we are in cookie banner hell.

They were crazy overzealous about not allowing these technologies for a long time. I'm pretty sure I had many posts about this complaining over the years.

Woah this is a MASSIVE deviation from FF's previous philosophy on allowing WebSerial. This is a GOOD thing!

What I don't quite understand is why would one of the most advanced AI labs use rudimentary broken text match heuristics to track and detect abuse. Why not run simple inference on actual turns out of band, and if abuse is detected, adjust the quotas semi-retroactively.

Yall remember https://en.wikipedia.org/wiki/Mystery_meat_navigation? Back in 2004-ish era, there was an explosion of very creative interaction methods due to flash and browser performance improvements, and general hardware improvements which led to "mystery meat navigation" and the community's pushback.

Since then, the "idiomatic design" seems to have been completely lost.

hahaha I’m glad I’m just a procedurally generated NPC

I built one for cross platform — using parakeet mlx or faster whisper. :)

I know that there's a deceptively high amount of engineering required for these kinds of things

I think there's a deceptively low amount of engineering required for most medical and medical-adjacent tech. The high costs are rooted in pervasive industry-wide centuries-long FUD campaigns.

ChatGPT, read this article and turn it into a AGENTS.md

Distinguished staff level trolling

Honestly, I found that the best way to use these CLIs is exactly how the CLI creators have intended.

Yeah, that's exactly what I started to do with mine. It runs local Whisper on a CUDA, on a graphics card. Whisper is actually better than any other model that I've seen, even things like Parakeet. It can do language detection. It automatically removes all the ahs and all the ohms unless I specifically enter them in my speech. I think this whole paragraph is going to take maybe half a second to process and paste without any issues.

(and it did it perfectly without any edits required for me at all.)

I built something similar for Linux (yapyap — push-to-talk with whisper.cpp). The "local is too slow" argument doesn't hold up anymore if you have any GPU at all. whisper large-v3-turbo with CUDA on an RTX card transcribes a full paragraph in under a second. Even on CPU, parakeet is near-instant for short utterances.The "deep context" feature is clever, but screenshotting and sending to a cloud LLM feels like massive overkill for fixing name spelling. The accessibility API approach someone mentioned upthread is the right call — grab the focused field's content, nearby labels, window title. That's a tiny text prompt a 3B local model handles in milliseconds. No screenshots, no cloud, no latency.The real question with Groq-dependent tools: what happens when the free tier goes away? We've seen this movie before. Building on local models is slower today but doesn't have a rug-pull failure mode.

Thanks for surfacing this. If you click to "tools" button to the left of "compile", you'll see a list of comments, and you can resolve them from there. We'll keep improving and fixing things that might be rough around the edges.

EDIT: Fixed :)

Eh. This is yet another "I tried AI to do a thing, and it didn't do it the way I wanted it, therefore I'm convinced that's just how it is... here's a blog about it" article.

"Claude tries to write React, and fails"... how many times? what's the rate of failure? What have you tried to guide it to perform better.

These articles are similar to HN 15 years ago when people wrote "Node.JS is slow and bad"

I hate this type of headline.

Imagine if we had something like:

    "google downloads and executes malware"
    "outlook downloads and executes malware"
    "chrome downloads and executes malware"

    "a person using a computer downloads and executes malware"

It's crazy how the etymology of "Kentucky" cannot be traced with certainty. Goes to show how much of the native American culture and language is now untraceable and how fragile our record-keeping is, even in "modern times".

I always wondered how something like AWS or GCP Cloud Console admin UIs get shipped. How could someone deliver a product like these and be satisfied, rewarded, promoted, etc. How can Google leadership look at this stuff and be like... "yup, people love this".

This is nuts and I absolutely love this. So you convert the PDF into image, edit the image, then convert the image back into a PDF.

Object.defineProperty on every request to set params / query / body is probably slower than regular property assignment.

Also parsing the body on every request without the ability to change it could hurt performance (if you're going for performance that is as a primary factor).

I wonder if the trie-based routing is actually faster than Elysia in precompile mode set to enabled?

Overall, this is a nice wrapper on top of bun.serve, structured really well. Code is easy to read and understand. All the necessary little things taken care of.

The dev experience of maintaining this is probably a better selling point than performance.

I barely use a keyboard for most things anymore.