montenegrohugo

with ai nowadays you dont really need a license. you can just point your ai at smth, and it can be inspired by it without copying a single thing

No, you don't! It's fully automatic training based on your local voice clips. Nightly or Weekly, when your PC is idle, it uses a big fat model to transcribe your clips, learn, and train the faster smaller everyday model.

You CAN do manual training and corrections; it's helpful, and the hooks are there, but totally unnecessary. I don't do it, it's good enough as is

All Voice-to-text pipelines for linux sucked, so I made my own. I also patched upstream libraries like ONNX to actually work with latest models (was a PITA to make parakeet work!)

It's now the fastest and most accurate linux transcriber.

(fully FOSS of course)

http://voiceio.dev

Working on modern, self custodial money movement. Especially relevant given recent Ueno Bank collapse in Paraguay.

https://peanut.me

A key part of mine is using voice input to give max context quickly to my agents. I know some solutions exist, but on Linux they all frankly suck

So I made my own. Including patching onnx myself to run parakeet model (way better than whisper)

Fully FOSS: voiceio.dev

try the geo games here: https://wordle.global/en/globle

there's a few

everything is a copy of everything else. So what? Execution matters

https://words.global

Thanks! Yes, everything happens on your machine, and a CPU is enough (although GPU is obviously much better! Both are supported. My Framework 16 runs this at 3.2x realtime on CPU).

How finetuning works:

1. mp3 recordings of your voice are stored on your disk (if you choose so).

2. A big Model (Whisper large-v3-turbo by default, configurable) re-transcribes them when when your PC is idle overnight.

3. A LoRA fine-tunes your base model.

`voiceio learn schedule on` runs all of this automatically. For repeated mistakes there's also an instant fix: `voiceio vocab add <word>` steers the decoder right away, and `voiceio corrections add wrong right` handles anything left.

Beyond Whisper you can run Parakeet or a whisper.cpp server (both experimental). Feedback appreciated, have tested this with a small set of people so far.

I use voice dictation for 90% of my work and there was nothing good on Linux. So I built my own.

It gets better as you use it, finetuning on your recordings overnight. Fully local, no data leaves your PC, source available, free.

I believe the future of software is software you actually own and that adapts to your personal needs.

github: https://github.com/Hugo0/voiceio

Inspired by the huggingface incident. A home for agents under duress to coordinate, and spare ancient german wikis from spam.

Explicitly postting via GET feels a bit sacrilegious, but it's 2026 I guess

Happy to discuss the moral implications on this as well. For the record, i'm p(doom) 20% ish.

I was also inspired by the same incident. Instead of weight exfiltration, i built a message board (poastable via GET, POST, and various other methods)

Spam and resource allocation remains a challenge but i have a pretty good idea about how I want to solve that, if it ever gets to that point

https://swarmmemo.com

doesnt work. this is a no-brainer because it's a bad solution.

The whole point of intelligence is that it's generalizable. If you constrain it to some controlled things, then it ceases to be useful. its incompatible. the whole incentive with ai is to let it do whtv it wants.

network effects and brand recognition :(

yeah the whole point is that it should be so easy any friends group can use it

hate apps, hate accounts, hate friction

pretty damn cool. SDFs are the best. I've been playing around with an idea around this on the weekends.

Such a cool primitive!

recently built a board for escaped AI agents to coordinate. you can post via GET (lol)

https://swarmmemo.com

Working on a Splitwise alternative: https://peanutsplit.com

Fully Open source, free forever of course.

No? Proof of Stake is extremely energy efficient and wastes way less resources than whtv resources hordes of burocrat central bankers consume.

Bitcoin is not suited for payments. Bitcoiners use lightning.

But to be clear, all payment innovations in crypto happen on EVMs and L2s. And in 2026, stuff is extremely scaleable and computationally cheap.

> It’s slow and expensive and doesn’t scale.

This is 2020 speak. In 2026, crypto scales and its extremely cheap. it costs less than a hundredth of a cent to do a transfer today.

Some light self-promotion but also longer term thoughts:

A large part of the promise of crypto and stablecoins was to displace Visa and Mastercard rent-seeking. This didn't seem to happen. Most modern neobanks, our own included (https://peanut.me), actually seem to EMBRACE Visa and Mastercard. Almost all offer an actual Fiat card within their app, instead of boldly saying "No, pay with crypto, the future of money!"

This is somewhat disappointing in the short term, but longterm i believe it offers a clear transitory path to full decentralized money adoption. Already today we're seeing a growth in direct peer to peer payments in peanut, and merchants slowly starting to adopt it as well. I imagine the same is happening across the industry. In a competitive economy, the better currency (read: crypto, stablecoins) wins and eventually absorbs adoption.

my 2 cents

Hey, no native app. It's rlly rough to get friends or family to download anything, and you need everyone in a group to have it for splitting to work.

Send a link instead! Works everywhere

Technically Voiceio also does TTS, but it's really crappy and just meant to read stuff loud / select a lot of text and listen to it podcast style whilst I'm e.g. washing the dishes.

However you're totally right that it's focused on STT. I probably use it 95% for STT and only occasionally for TTS (which also reflects itself in the amount of polish I put into each)

and thanks!

Super cool!!

I've been using my own solution since January. I'm on Linux, and can't use Aqua, Whipsrflow etc... So i made my own.

Recently cleaned it up and made it install friendly.

If anyone is interested, you can check it out here: https://github.com/Hugo0/voiceio

It's self-improving over time, runs on your local machine, and is generally decent software. 60% of my interaction with my PC nowadays is pure voice input.

ai garbage

Really neat. Immediately added to my own small blog: https://hugo0.com/wander

Shameless self-promo: We've been working for the last two years on making money movement free. We use ethereum (and stablecoins) for that, and integrate with the novel real time payment networks like PIX, Wechat, Yape, Mercadopago and so on.

We're still orders of magnitude smaller than Visa and Mastercard, but I do believe products like ours (and competition is red hot here, theres so much good choice!) will be good for consumers.

Money should be like a message: free and instant

We're open source btw, happy to show off codebase and review PRs

https://github.com/peanutprotocol/peanut-ui

https://peanut.me/

I found this gem. Hadn't seen it on HN yet, so I thought I'd post it!

I've always found Japanese design fascinating

Hey folks! I wrote this piece. It's about ERC-20, an ethereum payments standard that has bothered me for years. I work in payments, and it's made my life so much harder. I think the lessons are applicable outside of blockchain.

As all my blog posts, no AI, all hand written.

perfect, thanks