NVIDIA Parakeet and Canary models now available on Speed of Sound.
New Speed of Sound release is out!
You can now use NVIDIA Parakeet and Canary models for on-device transcription, alongside Whisper. More models = better accuracy across languages + more offline/private options.
Flathub/Snapcraft links & install instructions on www.speedofsound.io/getting-star...
@zugaldia@mastodon.social :) but I won't pretend I'm very active there (yet?)
On unrelated news, you should read TWIG this week. I hear it comes with particularly good app announcements this week. thisweek.gnome.org//posts/2026/...
Now featuring much easier installation :) bsky.app/profile/zuga...
It only took me 9 months, but I finally did it! bsky.app/profile/zuga...
Great news: My little voice typing app for Linux is now available on both Flathub and Snapcraft. Links for both here: www.speedofsound.io/getting-star...
NEVER FORGET.
Instant pre-order. Love the form factor, and the hackability even more.
@technologyreview.com is the new @wired.com.
Good move about the camera, bad move about the speaker www.zugaldia.com/notes/smart-...
This is exactly my setup but with a more modest model and a Jetson Thor dev kit. I also set up Riva with Whisper for local speech recognition (I'm surprised Ollama does not support audio input yet).
Love this. It's a win-win, we listeners get no ads and the publisher gets a stickier subscription. If you're looking for another high quality podcast with a similar no ads option check out @latenightlinux.bsky.social .
@8none1.org I just heard your comment about voice being the future of typing in a recent LNL episode. And I agree 100%. That's why I built github.com/zugaldia/spe..., which I've been using daily for most of my typing for months now. Feedback/contributions always welcome.
Benevolent organizers for life.
DevFestDC is today! Find me to talk AI and @mapbox.com (and, oh yes, we're hiring).
My experience using the Meta Ray-Ban glasses for the last six months or so, where I make the case that audio, not vision, is the current killer feature for smart glasses: www.zugaldia.com/notes/smart-...
I've been using Gen-1 for the last six months or so, but my (generally positive) notes were getting too long for Bluesky, so I posted them here: www.zugaldia.com/notes/smart-...
I'm ready to throw some money at whoever is printing t-shirts with that ultrathink design.
You spent the last 10 years building robots, in retrospect it was unavoidable one day the lines blurred and you joined them.
With those options it's just impossible to go wrong.
I'll be introducing our new Feedback Agent later today @ 12.30 PST, ping/DM me here if you have any questions!
"Without music, life would be a mistake." -- Friedrich Nietzsche.
> and says it has a $5B+ revenue run-rate (Anthropic)
And about half of that is me paying for Claude Code.
I have no idea what happened but I can't help but feel disappointed. I liked your style, transparency, and direction you were setting. Wishing you the best!
I designed the biggest binoculars in the world (that you can buy) and they're for looking at outer space 🔭 🌌. Early access is a big step for our nano-scale mom-and-pop business to bring this model to life! www.analogsky.co/bloom
My favorite use case is creating custom map styles, which I'm pretty bad at. “Create a Halloween-themed map. Make it look spooky and fun.”
It's like these PhD-level machines haven't lost their inner kid after all.
I've spent a lot of time trying to get LLM to generate things in a specific way, the biggest take away I have is, if you tell it "don't do xyz" it will always have in the back of its mind "do xyz" and any chance it gets it will take to "do xyz". When working on art projects, my trick is to specifically give all feedback constructively, carefully avoiding framing things in terms of the inverse or parts to remove.
Great advice for LLMs in particular and life in general (context: news.ycombinator.com/item?id=4488...).
DevFestDC is back! In collaboration with @developers.google.com, this year we're running a one-day event themed around generative AI and emerging technologies. More information about tickets and how to apply to speak on www.devfestdc.org
