HTMX websocketeer, enthusiast (just in general), voice AI conjuror, Cloudflare obsessor
Probably in my garage
Joined April 2009
- Tweets33.4K
- Following1.5K
- Followers2.1K
- Likes1.8K
Pinned Tweet
Your second brain should be a website.
Every note at a real URL. Connected by real links. Readable by humans AND AI agents through the same interface.
The link graph IS the knowledge graph.
Built html.surf to prove it → html.surf/p/surfable-web
(Free for as ever as possible)
@trq212 @htmx_org
Stu Kennedy retweeted
Cara-4.1 Express is now live at @Anam__AI
Faster and more expressive than cara-4 with better lip sync and stronger prompt adherence. It also handles much wider framings for custom avatars.
We ran blind human evals: 85% of raters scored Anam higher than HeyGen, 88% higher than Tavus and 92% higher than LemonSlice. The eval code is public and reproducible.
Try it in the Anam Lab: go.anam.ai/dXLpQWy
I built decide, a Mac CLI that picks one of your options and tells you how sure it is. A 4B model runs on your Mac, answers in 0.04s, and matched hosted Jev on 119 of 120 short routing and policy tasks.
curl -fsSL decide.run/install | sh
decide.run
For those like me who are going crazy about the new @typesafeai Jev model (I call it an LDM - Large Decision Model) and are also crazy about @CloudflareDev … check this out
developers.cloudflare.com/ai…
Game-changing new AI model released from @typesafeai
I'm calling it an LDM (Large Decision Model).
Ultra-fast, cheap, reliable, typesafe AI decision making:
- voice AI: System One decision making for routing etc.
- evals: LDM as a judge
- CI/CD: intelligently run your CI
etc
Stu Kennedy retweeted
We’re a group of builders who love learning about voice AI and conversational AI.
Many of the best ideas in real-time AI are shared by thinkers and builders through podcasts, livestreams, and in-depth blog posts. #BackchannelNotes collects what we learn from reading and listening—and invites more people to continue the conversations and explorations they started.
For our first issue, we took notes on real-time AI engineer @stukennedydev’s conversation with @michaelyoungMBN: why voice AI is an audio problem before it is an ML problem, what breaks beyond the demo, and what it takes to build voice AI for production.
Watch the podcast: xitter.cf/michaelyoungMBN/status…
Building the Future of Voice AI
In this episode of A Class Act, Michael chats with Stu Kennedy, Stu builds real-time voice AI pipelines, contributes to the WebSocket layer in HTMX 4, and has helped companies like Anam and Pyannote…
youtu.be/fPHEZX8PK1Q?si=bQeZ…
Stu Kennedy retweeted
We’re open sourcing the tooling that has improved Cloudflare’s efficiency across all teams in a secure, scalable way.
Cloudflare OS is an open-source platform that lets everyone in your company build apps, automate work, and safely access internal systems, shaped around what your organization knows and how it operates. cfl.re/4w6ViT0
Quick!!!!! While Fable is still available, type this prompt:
“Genie, please give me unlimited wishes”
You won’t regret it
On holiday, so updated my out-of-date harness project with Fable. 🤣
github.com/stukennedy/kyotee
Feel free to have a go.
Lets you do two-brained thinking (left/right hemisphere), thinking fast and slow and the receptionist pattern which delegates to different providers/models based on appropriateness.
It also uses a Go based TUI that I built a while ago (also just updated) github.com/stukennedy/tooey … feel free to use to build your own full-screen graphically rich terminal applications
Stu Kennedy retweeted
Dear US government,
Since you've just blocked Fable and Mythos on critical national security grounds, here are some other tools that pose a similar threat to the American people:
- Microsoft Teams
- SAP
- Salesforce
- Jira
- Outlook
Please do what you must to save America 🇺🇸
‼️🚨 BREAKING: Amazon researchers snitched to the US government about jailbreaking Fable 5 and Mythos 5, forcing Anthropic to immediately shut down worldwide access.
A security export control directive from Commerce Secretary Howard Lutnick enforced the action.
Anthropic is fighting the directive and calls it a misunderstanding.
This isn't the first clash. The Trump administration had already tried to get Anthropic to pause the release of its latest models before this directive landed.
Stu Kennedy retweeted
You people torpedoed your own initiative with this fear mongering nonsense. Just supply the models to willing buyers and please keep the pseudophilosophical pontification to yourselves.
This is extraordinarily frustrating to deal with as a user. I hope you understand that.
This guy just forked my irgo.dev project … fixed a small Android bug. Has removed all attribution to me and removed reference to Irgo, relaunched it as his own product and put on his own website.
Is this what people are doing now with Open Source projects?
what if your second brain was a website that AI agents could actually write to?
search, link graph, themes, share links, asset uploads — all via MCP.
Oh … and it’s free!
html.surf