Human performance × AI
Network School
Joined June 2013
- Tweets410
- Following763
- Followers135
- Likes513
I'm a little concerned about letting Alfred know about this collusion.wiki/
Cloudflare built a browser for AI agents and called it Kitesurf.
Two things I actually do, finally in the same sentence.
Happy kitesurfing.
blog.cloudflare.com/kitesurf…
One WhatsApp chat was using 54 GB of my disk.
It was one video. 4 seconds long. Saved 58,000 times.
That chat has 10 messages in it.
All 58,000 copies were written in 19 seconds. It never noticed it already had the file, because it gave every copy a new random name.
The twist: it was never really using 54 GB.
Your Mac can make a file that claims to be a full copy but just points at the original. Costs nothing, still reports full size.
One 0.8 MB video wearing 58,000 name tags.
I tested it. 500 fake copies read as 415 MB. Real cost: zero.
Listen to Uncle Bob.
I’m significantly older than you. I started coding in the late 60s. My current strategy is to not read any of the code written by my agents. That’s the only way I can take advantage of their productivity. What I do instead is to surround the agents with extreme constraints. Unit tests, gherkin tests, QA procedures, quality metrics, mutation testing, test coverage, and a plethora of others. In the end, I have very high confidence in the code they produce because they’ve had to run the gauntlet of all of my constraints and tests.
Got my mum to deploy her first website with with Cowork.
It processed her notes from the trip to Malaysia, selected photos, assembled html and deployed it to netlify.
While I had to help her with the setup a bit, she deployed herself!
wakacje-w-utopii.netlify.app…
mateusz sawka retweeted
I built a skill to let my Claude Code build premium landing pages like this in one shot. The 3 sites in the video are one-shot results, ~$10-15 each.
Register @higgsfield_ai , install the CLI and scroll-world skill, hand the rest to Claude Code.
Fable level model running locally?
Subscription is still way cheaper, but if privacy is your main concern...
GLM-5.2 can now be run locally!🔥
The 2-bit model retains ~82% accuracy after we shrunk it from 1.51TB to 238GB (-84% size).
Run on a 256GB Mac or RAM/VRAM setups.
GLM-5.2 is the strongest open model to date.
Guide: unsloth.ai/docs/models/glm-5…
GGUF: huggingface.co/unsloth/GLM-5…
I keep posting videos everyday now, without opening the video editor.
A push-up reel still posted itself this morning. Again.
The render, the caption, the tagging, the staging: not me.
Same pattern I build into every client automation.
From unreal images to very real images.
teapaw.org/midjourney/status/2067…
Won Best Use of Exa at SuperAI NEXT hackathon.
We built Relevnt: a second brain for your conversations. Or a surveillance device, depending how you feel about it.
36 hours, 4 people, one working product.
Setting up the whole stack on aws with cdk still feels like magic. Convo prep with people search, Live transcriptions, conversation analysis and live chat coach in a couple of hours feels unbelievable.
impressed by anthropics marketing stunts
teapaw.org/ClaudeDevs/status/2065…
As a result of a US government directive, we are suspending access to Claude Fable 5 for all users. You can continue to use all other Claude models.
Here’s what this means for you:
Across Claude products, new sessions will run on your selected default model or Opus 4.8, and existing Fable 5 sessions will end with an error.
On the Claude Platform, requests to Fable 5 will also return an error. Please update your integrations to other Claude models.
We know this is a disruption to your workflows; we appreciate your patience and support.
