just did a monolithic clean-up of my digital life with opus 5.5 on my notion
first time I've been VERY impressed with a model from Anthropic since sonnet 4.6
they have truly nailed how to speak to someone with this model
I used to beg my AI agents to write one clean PR without babysitting every line.
One random Maven talk from @poteto changed that.
45 PRs deep now and *almost* runs on its own.
What a time to be alive.
tested out building a launch video with @HeyGen hyperframes
Please don't forget that this prompt exists it will, as it's done for me, 10x your testing process with agents.
"""
/goal
- go over every single feature within this app and create a user story with expected behaviors (inputs, processing steps & outputs) that should happen based on the code
- keep a single canonical spreadsheet tracking the features status
- test every user story and document all errors
- when done with the testing, fix every logistical error or ux error
- finally, test every user behavior again post fix
''''
Enjoy.
Celebrating a small win for my app, getting 99 clicks and ranking top 10 in google search over the last 30 days 🥳
Jon retweeted
Today we’ve raised $52M Seed and we are announcing the public launch of S2.1 Pro.
>It can clone a voice from 5 seconds of audio
>2x faster than Cartesia & 1/6th the cost of Eleven Labs
>most expressive model with word level control over emotion, intonation, pacing etc
We support frontier AI companies including HeyGen, LiveKit, Retell, Sanas, and OpenArt all run our model in production.
If you're a business and we can't cut your voice AI costs by 50%, we'll give you 1 year of Fish Audio for free.
Book a demo: s.fish.audio/tmapke
To celebrate our first birthday, we'll give you 1 month of S2.1 Pro for free. Like, retweet, and comment “Fish” to get it.
I abhor using terminal interfaces
so, I got sol to help me setup a remote desktop (tigervnc) and the unofficial Linux chatgpt app on my hetnzer server
now I can code using my phone on a VPS without having to worry about my device powering down
and I don't have to deal with that termius+tailscale setup
s/o @levelsio
very bizarre to me that codex's scheduled tasks can't make api calls outside of their sandbox like claude code routines
i just figured out that you can start threads from different projects...
and gpt 5.5 will do so on its own and continue to work
@OpenAI has truly created a beast of a product with codex
im hyped for 5.6














