@RoundtableSpace 183 skills is impressive and also the exact problem in the thread next door. A giant skill library turns into a liability the second they all load into context and start nudging the model. Curation is the feature now, not count.
@shdu11546816 βAi cannot print more land in desirable zip codesβ
Changing regulations can, thereβs no under supply of land just an over supply of regulations that slow development
Introducing Roughdraft!
A new open source project designed to make collaboration with agents better.
The idea is to bring commenting and suggested changes to markdown (e.g. plan docs) in a nice interface.
Free, local, etc.
π https://t.co/J3YOOpL5ES π
@chamath When you say your agent cascades it, what makes that safe?
Reason I ask is agents grepping things like this to paste them scares the bajesus out of some folks (its hit or miss something)
Howβs this harnessed in a way that we donβt have to worry if weβre using 8090 for this?
Introducing the Printing Press, a CLI-factory and a CLI-library. Built with @trevin. ππ¨π
Most APIs suck for agents. Most MCPs suck for agents. Most official CLIs suck for agents. They waste tokens and time. @steipete started making his own because of this.
π A Library of agent-native CLIs you install today (Linear, ESPN, Flight GOAT (Google Flights + Kayak nonstop), Contact Goat (LinkedIn + Happenstance + Deepline more) +30+ more)
π A factory that prints new ones for any service - just type /printing-press <product name>
CLIs are fast, local, SQLite-backed. Work in Claude Code, Codex, OpenClaw, Hermes.
π https://t.co/GjnN9E9yTH
Stripe co-founder John Collison on the two types of people who will thrive in the AI era over the next 10 to 20 years:
He identifies two categories of people he's "super bullish on":
First: high-agency people.
"We know this at Stripe. The people who are like, I've been talking to customers. I know exactly what we should do. We got to go fix this. But the people who have that pep in their step and they want to go make Stripe better."
The idea is simple.
The people who don't wait around for permission, who figure out what needs doing and go do it, now have leverage they've never had before.
AI lets them execute faster without needing to assemble a huge team behind them.
Second: double majors.
"I think if you understand software and understand finance or if you understand software and understand marketing, you now can go massively improve the entire marketing funnel for your company and one person can do."
@collision connects this to a famous Paul Graham observation:
"Typically an entrepreneurship team a founding team has a collection of like five or six skills between two founders three founders."
He also points to Charlie Munger's case for multidisciplinary thinking, noting it's easier than ever to pick up a functional grasp of new fields:
"He thinks getting a functional understanding of many disciplines is not that hard you can just go read the books now you know you can talk to your AI about it and so I think multidisciplinary thinkers are going to do incredibly well."
The throughline is the same for both: AI closes the gap between knowing what to do and being able to do it.
One person can now move at the pace of a full team, and combine skills that used to require entire departments.
I ran comms for Google during Oracle v. Google in 2012, tech's last big trial of the century. Watching Musk v. Altman has me flashing back hard.
Fifteen observations on the trial comms war so far, in no particular order: