Jason Sich

Developer, AI pilled, obsessed with going fast ⚡︎

[Jason Sich @jasich]

Your choice — be a "helicopter parent" or let AI crank.

[Jason Sich @jasich]

How to make sure the slop isn't yours 👇

https://t.co/MGi62K3q49

[Jason Sich @jasich]

Slop is the default.

AI hands you bland, broken, generic unless you fight for better.

The fight isn't generating more.

It's making AI prove its work like a human would.

[Jason Sich @jasich]

It's like the kid that would always try to change the rules for the game you we're playing once they started winning.

[Jason Sich @jasich]

@sergeykarayev I know people like to jump onto the sandbox bandwagon, but why not focus on trusted sources first.

[Wes Bos @wesbos]

the four horsemen of the apocalypse

[Jason Sich @jasich]

@LeaVerou @claudeai Yes! I’ll tell agents to cut code or even responses by 50% and I end up with better results. Sometimes I just repond with “simpler”

[Lea Verou, PhD @LeaVerou]

💡Recent insight: gaslighting @claudeai seems to improve code quality >90% of the time.

“You overengineered this, there is a simpler way”
“There is a smaller delta that buys us most of the benefits”
“There is a more elegant way”
“This is not architecturally coherent”

…before I even read its code. 😆

[Jason Sich @jasich]

I know it's going to be a great day to use AI when it starts like this...

[Jason Sich @jasich]

As we see companies make more while at the same time lay more people off, the temptation becomes to to do a redistribution of wealth.

Take what these companies are earning and redistribute it back to people as UBI.

The actual move is a redistribution of work.

All because your old job is dead doesn't mean that you're useless. You just need to find a new job.

Probably one that didn't even exist before.

[Jason Sich @jasich]

just trying to tie lessons from a horse farm to how I use AI now, https://t.co/Xz9zDbEdQz

[Jason Sich @jasich]

Your choice — be a "helicopter parent" or let AI crank.

[Jason Sich @jasich]

WisprFlow keeps spelling Vite as “VeeT” and I don’t hate it

[Jason Sich @jasich]

wow, what an addition to the team 🤯

[Andrej Karpathy @karpathy]

Personal update: I've joined Anthropic. I think the next few years at the frontier of LLMs will be especially formative. I am very excited to join the team here and get back to R&D. I remain deeply passionate about education and plan to resume my work on it in time.

[Andrej Karpathy @karpathy]

Personal update: I've joined Anthropic. I think the next few years at the frontier of LLMs will be especially formative. I am very excited to join the team here and get back to R&D. I remain deeply passionate about education and plan to resume my work on it in time.

[Jason Sich @jasich]

For me, X is really about information and sharing of ideas. Grok is a 10x catalyst for making it more useful.

[Jason Sich @jasich]

@addyosmani I get these types of arguments but I also don’t lose sleep at night because I don’t understand how my program translates to machine code. We’re moving up the abstraction layers and at some point we’ll need to let go.

[Peter Steinberger 🦞 @steipete]

Try https://t.co/8xIcXuojse on one of your repos and let codex work its magic. It's amazing at uncovering bugs you didn't know you had.

[Jason Sich @jasich]

@coreyhainesco so awesome that you're sharing these skills for free to the world!

[Jason Sich @jasich]

It's not even a question if my OC could do these types of automations.

I only ask myself how much could it do, how well can it do it and would it be worth it.

[Peter Steinberger 🦞 @steipete]

People freaking out over my AI spend. What nobody sees: Part of what excites me so much about working on OpenClaw is that I'm trying to answer the question:

How would we build software in the future if tokens don't matter?

We constant run ~100 codex in the cloud, reviewing every PR, every issue. If a fix on main lands, @clawsweeper will eventually find that 6 month old issue and close it with an exact reference.

We run codex on every commit to review for security issues (as it's far too easy to miss).

We run codex to de-duplicate issues and find clusters and send reports for the most pressing issues.

We have agents that can recreate complex setups, spin up ephemeral https://t.co/Q1NRXLemEy machines, log into e.g. Telegram, make a video and post before/after fix on the PR.

There's codex that watch new issues and - if it fits our documented vision well, automatically create a PR of it. (that then another codex reviews)

We have codex running that scans comments for spam and blocks people.

We have codex instances running that verify performance benchmarks and report regressions into Discord.

We have agents that listen on our meetings and proactively start work, e.g. create PRs when we discuss new features while we discuss them.

We build https://t.co/bmA1XnoB7P to split all our projects into functional units to review and find bugs and regresssions.

We do the same split for security with Vercel's deepsec and Codex Security to find regressions and vulnerabilities.

All that automation allows us to run this project extremely lean.

[Jason Sich @jasich]

@mitchellh @yacineMTB There are probably plenty of time bombs waiting to go off, but I still can’t help wonder how far you can push a black box mentality right now. I’m cut to find out