amac
@amac0.bsky.social
3K followers 2.6K following 2K posts
Lawyer, coder, baker. Govt (Deputy US CTO for Obama and Biden), non-profit (Trust & Safety Professional Assn & Foundation, Data & Society, Public.resource), startup person (Google, Twitter). Curious tinkerer. @amac on the other thing. Now in London.
Posts Media Videos Starter Packs
Pinned
amac0.bsky.social
Thanks to many great suggestions, my #InternetLaw starter pack is getting better. Please keep the suggestions coming!

If you are suggesting yourself and you are white, cisgender male & from the US or Canada (like me) please consider also suggesting someone else who is not.

go.bsky.app/HAjFEKn
Reposted by amac
alondra.bsky.social
Call for Papers! 🧵

The Public's Science–A New Social Contract for American Research Policy a Special Issue of The ANNALS of the American Academy of Political and Social Science

Editors: Alondra Nelson (IAS) and Jenny Reardon (UC, Santa Cruz)

Abstract Deadline: Sept 19

www.ias.edu/stsv-lab/pub...
Public Science
Call for Papers Special Issue of The ANNALS of the American Academy of Political and Social ScienceThe Public's Science–A New Social Contract for American Research Policy
www.ias.edu
Reposted by amac
go-brown.bsky.social
As you may have heard, there’s an important birthday this weekend. We can only hope that people across the nation — nay, the globe — engage in tribute and fanfare proportional to the occasion. 1/
amac0.bsky.social
And, LLMs opacity + non-deterministicness make evaluation (including human) even more important, as Lili Jiang effectively argues.

3/4
amac0.bsky.social
One-shots & waterfall encourage thinking of software as products not services, which is not good.

2/4
amac0.bsky.social
AI vibe coding best practices look a lot like waterfall development, with all of its pitfalls.

1/4
Reposted by amac
tante.cc
tante @tante.cc · May 26
The narrative that #AI will do all routine and we can just focus on important/strategic/big picture things works because it tell us that we are oh so smart and special. WE DO THE BIG STUFF.

Newsflash: If your big picture decisions are not grounded in real experience they are gonna end up bullshit.
amac0.bsky.social
Claude Code logs a bunch of stuff via jsonl. The logs are a little hard to read, so I made a rough and ready parser that can also show git commits in the log timeline:

github.com/amac0/Claude...

All coded via LLM, mostly o3 & when o3 had trouble, switched to Claude 4 Sonnet (in Claude Code).
GitHub - amac0/ClaudeCodeJSONLParser at main
A parser for Claude Code's json logs. Contribute to amac0/ClaudeCodeJSONLParser development by creating an account on GitHub.
github.com
amac0.bsky.social
That seems wrong and typical Claude ("I found a test I don't pass, I'm going to pass by removing the test") but I don't know bandit, so I look online and find out that it is a perfectly smart thing to do and that bandit shouldn't check tests.

See bandit.readthedocs.io/en/1.7.3/con...

2/
Configuration — Bandit documentation
bandit.readthedocs.io
amac0.bsky.social
A good example of Claude Code teaching me stuff.
Claude code says: "Let's try a different approach - instead of using a .bandit file, let's update the pre-commit config directly to skip bandit checks for tests:"
1/
Reposted by amac
alondra.bsky.social
"For me, the answer now lies in refusal, the withdrawal of participation from systems that require dishonesty as the price of belonging."

Today I am resigning from the National Science Board and the Library of Congress Scholars Council.

I wrote about my decision in TIME.

time.com/7285045/resi...
Why I’m Resigning from the NSF and Library of Congress
I cannot participate in systems that require dishonesty as the price of belonging.
time.com
Reposted by amac
costasamaras.com
Thank you for your leadership, courage, and example!
amac0.bsky.social
... What then, is the responsible course of action? For me, the answer now lies in refusal, the withdrawal of participation from systems that require dishonesty as the price of belonging."

time.com/7285045/resi...

2/
Why I’m Resigning from the NSF and Library of Congress
I cannot participate in systems that require dishonesty as the price of belonging.
time.com
amac0.bsky.social
Alondra Nelson (@alondra.bsky.social‬) is a beacon, as usual.

"To watch these changes unfold without naming them for what they are is to participate in a collective amnesia about how knowledge infrastructures shape power relations. ...

1/
amac0.bsky.social
Saw An Oak Tree last night at the Young Vic. Most unusual play I've seen in a while and I don't really know what to think of it -- which I'm enjoying a lot. Anyhow, I'd recommend it if you get a chance to see it.
www.timeout.com/london/theat...
An Oak Tree, Young Vic review: Tim Crouch’s experimental classic returns for its twentieth anniversary
Starring a different celebrity each night, Tim Crouch’s experimental meditation on the nature of grief remains a thing of wonder
www.timeout.com
amac0.bsky.social
7) this type of coding is a bit like social media scrolling in terms of dopamine slot machine (someone at Coding AI www.oreilly.com/CodingwithAI... said this and I agree but forgot who said it)

4/
www.oreilly.com
amac0.bsky.social
"yes the tests are important, you should still do the tests and not move on if some are failing.";
5) pay attention to Claude Code and intervene;
6) Claude Code will do better in areas that you know because you'll be able to tell when it is not doing good stuff and stop/redirect it;

3/
amac0.bsky.social
3) anything you would want to have at your disposal when coding, make sure Claude Code has and knows it has;
4) stop Claude Code often to point out obvious things -- "that is out of scope for this step", "mocking the test result doesn't mean you passed the test",

2/
amac0.bsky.social
Published a short thing on one of my Claude Code experiments.
www.bricoleur.org/2025/05/clau...
Key takeaways:
1) be prepared to throw everything out (also, learn & incorporate git);
2) make the spec and the prompts simple -- no, simpler than that;

1/
Claude Code + Theater Scraper
Am playing around with AI coding. Which is fun and frustrating and very educational. Am going to post some of my attempts and observations. ...
www.bricoleur.org
amac0.bsky.social
I think that's directionally right, but I'm loathe to generalize. I have been doing a bunch of programming with the models and they often are quite good on judgment-like stuff and especially so when I know what I'm doing in the area.
amac0.bsky.social
I do think that a lot of what passes for AI "intelligence" these days is really AI producing a bunch of results and skilled humans choosing whether it is worth using / done / needs more prompting to get a better answer. Deep knowledge can be extremely helpful for the human part of that equation.
amac0.bsky.social
OK, learning git wasn't so bad. I used this video by Gwendolyn Faraday, which starts very basic but gets the job done:
www.youtube.com/watch?v=RGOj...
then this one on pre-commit:
www.youtube.com/watch?v=nX6w...
Git and GitHub for Beginners - Crash Course
YouTube video by freeCodeCamp.org
www.youtube.com
amac0.bsky.social
Ironically (?), I want to learn this stuff so that I can feel comfortable letting an AI coding helper remember it for me.