jedwhite

6 ms·

31.
by jedwhite 9mo ago
My view is that readability and ease of understanding have a real impact on auditability. Nondeterministic output also clearly has a significant impact on auditability. The balance between readability and determinism for auditability partly
32.
by jedwhite 9mo ago
I think there are a number of interesting ideas raised in the discussion about the use of LLMs and markdown for scripting. I'm doing my best to reply quickly and engage on the substance. I get the intent with pointing out my over-use o
33.
by jedwhite 9mo ago
The same as running `rm -rf $HOME`. Executing that in a bash script or in a markdown script are nearly functionally equivalent, with the difference being that the markdown would require you to also add explicit permissions to allow it to ex
34.
by jedwhite 9mo ago
Claude Code supports a set of flags to control behaviour such as permissions, and both `--permission-mode bypassPermissions` and `--dangerously-skip-permissions` are examples of those. The claude-run helper supports passing in those flags s
35.
by jedwhite 9mo ago
Thanks, the runme tool looks useful. You could use that along with this approach for testing. Being very meta, it could extract the code blocks from your executable markdown files to test they run correctly as part of the unit tests for a m
36.
by jedwhite 9mo ago
The markdown format provides a way to clearly format and structure instructions that Claude Code does well understanding, with clear markers for structure/headings, tables and code blocks. Making a markdown `.md` text file executable w
37.
by jedwhite 9mo ago
I can never tell :) But it's a good chance to explore the issue.
38.
by jedwhite 9mo ago
In practice after using this for real-world test suites and evaluations, the results with Claude Code if you do this sensibly are remarkably consistent. That's because you can still write the deterministic parts as the `./run_test
39.
by jedwhite 9mo ago
One thing to consider is that the steps in the pipeline can be deterministic (the code executed) while the outputs (summaries, reviews, evaluations, explanations) may be nondeterministic. An example would be summarizing data calculated via
40.
by jedwhite 9mo ago
Thanks, it is definitely an underutilized concept. I think Pete Koomen is right that we will see many more tools adopt this approach. And I hope that Claude Code and Codex add direct support for this themselves too.
41.
by jedwhite 9mo ago
You can use this without letting the markdown scripts you write execute any code at all, whether that is via Claude Code or other AI tool in future. The default permissions are to not allow execution. Which means that you can use the eval a
42.
by jedwhite 9mo ago
This is absolutely a new type of nondeterministic tool, so you're spot on there. One of the key things we realized starting to use it is that the approach allows you to mix deterministic and non-deteministic tools together as part of a
43.
by jedwhite 9mo ago
[flagged]
44.
Show HN: Executable Markdown files with Unix pipes
126 points by jedwhite 9mo ago|101 comments
45.
Grok 3 now available via API(docs.x.ai)
1 points by jedwhite 1y ago|0 comments
46.
by jedwhite 2y ago
Congrats and welcome to the new public status! Forgive the non-substantive comment but that's awesome :-)
47.
California's AB 412: A Bill That Could Crush Startups, Cement Big Tech Monopoly(eff.org)
12 points by jedwhite 2y ago|0 comments
48.
Being Responsible with Chinese AI Hype – By Peter Wildeford(peterwildeford.substack.com)
2 points by jedwhite 2y ago|0 comments
49.
by jedwhite 2y ago
I'm Jed, the author. I wrote this guide up because I'm working on an AI search startup using agents, and have been asked by other founders a few times why their content doesn't show up well with AI search engines like ours or
50.
AI optimization: How to optimize your content for AI search and agents(searchengineland.com)
6 points by jedwhite 2y ago|1 comments
51.
by jedwhite 2y ago
Congrats on the launch. Adding a memory layer to LLMs is a real painpoint. I've been experimenting with mem0 and it solves a real problem that I failed to solve myself, and we're going to use it in production. One question that I&
52.
by jedwhite 2y ago
Could this be used for doing team visualization dashboards - like metrics displays for a team with KPIs? Or is that outside the ambit? I'm asking because I feel like there is a gap between the easy set up and configurability of somethi
53.
by jedwhite 2y ago
We migrated microservices from Heroku to Porter, and also from standalone VMs and K8s running on AWS to Porter. As a coder trying to do both dev and devops on a tiny team, it was life changing for me. The key benefits for a small startup te
54.
Sympathy for the spammer. The desperate and credulous by Cory Doctorow(doctorow.medium.com)
19 points by jedwhite 3y ago|0 comments
55.
by jedwhite 3y ago
https://archive.ph/vr4M3
56.
The Remaking of the Wall Street Journal(nytimes.com)
5 points by jedwhite 3y ago|2 comments
57.
US science agencies on track to hit 25-year funding low(nature.com)
72 points by jedwhite 3y ago|50 comments
58.
Generative AI could make search harder to trust(wired.com)
232 points by jedwhite 3y ago|246 comments
59.
by jedwhite 3y ago
https://archive.ph/dFd7A
60.
Apple Considered Switch to Search Engine DuckDuckGo from Google(bloomberg.com)
39 points by jedwhite 3y ago|26 comments

More ›