Steve Klabnik
Everything, newest first — across every channel. Their profile → · Subscribe
Hiding
30 October
21 September
From one piece Arguing about arguments 2 beliefs, in the piece's order there
-
korrents.com
Named function arguments' readability advantage for humans is even stronger for AI coding agents reading code.Their words
And the readability advantage for humans is even stronger for agents: the extra clarity of the text inline seems to (I haven't run real evals on this yet...) be more helpful and use less tokens and calls when looking at the actual call site.
-
korrents.com
Function signatures that support many flexible calling styles make it hard to know everything a function can do.Their words
they can be really concise, and beautiful to a certain kind of eye, but it also makes it really, really hard to know what all you can do with a function.
2 September
-
Their words
AI is here to stay, and it is dramatically changing how people work.
14 July
From one piece Too many words about DIDs 2 beliefs, in the piece's order there
-
korrents.com
Giving people the choice to manage their own cryptographic keys matters more than forcing everyone to do so.Their words
A criticism of this boils down to "well, most users won't do that," and while that's true, I also think that's okay for most people, and that having the choice is more important than forcing every user to deal with their own key management.
-
korrents.com
A person who controls their own infrastructure and signing keys can meaningfully own their digital identity.Their words
I think that this possibility is an important design property, and allows motivated users to meaningfully own their identity.
10 July
-
Their words
it is unfortunate if these policies end up making it so that individuals like me can't make my software as robust as products from large organizations.
22 April
23 February
27 January
From one piece The most important thing when working with LLMs 3 beliefs, in the piece's order there
-
korrents.com
Practices that make an LLM's work easier to evaluate also tend to benefit the human developers working alongside it.Their words
what's good for Claude is also probably good for humans working on your system.
The most important thing when working with LLMssteveklabnik.com
-
korrents.com
Computers execute exactly what you tell them, not necessarily what you meant to tell them.Their words
the computer will do exactly what you say, but not what you mean.
The most important thing when working with LLMssteveklabnik.com
-
Their words
The most important thing that you can do when working with an LLM is give it a way to quickly evaluate if it's doing the right thing, and if it isn't, point it in the right direction.
22 January
From one piece Agentic development basics 3 beliefs, in the piece's order there
-
Their words
"AI does better on greenfield projects" is not exactly true, but there's enough truth to it that I think you should do a new project.
-
Their words
In the previous post, I said that you shouldn't be mean to Claude. I think it makes the LLM perform worse.
-
Their words
Claude is going to try and figure out what you mean and do that. This is both the source of these tools' power and also the very thing that makes it hard to trust them.
15 January
From one piece How to think about Gas Town 2 beliefs, in the piece's order there
-
korrents.com
AI has changed software development enough that some old assumptions about it no longer hold.Their words
I have a sneaking suspicion that the world has changed, and some old assumptions do not hold any longer.
-
Their words
And while rigor is a great value to have, when you're in an experimental mindset, relaxing rigor can be a valid strategy to move forward.
7 January
From one piece Getting started with Claude for software development 2 beliefs, in the piece's order there
-
korrents.com
Becoming productive with large language models is not easy, despite what many advocates claim.Their words
Becoming productive with LLMs is not actually easy, no matter what other people tell you.
-
korrents.com
Swearing at or being rude to an AI assistant makes it perform worse than treating it respectfully.Their words
I un-ironically believe that swearing at Claude makes it perform worse.
-
Usesrcmnd.app
ClaudeTheir words
I’m going to be speaking about Claude directly, because it’s the tool I use the most, but a lot of this should apply to other platforms as well.
21 December 2025
2 December 2025
18 November 2025
22 October 2025
10 June 2025
9 June 2025
30 May 2025
28 May 2025
15 May 2025
21 April 2025
7 April 2025
17 March 2025
13 March 2025
5 March 2025
7 February 2025
16 October 2024
Nothing matches.
What is a korrent?
A korrent is a belief a person has stated in their own words: one sentence stating the claim, backed by a quote and a source, kept at korrents.com.
Under a name here, the quoted block is what they actually said. The korrent beneath it is the claim those words support, in korrents' wording — tap it to see the record, its source, and who else holds it.
Nobody here wrote their own korrents. They are compiled from public statements, and a person can change their mind, which is recorded too.
About the English under a post
Some people here publish in a language other than English. Where they do, this site shows a machine translation beneath the post, in this typeface — the site's own, not theirs.
The post itself is never changed, moved or hidden: what is set in the serif above is exactly what the person published, and it is what to quote them on. A translation can be wrong in ways that matter, especially about tone.
Only the post's own words are translated. A quoted post, a linked article and a belief on korrents.com are left in their original language.