As Obama once said: “If you've got a vibe-coded app, you didn't build that. Somebody else made that happen”

"I built [new app]" announcements are a dead giveaway that an LLM built the app.

The announcers can't help advertising themselves along with the app, expressing unearned pride that they can now produce something they previously couldn't.

I think it would be much funnier if security researchers immediately disclosed all vulnerabilities they find instead of reporting them to the vendor first. Holes that are exploited in the wild lead to much worse PR for companies that ship insecure code than just abstract CVE counts

Sorry for being That Guy, but in elliptic curve cryptography, the private key is just a random sequence of bits, not a product of two primes

Your private key must meet the following criteria:
- At least 2048 bits
- At least one prime factor
- At least one NIST SP 800-186 compliant elliptic curve parameter
- At least one capital letter
- At least one digit

If any CAs want to send me their private keys I can check that they’re extra secure and random

finally, a private key that current quantum computers can factorize

OpenBSD 8.0 Released https://undeadly.org/cgi?action=article;sid=20261020021401

When people claim that it’s no longer sustainable to write code by hand, they’re not entirely wrong, but only in the same way that it’s no longer sustainable to sleep in a bed once someone has taken a shit in it.

https://po-ru.com/2026/10/18/why-im-no-longer-updating-this-repository

RE: https://f.duriansoftware.com/@joe/117436980950698017

And sure, while I’m not a fan of the Nazis either, the top executives from all of my favorite corporations hang out there too!

My favorite take is “if you’re not adopting LLMs, you lack intellectual curiosity.” These are the same people who think all of theoretical computer science is the study of “big-O notation”

A foundation for typed concatenative languages, Robert Kleffner (thanks @and!)

https://repository.library.northeastern.edu/files/neu:4f186p17k/fulltext.pdf

Typing functional stack-based languages, Christopher Diggins
https://www.researchgate.net/profile/Christopher-Diggins/publication/228985001_Typing_Functional_Stack-Based_Languages/links/5506cb7f0cf2d60c0e6d9e0e/Typing-Functional-Stack-Based-Languages.pdf

The commit description for the 2014 commit which introduced The Error Message was a work of art:

“While we work out the remaining performance improvements in the type checker,”

https://github.com/swiftlang/swift/commit/71cf7580558db8d6f6e27eaf3c7b530ee2d58510

The first sentence of the abstract for the physics department colloquium this week should win some sort of award for understatement:

"Fusion energy promises to produce clean and safe electricity on a large scale, but it still faces some challenges"

And another thing: It's not lost on me that teaching and popularization, the act of cultivating students, is the work most similar to child care in math. It is the role to which the #patriarchy assigns women in the field. It reeks of #misogyny to declare that now that the "hard" part of doing math is over (which it's not, because LLMs are a farce) that we can *finally* turn our attention to these topics. If it's so important to teach, why are you only realizing this now that you need a new reason to keep getting funded? It sounds like you still think it's below you and are giving the women who disproportionately do it a backhanded complement.

#feminism

@atax1a @davidgerard I also have strong opinions about firing up terrestrial jet engines for the purpose of automating the kinds of engineers we used to avoid hiring: Those that copied and pasted code they found on the Internet.

But I choose for personal reasons to focus on the business model of AI: Paying opaque prices for consumption of tokens while being bound by their TOS and subjecting myself to their censorship.

Do not want. I’m a hacker from the era when the word didn’t mean “techbro.”

Every time I see a vibe coder say something along the lines of "if you haven't used these tools over the past five minutes, you don't know what they're capable of, because everything's different now!" I get a little bit perplexed.

You do know that we can see the output without using it ourselves, right? Some of us are still reviewing pull requests and reading through code repositories, y’know?

how do they even interview for "ai proficiency" anyway, do they check your prompts for "do not make mistakes"