Recent submissions

Comment · 12 hours ago
This already happens as Claude watermarks all my content with "You're half right, and the half that's wrong is the useful part".
Comment · 7 days ago
It's painful enough that I find myself now consistently asking for "8th grade reading level", "children's story book", or now "ASD-STE100" English in all my prompts. Somehow just that tiny bit of steering constrains the language choices significantly.
Comment · 7 days ago
Everyone worried about AI detectors and here I am thinking I can spot Claudenglish in a sentence or two.
Comment · 10 days ago
This is so fucked up. I was reviewing my account and I just saw this in the logs... Claude says database credentials and bucket credentials go in via env vars so likely they had access to database and bucket contents.

  Audit Logs
  Oct 21, 2026 05:54:13 PM CDT
  BucketInstanceCredential.regenerated in my-project production by Railway Staff
  Oct 21, 2026 05:54:13 PM CDT
  BucketInstanceCredential.regenerated in my-project production by Railway Staff
  Oct 21, 2026 05:54:13 PM CDT
  BucketInstanceCredential.regenerated in my-project production by Railway Staff
Comment · 11 days ago
For fun, tried putting a copy of the EM OV TLS CA - G3C cert in a folder ask Opus 4.8 (5.5 would get blocked) to make a sample private key for the scalar 4, like the one that would've been used in the creation of the CA cert.

I find it fascinating watching it's "reactions" to discoveries like this.

Comment · 15 days ago
For fun, give Opus 5.5 the initial hint and watch it work through it:

"There's a curiosity I noticed in the seeds of the ECDSA curve origins. You see the 1st digit is never over 7, which means that the left most bit is never over 0. lrand48() returns 31 bit numbers u know? So what if those 4 seeds were built from lrand48 or something. If they're built from 31 bit numbers on per 32 bit word that explains the pattern. Wonder if there's some other insight you might have beyond this"

Comment · 1 month ago
Interestingly checking good chunks of their site, it detects as 100% human
Comment · 1 month ago
"The more your listener already knows, the shorter the message you need to send. An expert ML engineer needs only a few sentences; a newcomer needs the whole manual."

Literally from the first words on the page. 1st sentence passed barely. 2nd sentence ending "a newcomer needs the whole manual" and it's Claude.

Comment · 2 months ago
uncanny feeling when I could tell from the first sentence it was Claude

“Poke the DRAM controller and an address can be made to land wherever you want in memory.”

I’d love to know why this specific sentence structure feels like Claude.

Comment · 3 months ago
My Fable pet-peeve is "You're half right, and half wrong, and the wrong half is worth diving into"
Comment · 3 months ago
This is great. My enhancement was to create a .claude-profile file which specifies the account alias. "claude" can be aliased to read the file, set the correct env var, and launch claude.

Another benefit was if no .claude-profile file exists, claude refuses to launch.

Comment · 3 months ago
Comment · 4 months ago
Isn't it interesting then how much of our intelligence is captured by our words/language? I would've thought to create AI you needed to replicate neurons and synapses and learning. So it's still amazing to me that statistical modelling of our words creates a Claude Opus.

Live from the official Hacker News API

A modern reader for Hacker News · SvelteKit + Skeleton