The documentation for this image processing library by @vruba is one of the most interesting things I've read in weeks:

https://github.com/celoyd/potato/blob/main/docs/personal.md
https://github.com/celoyd/potato/blob/main/README.md
https://github.com/celoyd/potato/blob/main/docs/concepts.md

Philosophical discussion of the nature of seeing and what am image is vs a map, fascinating technical details about how satellite imaging works and why it looks as bad as it often does, a lot of really thoughtful conversation about engineering and aesthetic process, and even an amusing unit of measurement — grams per terrapixel.

The Revolution Will Not Make the Hacker News Front Page

(with apologies to Gil Scott-Heron) If you get all of your important technology news from "content aggregators" like Hacker News, Lobste.rs, and most subreddits, you might be totally unaware of the important but boring infrastructure work happening largely on the Fediverse, indie web, and other less-centralized communities. This is no accident. The rough consensus of these spaces has been strongly in favor of the…

http://soatok.blog/2025/12/17/the-revolution-will-not-make-the-hacker-news-front-page/

New post on whether mastery in writing software still matters. I decided it does to me.

https://vickiboykis.com/2025/10/20/i-want-to-see-the-claw/

We’re speaking at the DeCompute Conference!

So many bad hashes, so little time. At 10:20 AM (GMT+8), Opal Wright will discuss a new hash construction to help people avoid common pitfalls when hashing multiple pieces of data together.

Catch Paul Bottinelli at 11:15 (GMT+8) for his talk on effectively combining MPC protocols with TEE deployments.

We will also be at Token2049 later in the week. DM us if you’d like to grab coffee with one of our cryptographers attending.

Cryptography is inherently political, as Rogaway argues in The Moral Character of Cryptographic Work.

The kinds of technical problems that one chooses to solve are often informed by one's own political views.

One cryptographer might work on encrypting consumer devices such that no one but the owner can decrypt them, thereby preventing unlawful access by overzealous governments. Bullets don't solve math problems.

Another might work on DRM for ensuring that copyrighted videos are only playable on hardware that has been certified to prevent unauthorized reproduction.

These problems have very different politics.

But consider: Using DRM-like technology to run software in the cloud such that not even the cloud provider can decrypt?

The work you do in this field is never in a politically neutral vacuum. Nearly every capability can be used in an inverted power dynamic.

I did a deep dive on current embedding sizes and architectures

https://vickiboykis.com/2025/09/01/how-big-are-our-embeddings-now-and-why/

#Microcoding Quickstart by @taviso 

https://github.com/google/security-research/blob/master/pocs/cpus/entrysign/zentool/docs/intro.md

See also https://bughunters.google.com/blog/5424842357473280/zen-and-the-art-of-microcode-hacking

We hacked Gemini CLI, Vertex AI, Assistant, and other AI systems by embedding prompts into images that are not visible to users.

When these systems auto-downscale images, the hidden prompts emerge from previously invisible pixel patterns. https://blog.trailofbits.com/2025/08/21/weaponizing-image-scaling-against-production-ai-systems/