blog.vrypan.net

Tag: ai

Verifiable AI inference

#en
#ai

· What is verifiable inference and why it would make AI outputs much more useful.

◉ Tue, 14 July 2026 18:06 // Continue reading →

A code-security-review SKILL.md

· We have more software tools than we ever had. But at the same time it’s harder than ever to know if you can trust them.

◉ Tue, 14 July 2026 17:10 // Continue reading →

LLMs Should Be Common Carriers

· And the right to to distill them should be protected.

◉ Fri, 03 July 2026 05:11 // Continue reading →

We are in the mainframe era of AI: A few centralized compute providers and millions of "dumb" clients connected to them.

◉ Thu, 18 June 2026 22:00 // note

The Cathedral, the Bazaar and the Kitchen

· Open Source in the AI era

◉ Mon, 11 May 2026 20:36 // Continue reading →

A library of agent-friendly specs

#en
#ai

· Is this the new piece of reusable code?

◉ Thu, 30 April 2026 17:15 // Continue reading →

LLM-assisted coding is not deterministic. It does it matter?

#ai

· Does it matter that LLM-assisted coding is not deterministic? No. We don’t really care about determinism (unless you are working in theoretical mathematics or philosophy). What we care about is predictability.

◉ Thu, 23 April 2026 17:12 // Continue reading →

The Edges of LLM Reality

#en
#ai

· An LLM is enabled but also constrained by its excution environment.

◉ Thu, 25 December 2025 17:52 // Continue reading →

robots.txt for the real world

◉ Mon, 02 July 2018 10:23 // Continue reading →