Blogs
ArchiveLatest
Free article for Arpeggio / LinkedIn
Thesis
We are not seeing a tooling gap. We are seeing a structural mismatch between probabilistic generation and loadābearing systems.
Despite heavy investment, multiple independent signals point to theā¦
I previously wrote a paper arguing that LLMs behave more like Type-3 (regular) machines than true reasoners. This challenge was, in part, a small experiment to test that claim against a familiar problem under unfamiliar constraints.
Zenodo author page:ā¦
Foreword
Well then, this is a story about the one time I allowed my AI agent to co-write a complex library, and what I learned from the experience.
The following post is my subjective experience, and part of the reason I published independentā¦
Why Language Models Fail When Truth Is Co-Defined
Abstract
Most commentary on large language models still frames their limitations in terms of intelligence, hallucination, or benchmark performance. That framing is too coarse. The moreā¦
Practical categories where AI still struggles
1. Ownership and lifecycle design
- Deciding who owns a value, not just how to wrap it
- Borrowing versus retention across module boundaries
- Lifetime management across async work
- Destruction asā¦