I highlight the limitations of machine translation by linking to classic examples of linguistic failures. These sources showcase why software often fails to capture context, resulting in the humorous errors frequently found on sites like Engrish.com.
I’ve curated a list of essential computer science texts that offer deep insight. It features foundational papers on lambda calculus, CSP, relational databases, and early programming languages from legends like Hoare, Knuth, and McCarthy.
Review essential truths and myths of software development, including the 28x productivity gap, the high cost of maintenance, and why software estimation often fails due to poor timing and incorrect stakeholders.
I shared a humorous link from the Telegraph about how grammar becomes the first casualty of war. It highlights linguistic errors that crop up in media coverage during times of conflict.
AI research works much better when you specify failure modes, define validation criteria, and guide the system with concrete examples instead of vague requests.
I explored ChatGPT's export limits, Zettelkasten note-taking, and AI agent architectures from the AI Engineering Summit. I also noted Tyler Cowen’s LLM writing workflows, new OCR models, and the 'Diffusion LLM' approach to fixing hallucinations.