Publications tagged "Ai"

AI coding assistants are like an extra gear on a bicycle: they help you go faster but don’t replace the rider. This post explores the hidden patterns, from backwards compatibility obsession to test regression and security gaps, and how to stay in control.

A former colleague asked me to provide advice to high school students about LLMs. Here’s what I told them about navigating this transformative moment in technology while avoiding the hype and protecting their future.

My Hugo blog’s front page suddenly started displaying every page instead of just blog posts. Rather than fix it myself, I decided to test Claude’s debugging capability. What followed was a hilarious descent into AI debugging chaos that taught me more about LLM limitations than Hugo templates.

Hallucinations in large language models (LLMs) are no longer a niche concern. With real-world incidents causing significant damage, the question remains: can you trust an LLM?

The summer of 2025 is seeing AI-powered IDEs move from novelty to default. Cursor, GitHub Copilot, and others promise vibe coding: rapid prototyping through natural-language prompts. Productivity is jumping — yet the trade-offs are already showing. Six patterns are emerging: prediction vs. planning, reactive patch loops, invisible gaps, shifting privacy promises, tooling drift under scale, and erosion of developer choice. Here’s what we’re learning so far — and where the tools need to head next.