TWIL - Week #22

This Week I Learned:

Comments

Post a Comment

Popular posts from this blog

This Week I Learned - Week 16 2026

This Week I Learned -  * From The Batch - - Alignment training teaches LLMs to behave like assistants, but it tethers them to that behavior only loosely. Beyond alignment training, system prompts act as behavioral guardrails, but motivated users can bypass them.  - The lengths of tasks completed autonomously by AI agents have doubled roughly every seven months, according to METR, an independent testing organization - LLMs' knowledge is still relatively limited with respect to infrastructure and the complex tradeoffs good engineers must make...finding infrastructure bugs — say, a subtle network misconfiguration — can be incredibly difficult and requires deep engineering expertise.  - Research involves thinking through new ideas, formulating hypotheses, running experiments, interpreting them to potentially modify the hypotheses, and iterating until we reach conclusions. Coding agents can speed up the pace at which we can write research code. - Meta has pivoted from its ope...
Read more

Google Maps Presents: Traffic Trauma Live

Image
Cartoon co-created with Copilot. See more of my AI co-creations Google Maps officially rolled out AI-powered audio notifications for accident-prone areas (Sample: "Accident prone area for the next 500 meters") and proactive congestion alerts  ("Congestion 1km ahead. You're on the best route") in India in November, 2025. While basic traffic alerts have been around for years, this specific update introduced "Accident-Prone Area Alerts" and "Proactive Traffic Alerts" as part of a major integration of Gemini AI into the app. Maps also displays official speed limits beside the in-app speedometer. India was a pioneer market for these specific AI-driven safety features. The proactive and accident-specific audio notifications were part of a specialized " India-First " push. What was "India-First"? Accident-Prone Area Alerts: This feature was developed specifically for the Indian road context. Google collaborated directly with loca...
Read more

Uncle Bob vs. Grady Booch: Rethinking Code Reviews in the Age of AI

Image
In response to a question about the feasibility of effective code reviews for large (e.g., 500-line) AI-generated PRs like those from Claude, especially when reviewers lack deep codebase familiarity in new projects or fast-paced environments, Uncle Bob Martin and Grady Booch have contrasting views Uncle Bob Martin advocates metrics-based oversight (test coverage, complexity, dependencies) and higher-level management over line-by-line AI code review, while Grady Booch stresses manual verification for vulnerabilities, dead code, and performance factors. Uncle Bob Martin : " I don’t review code written by agents . I measure things like test coverage, dependency structure, cyclomatic complexity, module sizes, mutation testing, etc.  Much can be inferred about the quality of the code from those metrics. The code itself I leave to the AI.  Humans are slow at code. To get productivity we humans need to disengage from code and manage from a higher level." Grady Booch : "Unlike B...
Read more