Elliot Betancourt

Code, AI, and the Craft of Building

Exploring software engineering, artificial intelligence, and the tools reshaping how we create.

LATEST JUL 31, 2026

The Coding Agent Exception, Six Months Later: A Scorecard for Your Next Project

Back in March I wrote about why coding agents are a best-case scenario, not a template. This week I turned that reasoning into six questions you can run against your next candidate project before you build it.

Elliot Betancourt

Recent Articles

JUL 24 2026

A Frontier Lab Just Became a PE-Backed Services Firm

Anthropic just co-launched a $1.5B enterprise AI services firm with Blackstone and Hellman & Friedman. What that means for independents already doi...

JUL 17 2026

The Wrapper Is Getting Commoditized. The Judgment Isn't.

OpenAI shipped ChatGPT Work a week after Claude Cowork kept shipping. The long argument behind Tuesday's post: what happens to bespoke agent work w...

JUL 04 2026

Why Coding Agents Are the Exception — And What That Tells You About Every Other AI Use Case

Coding agents genuinely work. That success is becoming a dangerous data point.

JUN 20 2026

The Model-Centric Exception: What Coding Agents Actually Prove

Anthropic confirmed 80%+ of their code is now AI-written. The wrong lesson is everywhere. Coding agents work not because the model is ready for aut...

JUN 12 2026

Why the Smartest AI Teams Are Moving from Model-Centric to Systems-Level Thinking

Most AI agent failures aren't model failures — they're architecture failures. Here's what the shift to systems-level thinking actually looks like i...

VIEW ALL ARTICLES