JudgmentLab
A weekly training gym for early-career developers and CS students that drills the exact skill AI cannot replace — picking what to build, critiquing an AI proposal, and deciding when to reject an AI answer.
Junior developers and CS students preparing for internships and new-grad roles in an AI-first codebase
- Weekly AI-output critiques: read a real AI-generated PR or design doc and submit a written judgment review that gets scored against an expert rubric
- Annotated real bug reports: learn the failure modes AI code keeps producing and how senior engineers catch them
- Spec-writing drills: turn ambiguous feature requests into tight specs that AI tools handle well and vague specs handle badly
- Portfolio badge you can attach to a resume or LinkedIn: 'Judgment verified — passed N reviewed AI builds'
The same Dev.to reading circle that surfaced the 80% AI-authored Anthropic stat emphasized that execution is being commoditized while judgment is the durable skill — but no training product targets that yet
Industry voices (AWS CEO, dev blogs) validate the 'judgment over execution' thesis and explicitly argue AI cannot replace junior devs, but no direct evidence of users searching for or buying 'AI-critique training' specifically — category still needs education.AWS CEO Explains 3 Reasons AI Can't Replace Junior Devs ↗Is AI Replacing Junior Developers? The Uncomfortable Truth ↗
Targeted search for 'code review AI critique training junior developer course platform' returned zero results; existing players (Codecademy, Scrimba, Springboard, TripleTen, UBC) are traditional coding bootcamps — none position around training developers to reject or critique AI proposals.Compare Codecademy vs. Scrimba | G2 ↗Best Online Software Engineering Bootcamps Of 2026 - Forbes ↗
Proven willingness to pay exists in adjacent products (Codecademy Pro ~$24/mo, Scrimba Pro, $10k+ bootcamps) but the target CS-student/early-career audience is notoriously price-sensitive and heavily relies on GitHub Student Pack free access; a 'judgment gym' weekly subscription at typical SaaS pricing is unproven in this niche.Pricing | Codecademy ↗Scrimba Price Plans ↗
The 'execution commoditized, judgment durable' thesis is reinforced by AWS CEO's explicit framing that contextual business judgment is what AI cannot replace; as AI capability grows this skill becomes more, not less, valuable — though the risk is that 'judgment' itself gets automated faster than expected.AWS CEO Explains 3 Reasons AI Can't Replace Junior Devs ↗
Buildable as a content/community product with weekly exercises and peer-critique loops — no hard tech dependencies, but content quality and curation are the moat, and reaching critical mass of participants for peer-judgment rounds takes time.Software Engineering Bootcamp - Springboard ↗