What do we fundamentally misunderstand about AI? Great demos, mediocre production systems
Conscious_Belt_8444 · reddit · 2026-09-03
A veteran software practitioner asks: why do so many good AI demos turn into mediocre production systems?
His take: teams jump straight into RAG, agents, fine-tuning, and vector DBs, then six months later discover the costly problems are the boring ones — bad data, poor evaluation, unclear requirements, no ownership of output, and models confidently doing the wrong thing.
He's looking for real production war stories, not textbook answers.
More from AGI Musings
- Engineer: anthropomorphic AI descriptions on podcasts are misleading and non-predictive — jamesdouma · 2026-09-03
- Sam Altman: One California Almond Uses More Water Than 38,000 ChatGPT Queries — Polymarket · 2026-09-03
- kalomaze: compute multipliers don't matter if objective specification is unsolved — kalomaze · 2026-09-03
- World Bank Nigeria RCT: GPT-4 tutoring boosted learning by 0.31 SD, worth 2 years of schooling — alexvoica · 2026-09-03
- Gary Marcus backs proposal to halt frontier AI training runs worldwide — GaryMarcus · 2026-09-03
- Tao: proofs verified by Lean but incomprehensible to humans shouldn't be published — SoloGen · 2026-09-03