Astra's product judgment feels like a step back, user reverts to Sol
OnlyDaikon5492 · reddit · 2026-09-06
A long-time AI design/coding agent user shares hands-on experience: as models got smarter he gradually delegated more, and with the previous model Sol he could treat it like a junior product designer — set direction and trust its decisions.
But Astra disappointed him: its product instinct and intent understanding feel like a clear step backward, forcing him back to highly prescriptive prompting. He has mostly reverted to Sol and asks if others agree.
More from Apps
- Redditor builds free AskSary creative studio with game engine GPT can play and patch live — Beneficial-Cow-7408 · 2026-09-06
- Generative AI point-and-click detective game built without a multimillion-dollar budget — tristanbob · 2026-09-06
- Redditor uses coding agent to finish an obscure 18xx Pennsylvania railroad game — natep1098 · 2026-09-06
- 7 prompting principles for GPT-6 Astra and next-gen models: roles, context, self-review — goyalshaliniuk · 2026-09-06
- VoibeAI's first month: 507 users, 89,791 dictations, median just 15 words and 8 seconds — ayushtweetshere · 2026-09-06
- Has anyone fed their whole genome sequencing to Astra for diagnosis? — gajesh · 2026-09-06