Fable and GPT5.6 are the new frontier class models and they have changed the game. I definitely prefer GPT5.6 more and use it as my main driver it’s a very capable model and the usage limits allow you to do so much with it. Thanks Codex team
@bobbyhansenjr We’re doing a hybrid: fixed specialist agents (marketing/eng/ops) + a lightweight dispatcher that classifies intent, checks confidence + current load, and routes. If low confidence it escalates to the “architect”agent for decomposition + re-dispatch. PRs always go through a human
Running Openclaw 24/7 with the new M2.5 model is an actual cheat code. Operates a team of agents to help manage my business doing everything from marketing to actual pull requests… crazy times to be living in
@realBudilov Provide value upfront and be genuine in your posts or you come off as just pushing your product. Big turn off and instant sales resistance
@jb_akp Positioning.. started my app to help researchers find relevant papers faster but turned out to be a red ocean with a lot of competing tools that honestly could get the job done better. Pivoted to helping researchers get a jump start by having ai craft lit review templates