Some users see financial gains from GPT-5.6's unrestricted Yolo mode while many others criticize its uncontrolled agent spawning, messy code output, and lack of edit controls and urge safer auto-review alternatives instead.
Based on 9 visible X reactions from 13 accounts; directional sample.
Ask a question below.
Published answers will appear here.
@_xjdr We aren't here yet, fable 5 high wrote absolutely atrocious entangled mess of a code I'm more and more moving away from coding agents back to some form of round based chat + tools to access the code base Anything more than trivialities is still just asking for tech debt
@_xjdr this is not a good change imo, yesterday I run 'ultra' it spawned like 8 agents and started doing bunch of stuff without stopping for four hours, my fans were blasting, I can't even review what it did because it touched every single thing
@_xjdr They’re all missing a auto-mode-excluding-edits I don’t want to approve every single cd or cat or whatever but I don’t trust you’ll listen to my “no edits” command either
@_xjdr i am worried for noumena if that is the 60% long term thesis, good luck tho!
@helscom @_xjdr im going to make so much money from this so, it's win for me
@AndrewGinns TIL! i will look into this now
my biggest take away from gpt5.6 so far (and fable to some extent as well) is that you only get a fraction of the value without running in --yolo mode (subagents that require approvals aren't any more useful than just running the prompts yourself) and as a staunch --yolo mode denier (legacy review every diff unc), i am finding this dichotomy difficult to cope with. i knew this was coming, and i've been preparing for it (its like 60% of the long term thesis behind http://code.noumena.com) but i didn't think it was already here.
@_xjdr would strongly encourage *not* running in yolo mode and instead run it with auto review instead https://learn.chatgpt.com/docs/sandboxing/auto-review
can confirm, this is a nice addition https://twitter.com/reach_vb/status/2075649924465590329
Some users see financial gains from GPT-5.6's unrestricted Yolo mode while many others criticize its uncontrolled agent spawning, messy code output, and lack of edit controls and urge safer auto-review alternatives instead.
Based on 9 visible X reactions from 13 accounts; directional sample.
Ask a question below.
Published answers will appear here.
@AndrewGinns TIL! i will look into this now