Comment by tyleo

> constantly clearing context and starting from scratch

And using multiple models. And cycling through roles/persona - brainstorming, discussion, critique, etc. And thinking in agentic terms - all agents, including humans, are badly in need of output review steps.

Yesterday I saw copilot, from scratch, instructed to critique an idea, go straight to enthusiastic "major innovation!" nuttery. If one lived inside that as a bubble, rather than it being a few minute transient, followed by another model listing difficulties... I have no trouble imagining people going nutter.

Context matters with nuttery. You can have someone respected in their subfield, who in a different subfield, where they don't actually work, and thus lack the incentives and relationships and reality checks that would create, be... far less grounded there.

Group-think bubbles as a service - now available in a new low-friction self-adjusting one-person size!

I mean, it's an obvious difference in the primary use cases - if you explicitly want an isolated answer, you might clear context and start from scratch, and if you explicitly want a discussion with a persistent companion (as these people did before any of that psychosis started) then you won't do that.

Yes but I think it's generally how non-technical people use it

Also these people are using the memory features. In technical circles I’ve seen people made fun of for having it enabled. It’s considered “cringe”.