@antonosika Great addition. I fear however everyone is converging into the product. Generalist AI assistant for all computer work. As a consumer, it’s becoming hard to differentiate and harder to decide what to buy.
@realBigBrainAI Then who buys the AI? Demand is ultimately created and driven by humans. And if humans cannot buy anything, there’s no need for AI to begin with. There will be a viscous circular unemployment and growth cycle that limits the capacity for AI to diffuse.
@andrewchen I agree. I now spend more time in Claude Code than Excel, and have reframed how I work. The next big hurdle is simplifying code management (e.g., storage, versioning, sharing, collaborating) for the average analyst you refer to. Those things are convenient with spreadsheets.
@aakashgupta Yet despite all that, I bet the M365 Cowork will somehow be a worse experience and overall product than Claude Cowork, the exact tech it tried to replicate.
@karpathy I agree. At this point, every “app” should have an agent and be fluid enough to allow max customization on the fly. Let’s call it a template app —> extremely personalized.
@claudeai Feature requests for chat that can probably take 10 minutes for you guys to implement.
(1) collapsible chat folders
(2) hidden chats… maybe option to also hide/remove from memory or just hidden folder
Please!
@WestsideLAGuy I think it goes even further. The question may become, "what have you built in undergrad" rather than just the historical output of undergrad (i.e., scores, knowledge of the domain, and perceived raw intellgience).
And yes, also agree that MBA prestige will decline.
@kloss_xyz Don’t be shocked if they release a new OpenClaw dominating product in <1 month. Likely in a way that the enterprise will actually accept.