They mentioned "microVM" in the live stream. Notably there's no browser or internet access. It makes sense, running specialized Firecracker/Unikraft/etc microkernels is way faster and cheaper so you can scale it up. But there will be a big technical scalability difficulty jump from this to the "agents with their own computers". ChatGPT Operator already does have a browser, so they definitely can do this, but I imagine the demand is orders of magnitudes different.
There must be room for a Modal/Cloudflare/etc infrastructure company that focuses only on providing full-fledged computer environments specifically for AI with forking/snapshotting (pause/resume), screen access, human-in-the-loop support, and so forth, and it would be very lucrative. We have browser-use, etc, but they don't (yet) capture the whole flow.
It's not our only focus at Modal but it's a big focus![1] Code agents are the killer use case for LLMs right now, and this complements our GPU inference and training capabilities.
I'm quietly betting that agents increase the leverage of deterministic, reproducible devbox tech (eg. Nix, lockfiles, package mirroring), and this will end up being a huge win for us human engineers too.
There must be room for a Modal/Cloudflare/etc infrastructure company that focuses only on providing full-fledged computer environments specifically for AI with forking/snapshotting (pause/resume), screen access, human-in-the-loop support, and so forth, and it would be very lucrative. We have browser-use, etc, but they don't (yet) capture the whole flow.