I think we might finally move away from subscriptions in software. You don't expect your toaster company to pay your electricity bill for you, but we do that for our software apps. I think that the rising token costs will adjust the way we consume software. Consumer behavior will shift to pay for AI as utility and software apps will compete to be more token efficient.
I'm approaching this problem via recursive task decomposition [0] mostly for coding but once that is polished, I can imagine it generalizing to other workloads. One idea I've been meaning to explore for a while is using the principles from Erlang/OTP [1] and mapping it onto multi-agent systems. Some nodes are supervisor trees that coordinate and handle restarts, gracefully degrade, are easy to observe + steer.
Yes! Contrary to the anthropomorphized subagents, I view them as ways of managing context primarily. I'm exploring this idea in Scope[0] to have observable subagents that recursively break down the task to avoid having to compact. One thing I haven't been able to figure out is how to evaluate/improve this planning step. I am using markdown files to encode heuristics for planning but it feels too unstructured for me to measure. Would love it if someone pointed me to some existing literature/projects around this idea!
It's pretty much that with the UX. Instead of opening a terminal - you can hit a shortcut (option + L) and select the time in a GUI within the browser!
I’m building Skim: https://www.justskim.in/,
A PWA that lets you read books as auto-swiping, short-form content on mobile. I use it to replace watching YouTube Shorts or Instagram with reading in the same form factor. It works offline and is entirely client-side.
This weekend I’m working on making the parsing more robust. The most common friction I’ve heard is that downloading books elsewhere and importing them into the app is distracting. I’m torn between expanding it to include a peer-to-peer book exchange or turning it into an RSS feed reader.