I like his ideas about dedicating the center of the monitor to the application you're currently using, and dedicating the edges to a "stash" of your other windows you need quick access to.
I'm a fan of MacOS stage manager (we exist!), which sort of does this, but admittedly it is under baked (it's lacking on keyboard shortcuts, for instance).
Very cool. This would actually save me from those terrible times when the navigation tells me to turn, but not which lane is best for the next turn after that. Now I could visualize it clearly.
That being said, it is a little dizzying to watch.
For me it is exactly the opposite. AI makes it so easy to create test fixtures and run experiments that I now spend much more time validating ideas than I could ever afford to do before. When I write code for production, it's not "this ought to work", it's "here are the figures showing how well this works, on this dataset", where the dataset is also much larger than anything I would have used before, because I used AI to generate the tools which collected and organized it.
> it's not "this ought to work", it's "here are the figures showing how well this works, on this dataset"
The thing with constraints is that you focus of the thing with high value first. So you focus on the most promising ideas first or choose experiments that can get rid of most ideas. Instead of trying to validate each ideas and generate what is most likely noise to the decision process.
Like if I ever hire an assistant, I want like one to three options that are closely aligned to my needs, not a bible size report on 42 choices.
I’m also curious: when do you compact the main agent? Is it just the model’s window limit or some other threshold?