• CanadaPlus@futurology.today
    link
    fedilink
    English
    arrow-up
    1
    ·
    edit-2
    1 year ago

    I don’t know. After a skim I don’t see a lot of supporting evidence (basically just “people are working on it”). It’s still hard to turn a physical interaction into a dataset you can run through 10,000 times a second in serial.

    It’s pretty telling that self-driving cars still suck years and billions of dollars later, and that’s easily the best common physical task for automation.