Follow

sometimes i think about how Stanford had an AI in Medicine experimentation system in the 70s/80s

they were a little too early

@gewt true, but that's still pretty fast and a lot of working memory compared to people.

@gewt still though, that probably wouldn't scale up to a whole hospital of usage

@KitRedgrave definitely not - you're also not gonna fit a complex medical compendium - unless you manage to attach an entire winchester disk system to an SA-10 somehow

@KitRedgrave error - you'd want later drives than an original Winchester - you'd want later IBM DASD - which SUMEX-AI used

@gewt there's AI and then there's what people are calling AI these days
@gewt I would trust ELIZA + a reference textbook more than 'machine learning'

@petra i saw a story once where someone encountered eliza in emacs and it helped them with their depression, so i others do too apparently

@gewt I'm not surprised! I used to spend ages talking to chatbots
Sign in to participate in the conversation
San Junipero, from gimme-sympathy.org

gewt's private instance. May occasionally reference television.