Post

every time

This post is unavailable.

254 reposts668 likes

Only time I felt positively with a “surveillance” AI project was the one for seizure detection alerts using a watch app’s vitals and temp plus visual signs and then a proposal to make it as a mini cam on a service dog. I can see where it would go *absolutely* wrong but I appreciate the sentiment.

I'm not familiar with this project, but that also sounds like machine learning and not an LLM!

It was a proposal at a psychology conference about two years ago. Couldn’t tell you the university though but I think it was a southern US university. I thought it had merit at least but security of health information was the biggest criticism and halted the progress last I ever heard of it.

yea, i fucking hate how tech bros co opted the term AI when there's actually cool shit with machine learning technology that you can do, but then people are like "fuck AI" because they can't differentiate between generative AI and LLMs and actually cool machine learning projects.