a few bits from Janelle Shane’s You Look Like a Thing and I Love You
“Researchers have discovered that something as seemingly insignificant a a small sticker can make an image recognition AI think a gun is a toaster …” p4
And vice versa, I presume. Well, that’s not alarming at all.
“A team at Stanford University once trained an AI to tell the difference between pictures of healthy skin and pictures of skin cancer. After the researchers trained their AI, however, they discovered that they had inadvertently trained [it to be] a ruler detector instead—many of the tumors in their training data had been photographed next to rulers for scale.” p23
So when your GP asks your permission to use AI assistance … just say no.
“If you give a job-candidate-screening AI biased data to learn from (which you almost certainly did, unless you did a lot of work to scrub bias from the data), then you also give it a convenient shortcut to improve its accuracy at predicting the ‘best’ candidate: prefer white men.” p26-27
Yeah. Garbage in, garbage out.
“AI has the approximate brainpower of a worm.”
“AI does not really understand the problem you want it to solve.”
“AI will take the path of least resistance.”
