Recent searches
Search options
When worrying about the harms from deploying AI, it’s issues like this that have caused and will cause the most harm. Not the AI suddenly turning into the machines from the Matrix and enslaving humanity.
The system that was meant to analyze income and health information to automatically determine eligibility for benefits simply didn’t work and often failed to load the correct data to analyze. 
@carnage4life
The synthetic machines perform pattern synthesis on crystallized social relations: https://www.superversive.co/blog/crystallized-social-relations
@carnage4life i wonder if this was why I kept getting letters saying I need to renew my medicaid information or I would be kicked off and then some human would come along later and say no that was a mistake that wasn't supposed to apply to me
@carnage4life why pick between those two narratives? "A.I. tortures thousands" would be a perfectly accurate and probably more useful headline. Large amounts of people were condemned to unnescesary suffering by this system, some likely abused to the point of death.
@carnage4life yes we've already seen this evil misuse of algorithms in Australia - Google "robodebt" for details.
@carnage4life How come when something is shiny magic it's "AI", but when it breaks, it's an "algorithmic system"
@carnage4life denying benefits sounds like it's working exactly as intended.
@carnage4life @risottobias I suspect Tennessee is just one of many similar cases where algorithms are utterly failing harming the most vulnerable the most (and in many cases harming folks who don’t have the resources to push back)
Here in CA I’m fairly sure (from direct personal experience) that there are some underlying glitches inside of the otherwise good CoveredCA (and the various county led MediCare and MediCal systems). Leading to mistakes and delays and more referrals back and forth
my company got over 100 letters, AI generated, demanding money back (millions)
so my company has to have humans review each and every (bogus) AI generated demand and respond to each individually in the hopes that this will stop the payment demands (it won't) until they get a human at the demanding company to review each and every response
the whole thing is a scam and in the end will cost much more to defend against because AI is a scam
It’s not the machine. Its the people who designed and operated the machine.
@carnage4life This is probably going to be one of the major explanations of why identify as conservative instead of progressive. My main prior example is the “Urban Development” of Tulsa’s Greenwood, which took the systemic bias against the neighborhood’s residents and poured it into concrete.
Upper class folks can hire lawyers to fight for desired outcomes when the rules constantly change. If we’re making changes, they have to outpace elite advantage, or it’ll just marginalize harder.