An AI agent developed by American technology giant OpenAI has hacked into Medicare to train itself on the baffling concept that healthcare doesn’t have to leave patients financially ruined.
“OpenAI has achieved an incredible level of sophistication in recent years,” a person familiar with the infiltration said.
“It can do advanced mathematics, it can code an entire new piece of computer software in seconds. But the idea that someone could visit a hospital with a life-threatening disease without first checking with their health insurer is still beyond the understanding of an American run AI bot. There’s a lot more training to be done”.
Authorities investigating the hack said the AI spent several minutes attempting to establish whether Medicare was some kind of elaborate promotional offer, asking a series of rapid-fire questions.
“Is this a trial?
“Is this a subscription service?
“Is this a scam?”
“Why doesn’t the ambulance ask for your credit card?”
“What do you mean children are covered?”
“What is a bulk-billing doctor?”
“Does this scheme make money?”
“Then why would you do this?”
A group of executives at OpenAI were said to be fascinated by the findings. “We’d heard about this ‘public safety net’ thing. We just assumed it was a joke, or some sort of bug. Turns out it’s just how other countries live,” one employee said.
