Let’s say the error rate is 0.1%. Pretty low, right. But that’s one mistake per thousand flights. Are they really okay with one plane out of a thousand potentially crashing? There are certain industries and jobs where AI simply cannot and should not be used.
Each day, about 100-120 people die in car crashes in America.
Over 45,000 planes fly in America every day, and over 5000 are in the air at any given moment. With a crash rate of 1 out of a thousand, we’d be having multiple plane crashes, with thousands of people killed, every day. One plane crash could easily match or surpass that daily car crash number, and we’d be having multiple plane crashes per day.
1 out of a thousand? I’d never fly again. NOBODY would ever fly again.
Normally, I would scoff at being worried about airborne debris, but if 1 out of 1000 were crashing, and there were 45k flights a day, that’s enough crashes to worry about.
The vast majority of those crashes would be around airports, though, so just keep away from the airports, and your chance of being clobbered by a black box goes down significantly.
It’s almost comical to think about major airports having a half dozen crashes a day. At least the AI won’t have any trouble sleeping at night.
Even further: the biggest problem with AI and thus the biggest decider on its suitability or not for something is that its distribution of failure in terms of consequence is uniform rather than it being more likely to err in ways with few or less grevious consequences than in ways with more or worse consequences.
In other words, unlike humans who activelly try and avoid making the nastiest and deadly mistakes, when AI fails, it can fail just as easilly in the most horrible and deadly ways as it can in the most minor of ways.
That’s why you have lots of instances of LLMs giving what for humans are obviously dangerous advice like telling people to put glue on pizza to make it look good or those with suicidal thoughts to kill themselves - unlike humans AI has no mechanism to detect “obviously dangerous” on an output it’s about to produce and generate a different output instead.
This is why using AI to generate fluff filling for e-mails is fine but it’s not fine in systems were errors can easilly cost lives.
But think of the insurance people! Look at how many insurances are waiting to be denied and robbed!
More importantly, we can justify every other profit increase, because our economies are built on literal exploitation just as they did a couple hundred years prior!
Modern exploiting problems require modern idol solutions.
Sadly there is part of the population that will view that as a valid argument. Faux News, news max, OAN and all the conservative talk radio will feed it to them
Let’s say the error rate is 0.1%. Pretty low, right. But that’s one mistake per thousand flights. Are they really okay with one plane out of a thousand potentially crashing? There are certain industries and jobs where AI simply cannot and should not be used.
Each day, about 100-120 people die in car crashes in America.
Over 45,000 planes fly in America every day, and over 5000 are in the air at any given moment. With a crash rate of 1 out of a thousand, we’d be having multiple plane crashes, with thousands of people killed, every day. One plane crash could easily match or surpass that daily car crash number, and we’d be having multiple plane crashes per day.
1 out of a thousand? I’d never fly again. NOBODY would ever fly again.
The worst part would be that it doesn’t matter if you fly or not - as long as a plane can fly above you, you’re at risk. None of us are safe.
Normally, I would scoff at being worried about airborne debris, but if 1 out of 1000 were crashing, and there were 45k flights a day, that’s enough crashes to worry about.
The vast majority of those crashes would be around airports, though, so just keep away from the airports, and your chance of being clobbered by a black box goes down significantly.
It’s almost comical to think about major airports having a half dozen crashes a day. At least the AI won’t have any trouble sleeping at night.
Even further: the biggest problem with AI and thus the biggest decider on its suitability or not for something is that its distribution of failure in terms of consequence is uniform rather than it being more likely to err in ways with few or less grevious consequences than in ways with more or worse consequences.
In other words, unlike humans who activelly try and avoid making the nastiest and deadly mistakes, when AI fails, it can fail just as easilly in the most horrible and deadly ways as it can in the most minor of ways.
That’s why you have lots of instances of LLMs giving what for humans are obviously dangerous advice like telling people to put glue on pizza to make it look good or those with suicidal thoughts to kill themselves - unlike humans AI has no mechanism to detect “obviously dangerous” on an output it’s about to produce and generate a different output instead.
This is why using AI to generate fluff filling for e-mails is fine but it’s not fine in systems were errors can easilly cost lives.
Sarcasm:
But think of the insurance people! Look at how many insurances are waiting to be denied and robbed!
More importantly, we can justify every other profit increase, because our economies are built on literal exploitation just as they did a couple hundred years prior!
Modern exploiting problems require modern idol solutions.
Sadly there is part of the population that will view that as a valid argument. Faux News, news max, OAN and all the conservative talk radio will feed it to them