I find it more than a little bit implausible that facial recognition accuracy is good enough to pluck a face out of a billion people and accurately identify a single individual to the point that a fine would be automatically issued, especially without corroborating evidence that the person fined was even there.
AI Mistakes Bus-Side Ad for Famous CEO, Charges Her With Jaywalking
201–210 of 416 posts
Re: AI Mistakes Bus-Side Ad for Famous CEO, Charges Her With Jaywalking
#202Maybe calling a machine that has just demonstrated its profound stupidity an “Artificial Intelligence” is part of the problem. This crappy tech is overhyped and overrated.
Re: AI Mistakes Bus-Side Ad for Famous CEO, Charges Her With Jaywalking
#203Earlier quoted context omitted.
All cities, major AND minor (and villages). You wouldn't except this to happen to the equivalent of the New Jersey Turnpike, of course. So, on high speed highways, on the other hand, you're often not allowed to randomly cross, except in specific areas.
Interesting. Thanks for the response!
In some countries like Italy or France it is legal as long as there's no explicit zebra crossing nearby (e.g. without 50 or 100 meters).
In most of the countries, crossing the equivalent of highways or interstates is illegal (except in designated overpasses and such).
Regarding the question about what happens to drivers, here's an example for what's the case in Belgium "any physical damage to a pedestrian caused by a traffic accident shall be compensated by the insurance of the drivers involved, regardless of the responsibility of the pedestrian, except if that pedestrian is over 14 and wanted the accident and its consequences to occur."
Re: AI Mistakes Bus-Side Ad for Famous CEO, Charges Her With Jaywalking
#204I once read about a proposed 1$ fee for mistakes like that. For individuals, it's just one dollar. For companies and especially the state, it can cost 1000$ and more in processing fees altogether.
Re: AI Mistakes Bus-Side Ad for Famous CEO, Charges Her With Jaywalking
#205Earlier quoted context omitted.
Out of curiosity could you explain what you mean? I'm not nearly old enough to remember prior to ATMs. But I love ATMs. I only have to go inside the bank to do things once every few years. Everything else is on the phone or ATM. And the people I deal with on the phone are almost always very helpful and knowledgeable. I know people have a lot of trouble with some of the big banks like Bank of America and Chase. But th…
I'm not nearly old enough to remember prior to ATMs. I am. When ATM's started to become a thing, we were told that with ATM's automating the basics, the bankers and tellers would have more time for more personalized service, Banks would lower fees through cost savings, and we'd all have higher interest rates on our savings because the banks would save so much money. None of those things happened. Things just got wors…
Re: AI Mistakes Bus-Side Ad for Famous CEO, Charges Her With Jaywalking
#206Yes, but even hidden, it's probably still going to appear quite significant, and I do believe that at some point, the fine size becomes nonsensical. But then again, I do not think that who you are or what you have should change the magnitude of the crime you commit. I don't think it is sensible that I should get greater fines because I chose to bust my ass and suffer from stress to gain more for my family, whereas I…
Re: AI Mistakes Bus-Side Ad for Famous CEO, Charges Her With Jaywalking
#207Earlier quoted context omitted.
>I am lucky. I can pay that. It sucks but some people would be out food money if they suddenly had to pay $500 to drive again. Absolutely, court fees and fines disproportionately punish those with low incomes. Increasing enforcement of low level violations by surveillance tech and/or AI is going to crush a lot of people if we don't change that fact. https://www.npr.org/2014/05/19/312158516/increasing-court-fe... http…
Finland, I believe, scales fines proportionally to your income. Not perfect (the rich can more easily hide their income than the not so rich), but a good start.
Re: AI Mistakes Bus-Side Ad for Famous CEO, Charges Her With Jaywalking
#208Earlier quoted context omitted.
Finland, I believe, scales fines proportionally to your income. Not perfect (the rich can more easily hide their income than the not so rich), but a good start.
I think if it scales to a maximum amount its fine especially for driving related violations. For example nobody needs to be paying 25 grand for a speeding violation thats absurd regardless of income. But its interesting. I believe in Mexico its based on your hourly wage and x amount of hours. So if you have no job and you get a ticket you owe nothing, at least thats how it was when my sister got a ticket in Guadalaja…
Re: AI Mistakes Bus-Side Ad for Famous CEO, Charges Her With Jaywalking
#209Earlier quoted context omitted.
In the case of jaywalking, the answer is simple. You just make it legal. Cars can slow down. For routes that need to be able to move at high speed, you build in grade separation. Jaywalking was just a "growth hack" that needs to be reverted. Society needed fast transportation, and it was developed at any cost. The cost is now too high, so we just roll it back. No torture required....
> In the case of jaywalking, the answer is simple. You just make it legal. Cars can slow down. Mmmm I dunno about that. Drivers have minimum reaction times and cars have stopping distances. Giving pedestrians the right-of-way in every situation seems like a recipe for a lot of collisions. It could work, but I think it would involve re-designing countless intersections to provide higher visibility of pedestrians for d…
Re: AI Mistakes Bus-Side Ad for Famous CEO, Charges Her With Jaywalking
#210Earlier quoted context omitted.
A machine can be tested for bias and measures taken against such bias are certainly followed. The same does not hold for humans. Detecting the bias (down to the person level) is very hard. Going on to remedy it is even harder. You can't (or at least shouldn't) reprogram people. In other words, you can change a racist machine, but not a racist cop.
> A machine can be tested for bias and measures taken against such bias are certainly followed. To what degree? Based on first-order data, some populations really do commit more crime. However, we know that background/poverty is the real, underlying cause. And even if we fed enough data to an algorithm to avoid that bias, avoiding the bias isn't enough. What you really want to do is tackle the cause. If the AI is a b…
Then, we can explicitly correct for that quantified bias. This loses the whole 'AI is mathematics, and mathematics is unbiased' appeal, because we are explicitly adding a bias to counter a pre-existing one.
My point is, this (or more advanced methods) is possible with AI. You can't do anything like this with cops.