Ive been thinking for years that if we could put the (absolutely enormous) privacy concerns aside think of the environmental benefit of every major city in the world having an “AI” controlling the traffic lights and variable speed limits. Using numberplate recognition cameras and gps on every vehicle to optimise flow, reduce bottlenecks and minimise time spent in traffic.
That won’t work because you’re approaching the problem from the wrong angle; you’re trying to “fix” traffic by encouraging more traffic. If you want to improve car traffic the only possible solution is to make other forms of transport more appealing. It doesn’t really matter which form of transport you focus on, it could be trains, busses, bikes, walkability, etc; just as long as you ensure it’s as or more efficient than a car for the majority of journeys.
The only way to fix traffic is for there to be less traffic.
Well you arent wrong but its not like its a “pick one” situation. With the unbiased data from the AI you could optimise all forms of transport. If you can see that theres clearly a lot of people driving from point A to point B you can examine the why and implement better solutions.
Society wastes a great deal of time looking for the perfect solution while some good ones sit right under our nose. If the AI solution has a city of 1 million drivers saving 5 minutes each way on an average commute of an hour. Thats the equivalent of 166k cars not driving that day and everyone saves 10 minutes.
I would fund a truly fair AI and a very gentle, but firm, self replicating robot army to enforce it’s benevolent will on everyone.
So basically SkyNet, after I make a pointer arithmetic mistake.
Stupid integer overflow!
Ive been thinking for years that if we could put the (absolutely enormous) privacy concerns aside think of the environmental benefit of every major city in the world having an “AI” controlling the traffic lights and variable speed limits. Using numberplate recognition cameras and gps on every vehicle to optimise flow, reduce bottlenecks and minimise time spent in traffic.
That won’t work because you’re approaching the problem from the wrong angle; you’re trying to “fix” traffic by encouraging more traffic. If you want to improve car traffic the only possible solution is to make other forms of transport more appealing. It doesn’t really matter which form of transport you focus on, it could be trains, busses, bikes, walkability, etc; just as long as you ensure it’s as or more efficient than a car for the majority of journeys.
The only way to fix traffic is for there to be less traffic.
Well you arent wrong but its not like its a “pick one” situation. With the unbiased data from the AI you could optimise all forms of transport. If you can see that theres clearly a lot of people driving from point A to point B you can examine the why and implement better solutions.
Society wastes a great deal of time looking for the perfect solution while some good ones sit right under our nose. If the AI solution has a city of 1 million drivers saving 5 minutes each way on an average commute of an hour. Thats the equivalent of 166k cars not driving that day and everyone saves 10 minutes.