Comments by "Tony Wilson" (@tonywilson4713) on "How Uber u0026 Lyft ROB Their Workers With Scam Law" video.

@mikejohnson555 A better description might have been "when they make quantum computers work." First if you want if mention NASA & supercomputers together, I did aerospace at U. Illinois where they had access to the first generations of Cray supercomputers as I was doing my degree. All that fancy CFD you see these days was started by one of my professors. I knew a guy via the swimming team who was doing his masters on compilers for parallel processing systems. These days I work in industrial control systems and robotics. So I am fairly well aware of where we've been and what we now have and what the advances have been. There's a couple of great vids on why they believe quantum computers will change everything. Compared to even todays supercomputers they'll be so far ahead in data processing throughput its almost impossible to compare. Its like comparing an abacus to the latest CPUs. What made the Crays so powerful in their day was the parallel processing in conjunction with vector processing. Todays CPUs have inherited those methods but they are all limited by physical constraints that doesn't apply to quantum processing. As for AI that's fantasy. There has been so much idiotic crap put forward via journalists and other proponents that they deserve to be slapped. Yes they have lots of sweet and interesting algorithms that mimic various human traits but as for anything like actual intelligence its idiotic. Go look for the honest TED & TEDx talks on the subject. There've been plenty of people in that field coming out and trying to tell people what AI is and what its not. In engineering we use words from the common lexicon that mean things in our field but the meaning just not the same. We talk about "teaching a robot" where we teach the robot where we want to it move. We talk about process controllers (PID & Fuzzy logic) that "learn" how to run the process better over time. BUT those things are totally different to teaching and learning in humans. I have actually done that kind of work, including at the code level. I find it quite annoying that people have blown what AI is out of all proportion.
2
You don't have to worry about that anymore. Uber sold the entire division for self driving cars just as every major manufacturer sold off or cancelled their work in that area as well. It was very quietly done a couple of years ago when they realised that being able to prove the onboard computers would be safe was next to impossible. Its comes down to all of the what ifs. What if A happens, what if B happens,....... and there's more variations than there are alphabets. What they basically concluded was that true driverless cars would need a supercomputer onboard. So until they work out how to build a supercomputer the size of a shoe box its not going to happen. Where driverless vehicles might actually happen is in long haul trucking when the trucks are on freeways with no traffic lights, pedestrians and lots of other stuff. Which would make it a lot more like an autopilot for a passenger jet. The pilot's are onboard the whole time. They do the take off, landing and taxiing, while the autopilot does the long boring part. There will still be drivers who do all the driving around cities and towns but once they are on the freeway/expressway/interstate they'll switch to autopilot. I'm an engineer who works in industrial control systems (including safety systems). I also have a pilots license and there's a misconception that having an autopilot for a car or truck is the same as an autopilot for a plane. The auto pilot in a plane pretty much only has to deal with speed, altitude and direction. A car or truck has to deal with the road, lanes, other vehicles, pedestrians, animals, traffic lights and millions of other things that are all changing every second. Its actually highlighted just how extraordinary the human brain is to do what it does. It also highlights how dangerous even a slightly distracted driver is.
1