Disruptive technology

Paul Brubacher

In Memoriam
Messages
296
Location
outside of Toronto, Ont
This weekend I heard an interesting talk on disruptive technology. It was all about artificial intelligence and automation with an emphasis on self-driving cars to reduce road congestion.

Automation is becoming cheaper, look at the CNC machines and 3D printers. It now costs $3.10 per hour to operate a robot while a worker in China earns $3.00 an hour. So it looks as if most manufacturing will become automated.
Apparently a company would make more money selling plans for $5.00 to print your own Lincoln Logs than to manufacture the logs and sell them to a retailer.

Slightly over 50% of cars entering the downtown core are for government and banking. So time shifting and working from home would be good methods of reducing traffic congestion.

About the banking, there has been an estimate that 2 million jobs will be lost in North America if or when digital currency becomes predominant.

He said that his father’s new car has automated stopping. His father’s comment was that he was only 77 years old and could still drive. Once while driving he spilled coffee in his lap and the car stop automatically. He hadn’t notice that the traffic had stopped.
The self driving cars can be safer because they have thousands of inputs. The one drawback is that they will not break any rules of the road, so if a child darts in front of the car, it will not swerve across a double yellow line.
 
The one drawback is that they will not break any rules of the road, so if a child darts in front of the car, it will not swerve across a double yellow line.

That may be the current position but it won't be the end state. I doubt that any manufacturer is planning to put a car in the wild that is so rule bound that it would cause an injury rather than a "minor" infraction.

And that's not even an interesting dilemna!

What about one person on one side of the road and two on the other.
Whatever you do somebody gets hurt.
If it was a person and they make a decision we consider "wrong" in that situation do we cut them slack because "no time to think".
Do we give a computer the same grace?
There are a billion questions to be answered about these big new technologies and we don't even know what most of them are yet. But for me - I hope the next new car I drive has a forward facing radar that keeps me from rear ending somebody in a moment of distraction. And the one after that I hope has a hammock and an alarm clock so that I can sleep until I get where I'm going.

What about companion robot working.
I do lots of tasks where I spend 30% percent of the time picking 1 screw out of a box and presenting it the right way in to a hole. The entire task is not currently (hopefully ever) amenable to simple robotics but if I could get a $3000 robot that could reliably pick up the right screw and present it to me in the right orientation it would pay for itself real quick.
Not replacing my labour - enhancing it.
Kind of like my circular saw robot and my belt sander robot do.

I've recently acquired - for pocket money really - a laser engraver. The kind of machine that a few years ago had a $5000+ entry price.
That technology lets me do a bunch of little stuff that means that my waste stream is now an alternative product stream.
Those alternative products - personalised knick knacks mostly - are going ino a price band that recently has been occupied exclusively by the kind of very cheap products that are made in the far east and imported in volume.
A market that 12 months ago was useless to me is now filling in the gaps around my main business and making both bits more viable.

All technology is disruptive - from fire on demand onwards.


Bring it on.
 
There is good and bad in everything. BTW, there are methods to orient a screw with automation.

People can make abstract decisions relatively quickly based on the situation at hand. AI needs to have a situation programmed in, interestingly, by a human.

People can make the wrong decision in an emergency situation, whereas an AI could choose the correct course of action.

Still, Personally, I am not ready for AI, when it comes to human safety.
 
Yes. Leo is right. There is good and bad in everything. No truer words were ever spoken. I know automation has it's place. But, when you have a computer driving a car, that is not setting well with me at all. I don't even like a computer backing a car into a parking space. If you can't drive, don't get behind the wheel. Simple. Period. A computer driving a car to me is a public nuisance and needs to be done away with. Technology is fine as long as we remember it was meant to be a tool, not a crutch or a human replacement.
 
This has been a major concern in the aviation industry for 25+years. When I was flying high technology aircraft I ran into people who "got lost in the magic" and forgot to fly the airplane. There was also the problem of getting so dependent on the technology that you forgot basic airmanship - flying the aircraft first and what do do when the magic fails. I've got lists of accidents that covered all possible problems. The scary thing is that the aviators involved in these accidents were highly trained on the technology. Drivers will never get the type of training that pilots get in technology management. There has already been one Tesla autopilot accident where the driver set it and then thought he could read a book while the car took him where he wanted to go. When the autopilot got confused over what it saw and hit a truck, he died. Technology must be monitored by the operator to make sure it's doing what you want it to do.

Mixing self-driving with human driven cars is not a good fix. You can program a computer to do X in Y situation but the Mark 1, Mod 0 human being is unpredictable at best and irrational at worst. I don't see how they can mix the two safely on a city street. I've never been a fan of autonomous self-driven vehicles on a public roadway. Too many variables that have to be accounted for.

What I think would be great would be technology highways. Instead of mixing self-driven and human-driven cars on the same road, the only ones on the road would be high tech autos. Instead of being autonomous, they would have to be capable of connecting to satellites such as GPS. The satellites would set follow distance, speed, and other factors that integrated high tech traffic. Program in where you want to go and the satellites could read your fuel state, tire pressure, and weather to allow for refueling, etc. If you tried to zip in and out of traffic, the satellite would steer you to the side of the road and shut you down. Those vehicles which couldn't connect with the satellites would not be allowed on the highway. The benefit of this would be higher speeds. Traffic could move at 75-80 safely because everyone would be deconflicted from other traffic. Once you got off the high tech highway, though, you'd be back on the street driving your own car with everyone else.

In aviation we discovered new human factors traps with technology every day. Taking marginally trained people and putting high tech in their hands is a recipe for disaster. We already kill 35000-40000 people on US highways every year right now. Self-driving cars are not going to lower that number unless everyone has one.
 
Self driving cars get to decide which one is sacrificed.

Not always, Darwin has some say:
https://www.usatoday.com/story/money/cars/2017/06/20/tesla-self-driving-car-crash/411516001/

I can't remember where it was, but there was a crash that involved a automated car, that the car sensors were blinded, due to suns reflection off of the other vehicle. Humans due that as well (keep driving when the sun blinds you).
Then, there will be times that vehicles with act as a deer:
https://www.theguardian.com/technol...ter-las-vegas-launch-truck-autonomous-vehicle
Not programmed to back up, no ability to adapt.
 
there was a crash that involved a automated car, that the car sensors were blinded, due to suns reflection off of the other vehicle. Humans due that as well (keep driving when the sun blinds you).

That was the Tesla accident I mentioned. A tractor/trailer turned in front of it. Big silver trailer registered as sunlight. Driver was reading up til impact.
 
Top