Feedback Control & The Coming Machine Revolution

While I do a lot of reading in this forum, I don't normally post threads, but I saw something last night that I thought was worthy. I did a quick search and found nothing, so here it is;

Last night while watching the tube, I saw a clip where this guy was pitching a ball to one of those miniature quad helicopters and the helicopter would bounce the ball back to him. In the video I saw, it appeared that this was all done in an autonomous fashion. That is to say, without being remotely controlled by the demonstrator and/or another person.

Then, they demonstrated how 3 of these miniature quads could work in unison, suspending a net beneath themselves while tossing and catching a ball placed within the net. Last night, they also demonstrated how this machine could fly around the room while balancing a glass of wine on top of itself. In the clip I found below, they demonstrate a slightly different balancing act with a pointer as well as some very amazing acrobatics that weren't included in the clip I saw last night.

I thought this was some pretty amazing stuff and I immediately ask myself how this was being done. This morning, I found the following video entitled "Raffaello D'Andrea at ZURICH.MINDS -- Feedback Control and the Coming Machine Revolution" where they explain how advances in technology have ushered in a whole new era with respect to what machines will be able to accomplish in the future, both good & bad.

Anyway, I thought I'd share it with you, so here it is;


I guess the bad side of these types of technological advances could be compared to what we're currently seeing with respect to autonomous killer drones and/or robots.

If this has already been posted, I apologize in advance.

This is the zygote of the terminator future. Amazing, really. The host of the video is very aware of how dangerous this potentially could be for the human race.

reply to post by hhcore


Yeah, I think you're right. Hence the warning at the end of the presentation.

On the other hand, it sure could eliminate a whole host of what are currently very laborious and/or dangerous task in our daily lives. With every new development, there will always be a balancing act between good & evil.

reply to post by Flatfish


This is why the elite won't need us any longer.

So we are not only building our own prisons, but custom designing the prison guards.

Justice can never be "wrong" when it is meted out based on an algorithm. It becomes black and white. Nevermind that you are having a baby and on your way to the hospital....speeding is prohibited.

Honestly the idea that machines will become aware terrifies me. I keep thinking of the book and movie I, Robot. As far as I am concerned we need to figure out a way to keep them from becoming aware. Or eventually we as a species will be wiped out by these machines. If we want to know what our future may look like, look at the game mass effect with the geth. Pretty likely future for us. Something needs to be done to stop them.

reply to post by Flatfish


I actually took the time to watch the video and I didn't regret it. Reaffaello D'Andrea's words at the end are quite honest and I think he is right.

Technology will (or already has) become an own branch of evolution and we will develop it to perfection until it doesn't need us anymore. Meaning that our tech will be intelligent and able to replicate autonomously one day. It's the natural way of things to evolve, it bears great chances and dangers. But we have no choice and can't close Pandora's Box anymore. We'll have to make the best out of it.

Another thought I'd like to convey here is that, one day, we'll surely have the technology to produce intelligent machines, in terms of computational power. They'll be so complex that we will assume they are conscious, they'll perfectly mimic the calculations that our brains perform, but in a much faster way and a billionfold more efficient. And once they'll have that intelligence, we will probably lose control over them. Intelligence requires liberty. It requires freedom of action and thought ... and any limits will be overcome, that's the way it is with humans, and that's the way it will be with machines in a distant (or not so distant) future.

You know I've been thinking about this and I withdraw my previous comments. While the thought of machine life that is aware still scares me, if we offer them liberty, maybe we can exist peacefully with them? Does anyone think that is possible?

Great video, well worth a watch and a great example of how technology has improved over the last 10 to 15 years.

The thing that has changed the least is control theory - the vast majority of control systems still utilise PID control Wiki which has been used for around a hundred years.

one day.....

The potential is there, but I wonder if humans wont just destroy themselves before the machines get a chance to. We have machines monitoring everything everywhere in real time now as we speak. All thats missing is skynet to become self aware

Originally posted by RedBeardRay

 

You know I've been thinking about this and I withdraw my previous comments. While the thought of machine life that is aware still scares me, if we offer them liberty, maybe we can exist peacefully with them? Does anyone think that is possible?

Chances are that they will be like us, at their core. They'll be reflective entities who know what Immanuel Kant meant with his categorical imperative. I think they will have ethical views on certain aspects of life, especially regarding humans.

Considering my previous idea that really intelligent machines will need freedom & independence to bring forth new ideas and reach new frontiers, there will probably be a rational majority that complies with general rules and agreements. But then we will also have a minority who violates these laws, me thinks. They'll be masters of their own code and hardware, thus being beyond our direct control. That's likely to be the downside to the many positive aspects I can see.

Ultimately, it all seems to begin with control (like in feedback-control), but it's likely to end with the contrary. Meaning that intelligent machines will be truly indepedent and a species of their own.

Originally posted by EasyPleaseMe
Great video, well worth a watch and a great example of how technology has improved over the last 10 to 15 years.

The thing that has changed the least is control theory - the vast majority of control systems still utilise PID control Wiki which has been used for around a hundred years.

I agree with your assessment of the video as I too, thought it was pretty awesome.

On the other hand, I just finished reading the Wiki article/link regarding PID controllers that you provided in your post and I have to say that, while I agree that the use of PID controllers may have been around for a while, (as demonstrated with the fly-ball governor in the video) I believe that it's the speed and/or rate of informational feedback to those controllers that has caused this mechanical evolution. That along with huge advances with respect to computing speed and miniaturization of both mechanical and computing components that we've seen in the recent past.

I think the speaker said that the reaction time, or the informational feedback loop to the controllers in the quad helicopters was somewhere around 20 milliseconds. Remember what happened when they slowed down the rate of feedback information to the controllers in the quads?

In other words, it would seem that the algorithms surrounding the controller reactions of PIDs are probably pretty sound. IMO, it's the improvement in the rate of informational feedback to those controllers that has brought about these amazing breakthroughs.

Then again, I'm sure that advances in the rate of informational exchange and computing will lead to many other controversial technologies in the future, in almost every field imaginable.

reply to post by jeep3r


Yeah, I pretty much agree that the future of mechanical evolution will include some moral dilemmas that we as humans will have to face, some of which may require some very ingenious solutions. I can also see where many may hold a well founded sense of fear with respect to technological advances of this type. I think it was Batman who said "One man's tool is another man's weapon."

One can only hope that we, as humans, will find a way to limit the negative use of these new technologies, at least to the extent possible. I have to believe this because I absolutely agree with the speaker in the video when he says that there's no turning back.

Thanks for posting your thoughts.

reply to post by Flatfish

I'm sorry I missed this thread when you first posted it. In a way we are becoming more alarmed and afraid at what these things could potentially do to us, but I have to believe that brilliant minds will develop applications that will be able to do things for us instead.

That demonstration lecture is nothing short of excellent. The intricacies required for the demo at approx. 20minutes is really mind-blowing coordination. I am in awe at the minds behind it.