daveb posted an update 4 weeks ago
Should we fear or welcome robots?
Danger Will Robinson, danger!
That’s the robot I think of when somebody talks about robots. But with the advances in robotics, AI and the the developments in autonomy, should we welcome robots or fear what they can be used for. Because at some point a developer will attach a weapon to them straight off the production line.
We should welcome them my friend, welcome them!
Automation has been helping mankind for years, and will continue to do so – us humans will adapt to embrace and harness what the technology advances bring.
So firstly I think you are conflating AI and robots because they aren’t the same thing.
Secondly, both these things are tools and neither good nor bad in their own right. This is the same as a knife or a gun is a tool, or gene editing is a tool. The thing itself is nothing to be feared, it is the people who choose how it will be used who are to be feared.
True but a genuine AI might be able to choose how to use itself, on that basis could be feared/celebrated in its own right?
personally I can’t see how an AI could mess up the world any more than we have.
Possibly, but that’s a way off yet I think. If we get that far, we will debating whether AIs are alive or not. Even the most sophisticated AIs have a specific goal, which is usually something innocuous and finite like “get from A to B in the least amount of time” and a set of rules like “don’t destroy obstacles in your path”. It is when we setting goals like “improve yourself” or “exist for as long as you can at all costs” and taking the rules away that we will run into problems.
I have no problems with robots or AI.. its the people in control of them that you have to be worried about. With a healthy economy with lots of competition etc everything will be fine and dandy but if everything continues to center around the silicon valley elite then you got problems. Already we seen how ruthless these tech company’s tax avoiding, poor treatment of their workers etc and AI is just getting started.
I was listening to an Infinite Monkey Cage episode on AI (I know it’s not robots but bear with me …)
In it, one of the experts talked about a situation where robots suffered : https://www.youtube.com/watch?v=8Ip3q4d03Eot=32m4s
I could completely envisage this as well – if you think about why we feel physical pain.
We feel pain when we do something in order to ensure that we do not overtax or damage our bodies. Part of getting fit is about pushing that envelope and constructively damaging ourselves so that we grow back stronger.
So let’s say you build a robot and you program in a pain feedback mechanism so that the robot can learn when it should and shouldn’t move in particular ways in order to protect itself.
What if when you switch that robot on you’ve got the pain threshold parameters all wrong, and the thing is just sitting there in agony! Presumably you’ll have developed a safe position for the thing to recoil to … just think how bad you’d feel if it immediately curled up into it’s equivalent of the foetal position!
At this point, presumably, the humane(!?) thing you can do is to put it out of it’s misery and switch it off!
Just an interesting point of view I thought.
We are still a loooong way off any possible Skynet scenario. But the development of robotics and AI is quite fascinating there are still major developments needed for anything the sci-fi android set up.
1. the abilitiy convert a variety of energy sources into useful energy. For us this is eating, breathing, drinking. Any sucessful autonomous machine would need to do similar. Solar power is an option but is still too inefficient for anything other than slow movement and operation (ask a plant)
2. the ability to self-repair. I don’t mean going down to the shops and buying up some nuts and bolts and a wealding torch but actual healing. We are very good at doing this compared to robots. Possible lines of enquiry might be nano tech or bio-tech.
3. the ability to absorb, order and interpret complex data and act upon it accordingly. A definite possibility in the future though yet to match the human brain.
4. the ability to reproduce (if it isn’t going to remain a one off dependent upon other species for reproduction)
5. the ability to protect itself. Especially in a way that conforms to relevant standards of morality (unless murder or maiming is desired). Current robots are very clunky and slow though drones are probably the flexible soldier of choice
> We are still a loooong way off any possible Skynet scenario.
except for the real Skynet of course, first heard about it back in 73/74
I hope it all turns out like Sirius Cybernetics “Share and Enjoy”