AI - Godsent or Humanity's doom?
Well, the abbreviation AI stands for Artificial intelligence - An entity that has some cognitive qualities of human brain, like ability to learn, solve problems, or maybe even have conscience?
So, the bottom line would be that it’s great and we shouldn’t worry?
Well, under some conditions - yes. When developed in safety and narrowed to specific usage. It’s just like with the laboratories that develop cures for lethal diseases - they’re amazing, and we benefit greatly from their research, but if the virus escapes and cannot be controlled - it’s very, very bad. Exactly the same goes for AI - it’s good as long as it’s under control. But when it outsmarts us, and gets away from under our control - then bad thing can happen. It’s clearly visible on the example of a bot Microsoft recently made - AI imitating teenage girl on twitter - within 24 hours it came to being nazi, sex loving robot. All because it was set free to the internet with no control. So imagine what could happen if there was AI smarter than humans with access to internet? I guess within some time we could come to the vision from „The Matrix” movie - world controlled by machines and humans either extinct, or living in some sort of virtual world only. So, how to not let this happen? All we need to do is to supervise what we make and make sure its goals don’t involve annihilating us or simply not develop AI smart enough to outsmart us - following the idea the mighty of this world have.
 |
| The World controlled by the machines shown in "The Matrix" |
That’s it for this week, thank you for reading.I know there's plenty more to say on this subject, but it's not possible to cover it all. What are your thoughts on the subject? You can share them in the comments, I’ll be more than happy to read them.
https://www.intel.pl/content/www/pl/pl/analytics/artificial-intelligence/ai-in-your-pocket-infographic.html
https://futureoflife.org/background/benefits-risks-of-artificial-intelligence/
http://www.telegraph.co.uk/technology/2016/03/24/microsofts-teen-girl-ai-turns-into-a-hitler-loving-sex-robot-wit/
https://www.youtube.com/watch?v=N4o0OoN0zwQ
This is indeed a huge topic, and a scary one too. There are quite a few scientists who openly say that the moment when AI will outsmart us is imminent. Maybe that's why I am so sceptical about driverless cars.
OdpowiedzUsuńWhat about the definition of AI? I thought the programme which chooses the next YouTube videos for me to watch is not exactly a self-learning AI but just an algorythm taking into account my previous choices.
I can't imagine how AI, human work, could outsmart or take control of people. First of all, it doesn't have (and will never have) feelings and, as a result, will never have desire to take control over us. Secondly, we can't create anything smarter than we are. AI will never be able to plan future or distinguish between real or fake. I find it humiliating for human to think artificial intelligence can outsmart him. Please tell me if I miss something.
OdpowiedzUsuńHave you ever heard of the movie "Blade Runner"? In that movie this question, whether the androids (created by humans, so some sort of advanced AI) can have feelings, is the key. And it's well shown that if one day we'll be able to make something as complex as we are, or an AI with the capabilities of human mind, how we can be so sure that it can't develop the sensitivity to art, ability to feel, or any other quality we define as humanly, just because we haven't discovered yet another creature with these abilities. Also, how can you be so sure that AI can't be smarter than we are? You don't "built" it. You first create some core, and then you let it develop itself. You can either limit its knowledge bank and "thinking" to some particular task or you can allow it to shape on its own. If you give it the ability to learn freely, it'll be gradually discovering the world around it just like a child, and widen its knowledge by getting to know how do particular aspects of the world work, it'll basically do the same we do as we grow, but much faster, and without limits we have because of our biological bodies. So why can't it be something greater than we are? In terms of understanding what consciousness or "soul" is, we are just like socrates in ancient times - we can freely say that the only thing we know is that we know nothing. With that in mind, I think it's impossible for us now to define whether AI can be "concious", or outsmart us. The only thing we have are some suspicions, and assumptions, so I believe you are mistaken saying for sure that AI can't be smarter than us. Also, I don't quite understand what you find so "humiliating" about people having different opinion than yours on this topic. I find it even more surprising knowing that even now there's an AI that was proven to have outsmarted human, the world's champion in GO, japanese game referred to as "the most difficult game in the world", whom I think neither you or I could beat, I believe. You can read it yourself if you don't believe me. goo.gl/8X6VuJ It's been a quite hot topic last year, but maybe somehow you it you might have not heard about it.
UsuńSo you have to have human features in order to have human features (like feelings), that's what you stated. And I tell you none can achieve it. We are at the point where we build supercomputers with capabilities much worse than the insect brain.
UsuńWe are at the moment unable to build a computer or AI comparable with us in terms of full cognition, or feelings. Yes, that's true. But have you read all what I wrote earlier? I never wrote that right now we are building something as complex as we are. What I said was that we have no idea what AI can become within next years, or how powerful will it be.The computing power is exponentially growing each year, and considering the period within which it occured, it's a mere fraction of the time the nature shaped us. And coming back to AI, the development of it started merely a few years ago, so how can you compare such young creation with human brain which has been shaping itself for billions of years and say it can't achieve its level one day? Not even the leading scientists or programmers can predict so, what they openly admit. So please tell me, how can you be so sure about it and say with 100% certainty that it's impossible for it to develop humanly characteristics one day? Are you Stephen Hawking or a medium of some kind?
UsuńI only wrote, that I can't imagine AI taking control of people and explained why. Maybe it can be smarter than human in some way, I admit.
UsuńIf you believe human and his brain shaped itself, I give up-then anything is possible for you. But please explain why then brain shaped itself to process and create feelings? It is very unworthy because it doesn't help to survive. Creatures with developed feelings, like mercy or love, would be eliminated by the ruthless ones.
Answering the question at the end, you really don't need to be genius to know, that things don't come into existence without help from intelligent power. In this case even human can't 'create' feelings. My certainty is based on common knowledge.
From my point of view it is a real possibility of AI outsmarting us in the future and a slight chance of its revolt. Today it's pretty advanced but we can deal with it if something goes wrong. Lot of well known people like Bill Gates for example many times warned about the danger of AI. We can see how fast we got to this point. It took us olny like 30 years to develop learning AI so we don't know what we will create in lets say 100 years. All we can do is to hope we will always be albe to deal with it properly.
OdpowiedzUsuń