The Instigator
lyokowarri0r
Con (against)
Tied
0 Points
The Contender
red_x
Pro (for)
Tied
0 Points

Self Aware AI is possible

Do you like this debate?NoYes+0
Add this debate to Google Add this debate to Delicious Add this debate to FaceBook Add this debate to Digg  
Post Voting Period
The voting period for this debate has ended.
after 0 votes the winner is...
It's a Tie!
Voting Style: Open Point System: 7 Point
Started: 5/3/2016 Category: Technology
Updated: 1 year ago Status: Post Voting Period
Viewed: 461 times Debate No: 90618
Debate Rounds (4)
Comments (7)
Votes (0)

 

lyokowarri0r

Con

Rules
1st round acceptance
2nd opening argument/no rebuttals
3rd rebuttals only
4th closing statement

My basic position is that the AI that many foresee that will be just like human, or even better, cannot happen. It is hard to ask for much proof in this matter as it is speculative, so don't be too intimidated.
Please ask any questions in the comment section.
red_x

Pro

Thank you for this opportunity to debate with you on this topic, although not much proof can be given for this specific debate I can't wait for this.
Debate Round No. 1
lyokowarri0r

Con

You are correct that there is not much proof that can be offered, but I still find this can be debated well. My argument will be mostly causal. Many people see AI on a equal scale to man as very much possible and even may be arriving soon. It is hard to claim something cannot happen at all but I am very skeptical of the idea of self aware and recursive improvement. I do not think these people are stupid. However, I discussed this topic a month ago in my debate club, and found that I was the only one perusing a computer science degree. I find that if we look at the current hardware and look within ourselves, we will se that AI to rival us is impossible.

First point: Hardware.
The issue is that NOTHING can rival the computing power of the brain. Not even close. A cpu powerful enough would have to be the size of Ohio(where I live) and would need the same power that the whole countries uses daily to even turn on. It is far from off the horizon. Furthermore, we have to understand what it means to be human. We learn and grow as individuals. Experiments on rats show that stimulation makes our brains larger and more complex. MRI and PET scans also show in humans that people that know two languages have more a more complex temporal lobe(deals with words). Our brains physically grow and more connections are made. Hard drives have a finite amount of space and does not grow with information. We could just hook up more hard drives, but the AI could never grow like we do. This same limit applies to everything computer. The cpu can only go so fast and does not get better with practice.

Second point: software.
My biggest issue is that I find that most people do not know how computer think and are programmed. Things are done one line at a time. However, our brain is a filled with connecting cells that work together to reconstructed memories or think thorough a problem. The problem is learning. We learn as people and make connections from that and things we do not understand. We have AI that learns. There is a robot that walks, and if it missteps, it corrects itself and will learn not to fall again. We have made computers that learn how to be expert at games through losing and learning from that. Here is the problem...we have to program HOW it learns. We have to put in that falling is not what is wanted and that it should avoid it. We have to tell it that losing is bad. We all learn this from experience but this is from society. Machines do not conform to society. Connections are also not inherent. If I tell you to connect water and "U.S.S", one could connect USS with ships, and ships float on water. A computer could be programmed to make that connection but we would have to make a connection for everything which is not possible.

Third point being human is more complex than we think.
Lastly, I would like to ask what does it mean to be human? We have ambition, love, passion, and reasoning. I would argue that there is no true logic and everything is about perspective. Ambition and the idea of working hard are ideals of man. Love and passion for things are inherently illogical. I love to paint, but why do I do it? Because I like it. This AI could have dislikes and likes but how would it obtain it? Would we write them in? Along with a whole personality? This is not a self thinking or independent being, it is a imitation of what we can be. Lastly...purpose. What is it? Something that is different for everyone. We have many of them. Some short term, some long. How will this machine find purpose? It will be a method with nothing in it. It will be a shell of humanity. Purpose is such a human thing, something that cannot be programmed.

That was a bit long winded. I look forward to the opening statements of the pro side.
red_x

Pro

Self Aware AI can occur, its brains would just need to be stored wirelessly such as on the internet. The internet contains all the information in the world(the world we know), emotions could be programmed but it would be insanely difficult, so instead of emotion you program it to react the normal way a human would, killing would be a problem but with any self-aware robot.
Debate Round No. 2
lyokowarri0r

Con

I was hoping for a bit more to work with, but I get what I get. I would like to point out that the first sentence of the pro side shows how ignorance in the subject leads to misunderstanding. A computer needs a cpu for processing information; it needs a hard drive to store information; it will also need a motherboard for the base firmware. We cannot just "have it on the internet." A processor is a physical thing. You might say that we upload so many things on the internet so I AI would be possible but the internet is just servers around the world that are connected; a physical computer is still needed. We might say that the AI would learn from the internet, but where would it start? Again, we have seen that connection are complex to program and impossible to make them all. If it wants to look up how to row a boat, it might just search it on Google, but what would it do with the info? Why would it want to know? We need to see that a self-aware AI has no reason to live...no purpose. Emotion is beyond even our comprehension, let alone being able to be programmed. Also the way people react is from emotion and a lens of cognitive input. Would this AI see and hear? If it did, how would it learn how to use these inputs? It takes years for us to learn how to use our bodies.

Why would it kill people? Why is that logical? Computers are pure logic, only using numbers to finish tasks. There is no reason why a computer would want to kill humanity. Is it scared? We had to program fear. Is it evil? We had to program that. We can try to program it all, then the result is not self-aware AI, it is just a program that spits output from a certain input without processing anything.
red_x

Pro

The problem I see is that I didn't explain myself well enough. What I meant was that yes, i know how a computer works, but while those are needed, all a self aware robot would need to be able to access the internet for information, main processing would be in it's body, movement and that would be pre-programmed in, but info such as knowledge would be found on the internet.

While nothing CAN rival the power of the human brain, that is what the internet would be for. We have the tech just not the ability. Besides just killing, viruses can occur is what I meant, i meant not whether we could but whether we should.

The whole, it takes years for us to learn to use our bodies thing I don't quite like because we aren't just learning during those years we are also, they have made working robots smaller than me, i am 5' tall. So if they made it taller and a little wider everything could fit. it just needs enough processing power to get the job done.

http://www.collective-evolution.com...

while this is kind of funny it shows how close we are to self aware AI
Debate Round No. 3
lyokowarri0r

Con

I would like to point out that there is a deference between information and knowledge. Knowledge applies information to things other than whet the original info defines. People can see that wood floats on water, and that if something else floats too, then it must be a similar bouncy. That connection is much more complex than many realize. We also need to see that cramming information into a machine does not mean it knows how to use it. Also, the Internet is a vast place, and even Google does not know how much there is available to see using its engine. A computer would not be able to go through ever single webpage and take all this data, and then apply it to something else. Computers are getting better and mire like humans but we still need to see that what they can do is still limited. The AlphaGo bot beat the best human we could offer at Go. But put a checkers board, a much simpler game, the machine would stumble. A computer can be given a task and do it well. It can become better than we intended, but it can only go as far as the base programming will allow. Hard ware and soft ware are the biggest limitation. Computers are not human. They can never be human. We think too differently. It might be good for a book or movie, but in reality, AI is just a fantasy.
red_x

Pro

While my opponent has put up some good points, it is true a self aware AI is possible . But the stipulation applies that as soon as it is self aware it is no longer an AI, as it would learn instead of be programmed.
Debate Round No. 4
7 comments have been posted on this debate. Showing 1 through 7 records.
Posted by red_x 1 year ago
red_x
Yeah i got that part, this will be interesting no matter what. See you on the battlefield.
Posted by lyokowarri0r 1 year ago
lyokowarri0r
No no. I just don't see the self aware part happening.
Posted by red_x 1 year ago
red_x
Ok, so basically once it becomes self aware it is no longer an AI...
Posted by lyokowarri0r 1 year ago
lyokowarri0r
I am of the position that the AI that I have described cannot happen: period. I see many people's definition of AI is not possible, and when I show them so, they change it to the point where it is no longer AI.
Posted by red_x 1 year ago
red_x
yeah imma challenge you
Posted by red_x 1 year ago
red_x
I would be interested in debating this with you, although I fail to see how this can be debated. It eventually could happen but we are to far away from this technology as of now.
Posted by lyokowarri0r 1 year ago
lyokowarri0r
I should clarify that by Self Aware AI. I mean it knows it exist, and that we have reached the technological singularity which allows for recursive self improvement.
No votes have been placed for this debate.