Anyone familiar with the Chinese Room argument against strong AI?
I think that a fundamental flaw with that argument is that it presumes that consciousness is localized, i.e. since the man in the room doesn't understand Chinese there can be no algorithmically based means to achieve awareness. But consciousness is not a localized phenomenon of one area of the brain. It's an emergent property. I would argue that the Chinese Room, as a system, does understand Chinese. And in a world consisting of other Chinese Rooms that interact with it, it does functionally understand Chinese.
So..people understand Chinese though, at least to the extent that people can agree on word definitions. Even then there is struggle in communication as different words have different associations for different people based on their experiences. It has an alphabet and pronunciation. What does consciousness have that we agree on? Do we fully understand it..where it begins and ends? We only know what we observe and to suggest that our observations are complete seems a far reaching presumption...You might be fooled into believing that a robot is a human being, but the only consciousness you can know truly is your own. Even then, there is always more to know
I agree. I know maybe it's not what you meant (soIipsism) but I have always felt a bit drawn to solipsism. Although I try to avoid it as I feel it would lead to dark places.
@towkneed You can either view it as drifting all alone or as a universalist. Maybe we find each other like a baby finds their toes.
@thinkwithme how so?
@towkneed Awareness is shaped by interactions
This AI tests are not designed for AI itself, but the perception of AI.
They work based that the only way that we can sense and explore the world are our senses and perceptions.
So if int he conditions of any test, we cannot distinguish a human from an artificial system, so the artificial system have an AI.
It does not judge what kind or how the system achieve this intelligence, it treat it as a black box, if the box can process the inputs and give outputs that mimic a human, then it is an intelligence.
And when you open the black box, it there is a human inside, it is a human intelligence, if it is an artificial intelligence is an AI. But until the box is opened, you can only say that it is an inteligence or not.
That's why those tests are strong, they are based on fundamental perception and interaction traits that we evolve to recognize. And refuses to give a definition of intelligence.
At the same time it is statistical. Maybe in the future, when humans start interacting more with AIs we will develop more refined perception for this.
I guess you are referring to the Turing Test
[en.wikipedia.org]
Posted by JettyPerspective
Posted by PontifexMarximusWhy Evolution Is True … I never realised that there was still so much opposition to science. [livescience.com]
Posted by NR92What is the reason to live? What are we living for?
Posted by NR92Is it correct that Nietzsche was Hitler's inspiration?
Posted by mzeeWhat is fear?
Posted by DonaldHRobertsThe Most Complicated question ever asked. WHY?
Posted by TheMiddleWayRussel, the greatest salesman the world has ever known!