(Forgive my awkward English)
Thursday, May 21, 2020
Sunday, May 4, 2014
"What's holding us" in AI?
"What's holding us?" An artical David Deutsch writes about artificial intelligence:
I am convinced that the whole problem of developing AGIs is a matter of philosophy, not computer science or neurophysiology, and that the philosophical progress that is essential to their future integration is also a prerequisite for developing them in the first place.
[David Deutsch's full artical from an external link]
David Deutsch is a physicist at the University of Oxford and a fellow of the Royal Society. His latest book is The Beginning of Infinity.
I would agree to him on the problem of AGI development. There are basic questions haven't been resolved. Here are some:
(1) Natural language, people uses daily, everybody understand it, but what's the nature of natural language, what is 'understading'?
(2) Is AI viable on a computer? In recently years, AI research seems to avoid to answer such question. There is no philosophical/mathematical model for AI, after Turing machine.
(3) The simple but complete cycle of human recognition, has someone done that? Let's take a look the evolution of the Personal Computer (PC) we have today. A PC has its hardware and software component, quite complex. But in beginning, it might be just a small chip that can do simple math like plus/minus. The input is a switch, inputs 'program', one bit each time. The output is a few digits displayed with flash LED.
If a simplified AI application is built, we can expand the function units as per needs. Because of the overall methodology is there, we know what to do, how to optimize. The target will be clear then.
I am convinced that the whole problem of developing AGIs is a matter of philosophy, not computer science or neurophysiology, and that the philosophical progress that is essential to their future integration is also a prerequisite for developing them in the first place.
[David Deutsch's full artical from an external link]
David Deutsch is a physicist at the University of Oxford and a fellow of the Royal Society. His latest book is The Beginning of Infinity.
I would agree to him on the problem of AGI development. There are basic questions haven't been resolved. Here are some:
(1) Natural language, people uses daily, everybody understand it, but what's the nature of natural language, what is 'understading'?
(2) Is AI viable on a computer? In recently years, AI research seems to avoid to answer such question. There is no philosophical/mathematical model for AI, after Turing machine.
(3) The simple but complete cycle of human recognition, has someone done that? Let's take a look the evolution of the Personal Computer (PC) we have today. A PC has its hardware and software component, quite complex. But in beginning, it might be just a small chip that can do simple math like plus/minus. The input is a switch, inputs 'program', one bit each time. The output is a few digits displayed with flash LED.
If a simplified AI application is built, we can expand the function units as per needs. Because of the overall methodology is there, we know what to do, how to optimize. The target will be clear then.
(AI) Technological Singularity might be understandable
Super-powered AI (Technological Singularity) may still follow the same 'recognition scheme' that our mankind is using. If that's true, people should be able to understand it; The super-powered AI may also be friendly to us.
[read more]
[read more]
Subscribe to:
Posts (Atom)