Contact Institute for Advanced Science & Engineering for event and ticket information.
Tuesday, June 12, 2007 from 4:15 PM - 5:45 PM (PT)
Conscious knowledge and other information is distinguished from unconscious information by being observable, and its observation results in conscious knowledge about it. We call this introspective knowledge; it's an internal form of experience.
A robot will need to use introspective knowledge in order to operate in the common sense world and accomplish the tasks humans will give it.
Many features of human consciousness will be wanted, some will not, and some abilities not possessed by humans have already been found feasible and useful in limited domains. We give preliminary fragments of a logical language a robot can use to represent information about its own state of mind.
A robot will often have to conclude that it cannot decide a question on the basis of the information in memory and therefore must seek information externally. Paradoxes, e.g. as mentioned by Montague, lie in wait for us here, but Godel's idea of relative consistency lets us formalize non-knowledge and avoid paradox. It turns out that relative consistency, is the basis of many other introspective abilities.
Programs with much introspective consciousness do not yet exist.
Thinking about consciousness with a view to designing it provides a new approach to some of the problems of consciousness studied by philosophers.
By Steven Ericsson-Zenith, Chairman of IASE
Is the manifest existence of experience in the world to be mastered only by poets and priests, or is its mystery one that science can disclose?
Before 1950 the answer was clear, experience lay at the foundation of scientific consideration. But the positivist agenda was abandoned, in part because the implementations of logic in computing machinery proved so successful.
In his 1950 seminal paper, Alan Turing wrote:
I do not wish to give the impression that I think there is no mystery about consciousness. There is, for example, something of a paradox connected with any attempt to localize it. But I do not think these mysteries necessarily need to be solved before we can answer the question with which we are concerned ...
Alan Turing, Computing Machinery and Intelligence. 1950.
But was Turing right that he could ignore the mysteries and still make progress to the ultimate solution? I think not.
Turing's goal was to take rigorous steps forward in reasoning about intelligence and it is a remarkable fact that we are able to capture certain aspects of our intelligence by implementing symbolic logic in computing machinery. But this approach, necessarily, changed the fundamental conception of logical construction from the non-local differentiation from the landscape of the entire embodiment of sense proposed by logical positivism, to the strong locality of integration of logical parts.
Upon these foundations was born a first view of how an artificial intelligence might be conceived. It a great opportunity for us then, and it is with some pleasure, that in this event we hear from the front lines of that revolution.
In this second of our lecture/discussions on the Foundations of Logic and Apprehension, Professor John McCarthy leads the discussion. His name is, perhaps, that most associated with the initial conception of artificial intelligence. He invented LISP, the language most used to that end, and famously coined the term "Artificial Intelligence" at the Dartmouth conference in 1955. He worked with Claude Shannon and many other founders of computer science to build the science as we know it today. He may even have known Turing, we shall have to ask him.
Professor John McCarthy is Professor Emeritus at Stanford University, more details can be found at his Stanford University home page.
Steven Ericsson-Zenith, Chairman.
 |
View other Institute for Advanced Science & Engineering events |
 |
Contact the Host |
 |
Subscribe to receive notifications of future events by this host |
Email
Facebook
Twitter
LinkedIn
MySpace
Digg
del.icio.us
Reddit