What Consciousness does a Robot Need?

In some respects it is easy to provide computer programs with more powerful introspective abilities than humans have. A computer program can inspect itself, and many programs do this in a rather trivial way. Namely, they compute check sums in order to verify that they have been read into computer memory without modification.

It is easy to make available for inspection by the program the manuals for the programming language used, the manual for the computer itself and a copy of the compiler. A computer program can use this information to simulate what it would do if provided with given inputs. It can answer a question like: ``Would I print ``YES'' in less than 1,000,000 steps for a certain input? A finite version of Turing's argument that the halting problem is unsolvable tells us that that a computer cannot in general answer questions about what it would do in steps in less than steps. If it could, we (or a computer program) could construct a program that would answer a question about what it would do in steps and then do the opposite.

Unfortunately, these easy forms of introspection are not especially useful for intelligent behavior in many common sense information situations.

We humans have rather weak memories of the events in our lives, especially of intellectual events. The ability to remember its entire intellectual history is possible for a computer program and can be used by the program in modifying its beliefs on the basis of new inferences or observations. This may prove very powerful.

To do the tasks we will give them, a robot will need at least the following forms of self-consciousness, i.e. ability to observe its own mental state. When we say that something is observable, we mean that a suitable action by the robot causes a sentence and possibly other data structures giving the result of the observation to appear in the robot's consciousness.

We will give tentative formulas for some of the results of observations. In this we take advantage of the ideas of [McCarthy, 1993] and give a context for each formula. This makes the formulas shorter. What , and mean is determined in an outer context.

• Observing its physical body, recognizing the positions of its effectors, noticing the relation of its body to the environment and noticing the values of important internal variables, e.g. the state of its power supply and of its communication channels.

  

  [No reason why the robot shouldn't have three hands.]

• Observing that it does or doesn't know the value of a certain term, e.g. observing whether it knows the telephone number of a certain person. Observing that it does know the number or that it can get it by some procedure is likely to be straightforward.

  

  

  Deciding that it doesn't know and cannot infer the value of a telephone number is what should motivate the robot to look in the phone book or ask someone.

• Keeping a journal of physical and intellectual events so it can refer to its past beliefs, observations and actions.

• Observing its goal structure and forming sentences about it. Notice that merely having a stack of subgoals doesn't achieve this unless the stack is observable and not merely obeyable.

• The robot may intend to perform a certain action. It may later infer that certain possibilities are irrelevant in view of its intentions. This requires the ability to observe intentions.

• Observing how it arrived at its current beliefs. Most of the important beliefs of the system will have been obtained by nonmonotonic reasoning, and therefore are usually uncertain. It will need to maintain a critical view of these beliefs, i.e. believe meta-sentences about them that will aid in revising them when new information warrants doing so. It will presumably be useful to maintain a pedigree for each belief of the system so that it can be revised if its logical ancestors are revised. Reason maintenance systems maintain the pedigrees but not in the form of sentences that can be used in reasoning. Neither do they have introspective subroutines that can observe the pedigrees and generate sentences about them.

• Not only pedigrees of beliefs but other auxiliary information should either be represented as sentences or be observable in such a way as to give rise to sentences. Thus a system should be able to answer the questions: ``Why do I believe ?'' or alternatively ``Why don't I believe ?''.

• Regarding its entire mental state up to the present as an object, i.e. a context. (McCarthy 1993) discusses contexts as formal objects. The ability to transcend one's present context and think about it as an object is an important form of introspection, especially when we compare human and machine intelligence as Roger Penrose (1994) and other philosophical AI critics do.

• Knowing what goals it can currently achieve and what its choices are for action. We claim that the ability to understand one's own choices constitutes free will. The subject is discussed in detail in [McCarthy and Hayes, 1969].

The above are only some of the needed forms of self-consciousness. Research is needed to determine their properties and to find additional useful forms of self-consciousness.

• Understanding and Awareness