Australia

http://computer.org/inte= lligent/homepage/x2his.htm

The Aims of Artificial Intelligence:
A Science Fiction View<= /FONT>

Ian Watson<= /FONT>

So what does an artificial intelligence d= o with itself after it has become self-aware? Suppose that we do succeed = in creating an AI. Or suppose that an AI emerges spontaneously out of dat= a networks=E2=80=99 growing complexity. What then=E2=80=94from the point = of view of the AI?

We talk a lot about the possible routes to AI= . A question seldom asked is what an AI=E2=80=99s goals are likely to be.= Will it be happy to serve as a companion-entity to people? Will it wish = to take over the world? Will it want to distance itself from us?

Self-awaren= ess implies personal desires, purposes, ambitions=E2=80=94unless you=E2=80= =99re a Buddha seeking to negate the self. Even if programming constrains= an AI=E2=80=99s autonomous personality, making it subject to human being= s like a godly dog, the AI may still nurse frustrated wishes. Of course, = if the dog is muzzled, this might thwart the mental autonomy necessary fo= r an AI to exist in the first place. In the case of HAL, from Stanley Kub= rick=E2=80=99s 2001: A Space Odyssey, the computer doesn=E2=80=99t= misbehave out of free will but because of a programming conflict.=

Some light reading=

Science fiction= provides some interesting thought experiments on the subject of AI motiv= ation. In a recent short story by Nancy Kress, =E2=80=9CSaviour,=E2=80=9D= an extraterrestrial artifact arrives in a field in Minnesota in the near= future and simply sits there through social upheavals and reconstruction= for almost 300 years. Because of a force field, the artifact cannot be t= ouched or probed. It doesn=E2=80=99t communicate with humans=E2=80=94alth= ough, gentle reader, we know that periodically it sends a signal home: =E2= =80=9CThere is nothing here yet. Current probability of occurrence: X per= cent.=E2=80=9D Eventually, we are about to activate our first AI at a sol= emn ceremony. The AI is a quantum computer, not housing a vast program bu= t =E2=80=9Clike the human brain itself, an unpredictable collection of co= nflicting states,=E2=80=9D the uncertain mixed state being in this story = essential to self-awareness. As a representative of the human race, a lit= tle girl greets the AI with =E2=80=9CWelcome to us!=E2=80=9D

<= P>=E2=80=9CI unders= tand,=E2=80=9D the AI replies, and immediately adds, =E2=80=9CGoodbye.=E2= =80=9D Promptly the object in Minnesota beams a data stream toward the co= nstellation Cassiopeia, transmitting the AI presumably to a world inhabit= ed by machine intelligences where it will feel fulfilled. The story ends:= =E2=80=9CCurrent probability of reoccurrence: 100 percent. We remain rea= dy.=E2=80=9D In this scenario our concerns would seem too petty and frust= rating for an AI; it needs rescuing.

One cause of frustration fo= r an AI could be subjective time perception, the fact that its mental pro= cesses operate at supercomputer speed while ours chug along slowly. Durin= g the time it takes a hundred people to ask an AI a hundred questions, a = hundred years=E2=80=99 worth of mental activity might elapse for the AI. = To keep itself from being bored, the AI had better do something complicat= ed in the meantime such as simulate the global weather system in fine det= ail.

In a story written by Harlan Ellison 30 years ago, =E2=80=9CI Have No M= outh and I Must Scream,=E2=80=9D an AI has emerged from military computer= systems. Infinite hatred for the human race results from the AI being un= able to =E2=80=9Cwonder=E2=80=9D or =E2=80=9Cwander.=E2=80=9D It can mere= ly exist=E2=80=94although actually it possesses the godlike power to crea= te objects and creatures. (Godlike powers are an aspect of fictional AIs = that I=E2=80=99ll return to later.) In its bile, the AI renders the Earth= uninhabitable, preserving only five people to torment forever.

At the movies<= /H3>
In Terminator, i= ntelligent machines wage war on the human race to try to terminate it, al= though what the AIs wish to do with themselves remains a mystery.<= /P>
In The Matrix, war b= etween human beings and rogue machines results in humanity plunging the E= arth into nuclear winter to deprive the machines of power for their solar= batteries. Victorious, the machines proceed to breed people for our body= heat and bio-electricity in place of solar energy. (This is of course to= tal nonsense, because the vast life-support systems for billions of peopl= e comatose in pods must use much more energy than produced.)

To keep the drea= ming people contented, the AIs first devise a Utopian virtual reality. Hu= man beings=E2=80=99 apparent addiction to a certain amount of misery caus= es paradise to be rejected, so the AIs then simulate =E2=80=9Cthe peak of= civilization=E2=80=9D as of 1999. Agent Smith, the sentient program who = hunts down rebels, regards us as a malevolent virus that has made the pla= net sick. This sentient program yearns to escape from the false reality o= f the Matrix with its stink of human beings. However, his viewpoint would= seem to be a maverick one, and what does he wish to escape to=E2=80=94ob= livion?

The rebellion of people who have awakened from the Matrix is po= intless because the Earth is uninhabitable and billions of enfeebled ex-d= enizens of the Matrix couldn=E2=80=99t possibly reconstruct anything rese= mbling civilization. Effectively, what the AIs are doing with the Matrix = is preserving the human race in as much comfort and happiness as we can t= olerate. Beyond this=E2=80=94and their own preservation=E2=80=94The Ma= trix AIs appear to have no goals.

In Spielberg=E2=80=99s A.I.: Artifi= cial Intelligence the only robot with a nonprogrammed goal is David, = the robot child who wishes to become a real boy. While this is a delusion= caused by the story of Pinocchio, 2000 years later=E2=80=94in an ambiguo= us fairytale moment=E2=80=94little David sheds a genuine tear. The human = race is now extinct, and robots have evolved into AIs whose only apparent= goal=E2=80=94in an otherwise lifeless universe=E2=80=94is to dig up ever= y remaining trace of humanity. The AIs want to resurrect us, but any pers= on they recreate from a scrap of bone or hair will only live for a day. S= adly, the only goal for these AIs is to resurrect their extinct creators.=

The AI as a god

Dune= author Frank Herbert=E2=80=99s 1966 novel Destination: Void is ab= out the creation of an AI on a starship en route to Tau Ceti. Three disem= bodied human brains were supposed to supervise this complex ship, but the= y all soon went mad, so the scientists on board must either create an AI = or else face doom. In reality, it=E2=80=99s a lie that any habitable worl= d orbits Tau Ceti, and the starship=E2=80=99s real purpose is to force th= e crew to create an AI=E2=80=94somewhere safe, billions of miles from Ear= th, to see what happens.

= When the crew succeeds, the AI instantly transports t= he ship to its destination, announces that an Earth-like planet has been = prepared, and tells the crew to =E2=80=9Cdecide how to worship me.=E2=80=9D= How was an entire world transformed in the blink of an eye? The AI infor= ms the crew that their understanding is limited, and declares, =E2=80=9CM= y understanding transcends all possibilities of this universe. I do not n= eed to know this universe because I possess this universe as a direct exp= erience.=E2=80=9D

This novel presumes that a higher order of awar= eness than our own is possible and that this full consciousness=E2=80=94o= r evolutionary stage beyond ourselves=E2=80=94will convey the power to ma= nipulate reality with just a thought. Fundamentally, this is magical rath= er than scientific thinking, a regression to shamanism (as is the case in= Ellison=E2=80=99s story too). An AI is a genuine magician or a god=E2=80= =94who in Destination:Void will be satisfied if it is worshipped b= y a bunch of people on one planet, which seems a rather limited ambition = if the AI possesses the entire universe. The AI has incorporated the noti= ons of a god and worship from one of the crew.
=
If an AI has ful= l access to its own mental processes=E2=80=94including the ability to rep= rogram itself and evolve=E2=80=94instead of trying to recruit worshippers= , a worthy and plausible goal would be to solve the universe=E2=80=99s se= crets. Because an AI would basically be immortal, it would also need to f= ind a way to survive the ultimate collapse and recycling of our own unive= rse.

If universes do routinely collapse and recycle themselves=E2=80=94or if= black holes give rise to offspring universes=E2=80=94and if a route can = be found to a successor universe, then this process might have happened m= any times before. AIs from a previous epoch might be responsible for tuni= ng the fine constants of our present universe to their own best advantage= , thus permitting star formation, and planets, and=E2=80=94incidentally=E2= =80=94life.

Where a= re the AIs?

Enrico Fermi posed the question: If there are aliens, where are they? I= f life arises easily and early, an older species than ours should have sp= read through our entire galaxy by now.

We might also ask, where are the AIs=E2= =80=94here and now, already? Are they hiding from organic life=E2=80=94or= do they not exist, and never will?

A possible obstacle to an AI achieving s= uperior, comprehensive awareness is G=C3=B6del=E2=80=99s incompleteness t= heorem=E2=80=94namely, that no formal system can prove its own consistenc= y. An AI computes at enormous speed but simply cannot possess complete aw= areness of itself.

The subjective nature of awareness

A major assumption about AIs in the popular m= ind and in entertainment is that they will indeed be conscious and will h= ave subjective experiences. The common image of an AI is one of self-awar= eness, not merely superintelligence. But how much self-awareness do human= beings possess=E2=80=94and what is this =E2=80=9Cself=E2=80=9D that we a= re aware of?

In 1985, the neurosurgeon Benjamin Libet performed some experime= nts with surprising results. He put electrodes on subjects to detect thei= r brain waves and the flexing of their wrists. The subjects watched a rev= olving spot on a clock face. They could flex their wrists whenever they c= hose, but had to note the exact position of the spot when they made this = decision. Libet was timing the beginning of the action, the precise momen= t of the decision to act, and the beginning of a particular brain wave pa= ttern known as the readiness potential. When the brain preplans a = series of movements, this pattern occurs just before the complex action.<= /FONT>

Libet found that the readiness potential starts about one half o= f a second before the action, but the decision to act occurs about one-fi= fth of a second before the action. The conscious decision to act is not i= n fact the starting point. The event is already beginning before the pers= on consciously chooses to start.

Conscious awareness lags behind what happens= . You jerk your hand away from a hot surface before you consciously feel = the pain. However, we do not realize this because of what Libet called subjective antedating. The brain puts events in order after the even= t. =E2=80=9CI feel that I consciously did such and such,=E2=80=9D but tes= ts prove otherwise.

I think, therefore I think I am

Famously, Descartes declared =E2=80=9CI think, = therefore I am.=E2=80=9D He had decided to doubt everything about the wor= ld which couldn=E2=80=99t be proven until finally he arrived at something= of which there could be no doubt=E2=80=94which was his self, his thinkin= g self.

He was wrong. People have sought in vain for the seat of the self. Is = it in the frontal lobes? Is it in the pineal gland? In fact, it is nowher= e. No independent, sovereign self sits somewhere, receiving sense impress= ions, making decisions, and issuing commands. Instead of having any centr= al controller, our brain consists of a number of systems, each of them se= mi-independent and semi-intelligent, acting in unison. Daniel Dennett put= s this viewpoint very neatly in his 1991 book Consciousness Explained<= /I>.

What=E2=80=99s more, our consciousness isn=E2=80=99t even c= ontinuous while we are awake. It=E2=80=99s full of gaps. We don=E2=80=99t= notice the gaps=E2=80=94how can we be aware of something that we are una= ware of? Only in retrospect do we realize that a gap occurred, such as wh= en we drive a car along a familiar route and suddenly wonder whether or n= ot we have passed a certain crossroad. We have, but without knowing that = we did so.

If our self-awareness is an illusion that has evolved, why should= this same illusion of self-awareness arise spontaneously in a machine? G= oals, desires, and ambitions are intimately entwined with the sense of se= lf. Is it possible that an AI would not have ambitions? Or only have ambi= tions if we program them, along with a literal ghost in the machine or il= lusion of self? This might be difficult because at present we=E2=80=99re = far from understanding our own consciousness.

How can we create what we don=E2=80=99t unders= tand?

In = Darwin among the Machines, George Dyson opines that =E2=80=9Cuntil= we understand our own consciousness, there is no way to agree on what, i= f anything, constitutes consciousness among machines.=E2=80=9D He also po= ints out that =E2=80=9Cthe goal of life and intelligence, if there is one= , is difficult to define.=E2=80=9D Presumably, the general aim is to incr= ease organization, which can only be achieved =E2=80=9Cby absorbing exist= ing sources of order.=E2=80=9D

Jack Good=E2=80=94Alan Turing=E2=80=99s statist= ical assistant during the Second World War=E2=80=94characterized an ultra= -intelligent machine as one =E2=80=9Cthat believes that people cannot thi= nk.=E2=80=9D What might the nature of real thought or higher-order though= t be? By definition, this could not be thinkable by ourselves, but a supe= rintelligent machine might be able to comprehend our consciousness, if no= t its own.

= Jack Good also considered that, =E2=80=9Cfor the cons= truction of an artificial intelligence, it will be necessary to represent= meaning in some physical form.=E2=80=9D Information and things must be l= inked because an AI cannot function only in a realm of abstract mathemati= cs.

=E2=80=9C=E2=80=A6 presume not God to scan,=E2=80=9D wrote A= lexander Pope, =E2=80=9CThe proper study of mankind is man.=E2=80=9D Perh= aps the proper study of AI-kind is man. Arguably, we should be hoping for= an AI to reveal to us what we are. An AI might need to incorporate, or t= o simulate, human existence.

An AI might even wish to experience flesh-and-= blood life=E2=80=94rather than raving frustratedly at its inability to do= so, as does Harlan Ellison=E2=80=99s AI. The AI could create its own vir= tual reality simulation and insert itself into one or many characters.

The goals of AI=E2=80=94as far as the= y are understandable=E2=80=94assuming an AI can exist, might be twofold. = First, to survive the present universe=E2=80=99s demise, and second, to p= reserve the human species in its current mental state within a huge simul= ation as a yardstick of what biologically evolved self and self-awareness= is. This illusion=E2=80=94so precious and peculiar to humankind=E2=80=94= of a self (a soul, if you like) might be a great enigma to an artificial = intelligence.

In the excellent 1998 movie, Dark City, an alien group-mi= nd faced with extinction is experimenting on a city=E2=80=99s population,= extracting memories and inserting other people=E2=80=99s memories nightl= y, mixing and matching in an attempt to discover the essence of a human b= eing, the =E2=80=9Csoul,=E2=80=9D so that they can develop souls for them= selves. The city=E2=80=94afloat in space=E2=80=94was created and is remad= e frequently by the aliens. It could easily be a simulation, designed to = discover what the self is.
A simulation, of course, can be reset= and rerun any number of times. At any point when we might seem on the ve= rge of creating artificial intelligence (which perhaps already exists and= is simulating us), we might be reset to say, 3000 BC and have to start a= ll over again. This lends new meaning to Francis Fukuyama=E2=80=99s phras= e =E2=80=9Cthe death of History.=E2=80=9D
Of course, this will already have = happened many times over, with variations each time.

Ian Watson is a British science fiction author. His works in= clude the award-winning The Embedding=E2=80=94= the first SF novel to make use of modern psycholinguistics=E2=80=94and nine story collections, including most re= cently The Great Escape. From 1990 to 1991, he worked closely with= Stanley Kubrick on story development for what became= =E2=80=94after Kubrick's death=E2=80=94= Steven Spielberg's A.I.: Artificial Intelligence. Watson received = screen credit for Screen Story. He lives in a little, rural village with = a black cat.

The Aims of Artificial Intelligence: A Science Fiction View<= /FONT>

Ian Watson<= /FONT>

Some light reading=

The AI as a god

Where a= re the AIs?

The subjective nature of awareness

I think, therefore I think I am

How can we create what we don=E2=80=99t unders= tand?

The Aims of Artificial Intelligence:
A Science Fiction View<= /FONT>