Moral Dilemmas of Artificial Intelligence

Don Brandes

Computers are on a development pace that will allow the creation of artificial intelligence to match the human brain.  Within the next twenty years, it will be possible to construct machines that communicate with us and give all the appearances of having equal or greater ability to think than a human being. 

Other technologies also may evolve to level that would allow such machines to be carried in convincingly human appearing bodies.  When dealing with such machines, we are likely to think of them as sentient creatures, if not as actual human beings. 

Would they really be sentient? Even if they appear to express emotions, how could we know if they really were experiencing happiness, sadness, or excitement?

What I am telling you is not fiction.  It is the stuff on which past fiction has been based but it is fast becoming reality and it could have major impact on our social system.

Computing devices are becoming pervasive in our society.  We have computers in our cars, our home heating and air-conditioning systems, games, even in the tiny confines of our watches.  The same technology that supports those developments allows us to develop devices with greater and greater ability to simulate the thought processes of the human brain.

It is estimated that by 2020 a $1,000 dollar computer will have the processing power to match the human brain.  By 2030 the average personal computer will have the processing power of a thousand human brains. 

That processing power will allow us to develop thinking machines with apparent consciousness indistinguishable from that of a biological human being.  They may even claim to be human if you ask them.  If you put that artificial thinking device into a human appearing body, people around it would treat it as human because that is what their senses would perceive. 

It probably even will be possible to make a humanoid robot that could be powered by the same foods you and I eat, taking the food in trough its mouth and converting it to energy by a chemical process in an interior cavity.  It would sit down to meals with us and join our conversation.

Or, we could at least design a system that would run on alcohol.  This could go over great in bars and at parties.  Imagine the bets you could win if you had a buddy who could chug a whole fifth of 100 proof vodka and not show any effects from it.

When such entities come into being, it will be necessary for society to make moral and legal decisions concerning how they should be treated.  Will they be granted civil rights or treated like machines that can be turned off by their owners?

Obviously, when you stop and think about it, you know this device is not biological.  But that would become a secondary matter, as it did in the Star Trek Next Generation episode in which a hearing heard to determine if Mr. Data should have rights.  It was obvious that Mr. Data was a machine.  The question of whether he should have rights was based on whether or not he was sentient.  Did he have consciousness?  The producers of Star Trek decided that he was and that he should have the rights of a sentient being.  But what would the United States Supreme Court decide.  

Even assuming a device as advanced as Mr. Data is actually a sentient being, we would still have to deal with an infinite gradation from obviously inanimate machines to sophisticated humanoid robots, and with equally intelligent devices that are not housed in humanoid bodies.  If we are to consider any artificial device as a sentient being, then we must have criteria for drawing the line between those which are and those which are not.  You can imagine the debates about that!       A thinking machine of a particular design may have human rights in one state or country and not another.

One convenient way to program such a machine would be to copy the brain patterns of a real human being.  If you do that, the machine would claim it actually was that person.  It would have all the memories of that person, including having lived a normal human life until one day when he put his head into a brain scanning machine and became transferred to an artificial body or trapped in a desk top box.

This situation will be complicated by the use of computers by natural human beings to enhance there own functions, thus creating a convergence of human and machine.  Geeks at MIT have been walking around for several years with computers strapped to their bodies and interface gadgets in their hats and eye glasses to try to make themselves into sort-of cyborgs, part human and part machine.  Sort of low budget $6 million men!

The creation of real cyborgs might begin with the use of implanted computer chips as prostheses for brain damage.  We already have seen the use of artificial eyes connected directly to the brain.  Artificial devices also can detect odors.  We could be looking at scenarios in which someone gets his head bashed in a car accident and gets a small data processor implanted to take over the lost functions.  Or, when Grandma’s memory starts to go, maybe you’ll get her some new RAM chips. Then you have the question of whether the guy with the new processor still Joe and whether Grandma is still Grandma.  

If you have a person’s complete memory and thought process backed-up somewhere, you could even have the option of totally replacing a lost loved one with a very convincing machine.  At least some people would focus their emotions on the artificial being and treat just as they had treated the lost loved one.  Maybe we could even make arrangements to achieve a sort of personal immortality by getting ourselves converted into machines.

Should we not allow such thing?  You can rest assured that some voices will say we should, especially those who will make money from it.

Here is another wrinkle.  Once machines reach a human level of intelligence, rapid advances in technology will move them quickly past their creators.  In order to stay ahead or even keep abreast of the thinking machines, humans will have to integrate artificial circuitry into their own brains, even if they have not suffered brain damage.   You probably would still say that such an enhanced person was still human.  But what if they decided that the advantages were so great that they had the all thinking parts of their natural human brains completely replaced with faster and more efficient artificial circuitry while maintaining the original body and personality of the individual.  Would it still be the same person?  Would the person still be human?

Those of you who believe in the concept of soul would have even more questions to ask.  Would he still have a soul?  At what point would the soul be lost on a continuum of progressively more of the original human parts being removed and replaced.  Could a completely artificial being have a soul? 

Now let’s go a step further.  Suppose we created a race of self reproducing artificial beings.  Would they be sentient or have souls?  Should they have the same rights as natural humans?

It will be possible to do this.  Computers in use now can design new computer chips as well as live engineers can.  Experiments have successfully created devices that are capable of changing their own programming to adapt to changing contions.  Other experiments also have successfully created machines which invent other simple machines without outside control.  We are in the process of creating a new evolutionary track.

This is not without danger.  From the psychosis of HAL, the computer in Arthur C. Clarke’s 2001 Space Odyssey to Arnold Schwartznegger’s Terminator, we have been warned many times of the potential for our creations to turn on us.

Isaac Asimov created the Three Laws of Robotics to protect us from such events.  Asimov’s three laws are:

1. A robot may not injure a human being, or, through inaction, allow a human being to come to harm.

2. A robot must obey the orders given it by human beings except where such orders would conflict with the First Law.

3. A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.

Presumably, these laws are supposed to be fixed somehow into every robot.  However, Asimov wrote about these things long before access to computers became nearly universal in our society and the specter of computer viruses raised its ugly head.  We know now that even if every robotic device is constructed to be fail-safe, someone will be working to override those safeguards. 

The makers of computer virus detection software are guaranteed an ongoing market by virus making vandals.   Research is being done to create auto-immune systems for computers that automatically will evolve new forms of protection for new viruses.  However, when such procedures come into use, the virus makers are likely to seize on the same techniques to make self evolving viruses.

If intelligent computers become viewed as a physical danger to society, no doubt some people would try to have laws passed against them. 

It is likely that certain segments of society also would see the use of voluntary brain implants as being ethically and/or morally unacceptable and work toward laws against those practices.

Which side would you be on?

Can artificial thinking machines be sentient?

Can they have free will?

Can they have souls?

What constitutes consciousness?

How would know if the machine had consciousness?

If machines can be sentient, should they have legal rights?

Are thinking machines a physical threat to us?

Should the use of artificial intelligence be legally limited?

Whatchathink?