Artificial intelligence/prospects

There is also an AI wikibook here: b:Artificial Intelligence

This learning resource is about Artificial Intelligence (AI). It is still in a very early stage of development, so any contributions are greatly appreciated. Although pages are currently very high level, this book should be expanded down to the algorithm level.

AI & What It Means Today

The Merriam-Webster Online Dictionary states the following of AI:

Main Entry: artificial intelligence

Function: noun

1. the capability of a machine to imitate intelligent human behavior
2. a branch of computer science dealing with the simulation of intelligent behavior in computers

AI cannot replace battlefield technicians or platoon commanders, yet. But AI research and development is continually advancing; and although it may seem to be moving at a snail's pace, computer technology in general is accelerating (in a pattern known as "Moore's Law"), and may provide the computational capacity required for truly intelligent programs within a mere decade or two.

There are many who believe that creating an intelligent machine is impossible. That's what they used to say about men flying, and after the Wright Brothers learned how to fly, people said it about putting men in space. And now there's a flag on the moon (courtesy of Neil Armstrong), and we have the International Space Station in orbit around the Earth. So now people are saying it about intelligent computers: that they are impossible. But history has shown that just because something can't currently be done, this doesn't mean it won't be done in the future.

Computers can now do many things that before only humans could do: play chess, recognize letters of the alphabet, check spelling, check grammar, recognize faces, take dictation, speak, win game shows, and many more. But the naysayers persist. It's odd that once a human ability is successfully automated, the general public no longer considers it to be an example of AI or AI research — this phenomenon is known as the "AI effect". At that point they say, "it's just another program. Computers will never be able to actually reason."

How many mental abilities does it take to make a machine intelligent? Nobody knows, but at the rate which new abilities are being added to computers' repertoires, it may not be very long. AI researchers see intelligent machines on the horizon, even though the general public does not.

Sony has created a robot named Asimo, that gives tours of Sony's headquarters in Japan. It looks you in the eye when it's shaking your hand, and it walks, even up and down stairs. NASA has developed an anthropomorphic robotic hand as a component of its Robonaut, including the software to enable it to do many of the things a human can do with his or her hands. Other companies or organizations are working on other body parts, such as robotic legs and feet, and visual systems, while others endeavor to reverse engineer the subsystems of the human brain, like the hypothalamus (which controls memory).

AI, Our Future?

Artificial Intelligence (AI) is a young field with massive potential for growth. This article encourages the reader to consider what philosophical and cultural impacts this growth could have upon our society.

Artificial Intelligence has already had a significant impact on our world, enabling developments that were unimaginable only a century ago. “Intelligent” telephone networks route calls more efficiently than any human operator. Cars are constructed in largely unmanned factories by automated robots. Artificial intelligence is even being integrated into common household items such as the humble vacuum-cleaner.

Artificial Intelligence has replaced human workers in a number of fields that involve relatively simple and repetitive tasks. Computers possess many advantages over human workers: they are consistent, they don’t tire and they never demand a pay rise (or a salary for that matter!). However, the computer revolution has resulted in the creation of jobs in a number of new fields such as software engineering and CPU design.

Although Artificial Intelligence is currently too simple to compete with human beings in these more cerebral jobs, this will likely change as the field advances. If Moore’s law continues to hold, we can expect the complexity of computers to rival that of the human brain by the year 2020. Furthermore, some computer scientists predict that, in much the same way that Computer Aided Design (CAD) facilitated the creation of more powerful computers, the AI field will accelerate as AI becomes increasingly able to assist in its own development.

Questions

• Could AI be developed to create artificial art? What impact would this have?
• How would the economy adjust if computers and robots supplanted humans in a majority of jobs?
• If human labor were rendered obsolete, what would civilization be like?
• Is there a difference between ‘having simulated intelligence’ and being intelligent? If so, what is it?
• A self-replicating technology could have effects reaching far beyond the act of creating the first one. If humans become capable of making machines as intelligent as humans, then it follows that such computers or robots could make more of themselves. Should measures be put in place to prevent or limit the development of such self-replicating technologies? Given the nature of competition and political conflict between nations, is this even possible world wide? Given the nature of the availability of computer know-how to the general public, is this possible at all?
• Would an artificially intelligent entity capable of replicating itself qualify as a life form?
• Strong AI is (hypothetical) artificial intelligence that is as smart as a human. Assuming that strong AI would be recursive, that is, capable of improving its own design, what effect would that have on human civilization?
• Could AIs have emotion? Would they? Should they?
• Would AIs based on human brain anatomy be subject to mood shifts, neuroses, or PMS?
• Could an AI go insane?
• Could an AI become evil?
• Could this be prevented? Could it be stopped if it occurred?
• What role would AIs serve in human government?
• Is it possible to control strong AI? Or would it become self-directing? If so, how?
• If an AI became self-aware and sued for rights, what would happen? What should happen?
• If you woke up one day to discover that you were a copy of yourself uploaded into a robotic body, and you were owned by a corporation, what would you do?
• If an intelligence were inherently designed to be satisfied with subservience, would that be slavery?
• To what extent could AI be constrained to human-like thinking? Assuming it could, should it be?
• Could AIs take over? How?
• If AIs acquired the ability to communicate via text (via inter-relay chat, email, web content, etc.), how would you be able to tell their text from text entered by a human?
• Based on how integrated computers and computer programs are in society today (especially within our communications systems), if AIs did take over, would we even notice it at the time? How?
• If AIs took over, what would happen to humans?
  • Would we be treated as equals, or as second-hand citizens, pets, pests, prey, prisoners (criminals, enemies), livestock, slaves, zoo exhibits, or medical specimens?
  • Would humans be moved to reservations? Or be domesticated, trained, fought, hunted, exterminated, imprisoned, raised (to be milked, shaved, or slaughtered for food, fuel, or parts), enslaved, placed on exhibit, or experimented on?
  • Would humans be modified? Limited, improved, or upgraded?
• Would AIs become capable of space travel?
• What would happen to intelligent species they encountered out there? Should we care?
• Did a robot write this list of questions? How could you know? Were you there to see who wrote it?

These are questions that the human race may well face some day, and the foundations are being laid today. Most of them relate to one overarching theme: with artificial intelligence in it, what would the world be like?

Additional reading

• Man and the Machines