What’s at Stake in the Debate over AI

googleplus Google+
arroba Email

What’s at Stake in the Debate over AI

A Briefing Paper

Human exceptionalism, the view that humans have a unique status in the created order, is foundational to the Judeo-Christian tradition. According to this tradition, humans alone among physically embodied beings are created in the image of God. The superiority of human intellect and judgment is not merely a difference in degree with other living forms but a difference in kind. Notably, human language ability and moral agency are without parallel in the physical realm.

The accelerating performance of artificial intelligence (AI), also referred to as machine or computational intelligence, is increasingly cited as reason to doubt human exceptionalism. Artificial intelligence has defeated Garry Kasparov at chess, beat world champions at Jeopardy, and most recently won a top level contest in the game Go, the most difficult of popular board games. Computers inarguably excel over humans in many areas.

Well-respected scholars have advanced the idea of a singularity, in which computers will soon exceed the abilities of humans in all areas, whereupon humans will either become extinct or either be relegated to the status of a computer’s pet or have the good sense to upload themselves and thereby in fact become computers. Such speculations reject the strong evidence that humans have abilities that exceed any capacities of computers. This includes both computers as they exist now and as they might exist in the future.

Interest in artificial intelligence and concern about its impact is widespread. In the political realm, Russian President Vladimir Putin warns that “the one who becomes the leader in this sphere [artificial intelligence] will be the ruler of the world.” Reputable professional societies speculate about whether artificial intelligence will one day be granted constitutional rights. Entrepreneur Elon Musk warns that artificial intelligence is humanity’s “biggest existential threat.” And Microsoft founder Bill Gates does not “understand why some people are not concerned [about artificial intelligence].” Although much angst about artificial intelligence is highly speculative, there is no doubt its impact will be transformative, whatever the underlying reality. The investigation of this underlying reality being precisely the point of the Walter Bradley Center for Natural and Artificial Intelligence.

The Bradley Center for Natural and Artificial Intelligence is firmly on the side of humans being exceptional and forever remaining exceptional in the face of machines. Human exceptionalism against machines has been advocated at least since the time of Descartes, who hit the nail on the head in his Discourse on Method (Part V) when he wrote:

While intelligence is a universal instrument that can serve for all contingencies, [machines] have need of some special adaptation for every particular action. From this it follows that it is impossible that there should be sufficient diversity in any machine to allow it to act in all the events of life in the same way as our intelligence causes us to act.

Lady Ada Lovelace, the first computer programmer, working with Charles Babbage’s nineteenth century mechanical computer, made a similar point. Her argument has been paraphrased thus:

Computers can’t create anything. For creation requires, minimally, originating something. But computers originate nothing; they merely do that which we order them, via programs, to do.

Are there capabilities of the human mind that will forever escape computer performance? Current indications say yes. Creativity, understanding, natural language, and consciousness are examples. Other proponents of natural over artificial intelligence include:

  • The mathematician Jacques Hadamard, who studied human creativity and described it as beyond the reach of computation;
  • Information theory pioneer Leon Brillouin, who wrote: “The [computing] machine does not create any new information, but it performs a very valuable transformation of known information”;
  • Selmer Bringsjord at Rennselaer Polytechnic Institute, who dismisses modern claims to have solved the Turing Test as depending on the “strength of clever but shallow trickery”; and
  • Celebrated mathematician/physicist Sir Roger Penrose, whose numerous thought-provoking books, papers, and lectures on the topic locates the source of human exceptionalism in quantum-mechanical aspects of the human brain, thereby rendering the brain’s performance beyond the ability of any computer algorithm.

Contrary to the reports from some media, computers are not getting more creative. They are simply getting faster and acquiring more memory. According to the Church-Turing Thesis, Alan Turing’s conceptual computer from the 1930s (the “Turing Machine”) can perform any operation performed by any of today’s computers. Today’s computers can perform operations zillions of times as fast as the computers of Turing’s day. But the limitations of present-day computers, including quantum computers, are the same as those of Turing’s original machine.

Computers are only able to execute algorithms—step by step recipes for performing a task. Turing showed there exist many operations that are nonalgorithmic and therefore noncomputable. For example, a diagnostic program cannot be written to determine what a second arbitrary computer will do, including whether the second program will eventually halt or run forever (this is “the Halting Problem”). Likewise, it appears that the creativity of humans is noncomputable and is therefore unable to be reproduced by artificial intelligence. John Searle’s Chinese Room allegory illustrates that computers do not understand the operations they perform. The inability of computers to gain consciousness or sentience seems insurmountable to artificial intelligence.