Debates about artificial intelligence
Since the beginnings of artificial intelligence, researchers have been trying to determine whether a machine can be conscious; this is still unknown. Some researchers predict computers will reach the overall competencies of humans in the 2050's, while others in the field predict that it will happen during the second half of this century.
Researchers in this field have always debated how consciousness should be defined. Every few years a research team announces a newsworthy breakthrough in artificial intelligence, and invariably critics and skeptics are quick to list the many human abilities computers "will never have."
One of the questions that is common in the debate is, "Does an entity have to be a wet, mortal, organic, living being in order to be conscious?"
Another question that arises is, "Living beings are composed of atoms and molecules. Atoms and molecules are not living things. How does life emerge from non-living things?"
Some scientists study the biological chemistry that may have existed when life began four billion years ago. One on-going debate in that field is the question, "Which kind of molecules initiated life? Proteins? RNA? DNA?"
There is one school of thought that says that RNA is the answer, and its theory is known as the "RNA World," and it suggests that DNA and proteins proliferated as a result of RNA.
Scientific evidence has been found, however, that suggests that either proteins or DNA served as the chemical starting point.
Three of the most important kinds of chemical reactions required for life are the replication of nucleic acids (i.e. RNA and DNA), transcription of nucleic acids, and enzymatic reactions; but it is unknown which of these was the most important in initiating life.
Biology has always influenced artificial intelligence research -- for example, one idea for the creation of intelligent programs has been to create a progenitor program from which more intelligent programs will evolve from -- and this idea brought about the field known as artificial life, which is the simulation of evolution by a computer program. Such researchers apply the science of evolutionary biology and population genetics to their programs, and draw inferences from the outcome of the simulations.
Subsequent to the advent of artificial life, researchers applied these concepts to the field of virtual reality. Some call this field virtual intelligence. One idea in these fields is to explore the interactions between virtual humans and virtual, artificially intelligent characters (sometimes called virtual agents). These characters will evolved via the computer science of artificial life, and they will be responsive to the virtual environment, virtual human(s), and other virtual agents. Those in this field feel that a strong aspect of their approach, compared to that of robotics, is freedom from the need to re-build in order to introduce design concepts. Another strength, they say, is the ability to simulate rich sensations experienced by the virtual agent, which, they hypothesize, exceeds those of robots. A third strength claimed by these developers is that it puts the virtual agent and virtual human on a "level playing field" with regards to sensations and experiences.
A related idea is 'foglets," or miniature fogs of nanobots with collective (i.e. swarm) intelligence and morphing ability, a technology with currently rudimentary beginnings which will benefit from artificial life and virtual intelligence research. Ray Kurzweil discusses foglets in his book The Singularity is Near.
The idea of foglets is influenced by the idea that living beings are composed of millions of neurons and cells which all work together as a system. It is a type of robotic system, but its strength is flexibility and physical adaptability. Since it is a kind of robotic system, it meets the need for embodied cognition, which many researchers insist is a requirement for the long-term goals of artificial intelligence. 'Embodied' means "having a body."
Artificial life and foglets are bottom-up approaches. A third such approach will draw from artificial chemistry, which is the simulation of chemical reactions at the molecular and atomic level. The idea is "life evolved from the bottom upwards from simple components and processes toward higher complexity."
When artificial intelligence began, computation was the goal. It did not take long for the development of programs to solve mathematical problems and proofs, a multitude of which had never been solved prior. This convinced the community of researchers that virtually any logical or mathematical problem could theoretically be solved by computers.
There was still debate, however, as to whether a computer thinks, and whether a thinking computer could be made. Turing in 1950 listed some abilities which had been brought up by skeptics:
"Be kind, resourceful, beautiful, friendly, have initiative, have a sense of humour, tell right from wrong, make mistakes, fall in love, enjoy strawberries and cream, make some one fall in love with it, learn from experience, use words properly, be the subject of its own thought, have as much diversity of behaviour as a man, do something really new."
Whether a computer or robot will ever be able to do any of these things is still being hotly debated in the artificial intelligence community and in the field known as philosophy of artificial intelligence.
Some believe that a list similar to the one above should be compiled by worldwide Wikipedia contributors, with the intent to inform researchers and skeptics; students and professionals; scientists and sci-fi writers.
Such a list would also be useful to the philosophy and cognitive science communities -- the idea being to explore the mind and human behavior far beyond the notion of calculation, and to stimulate new conceptual approaches to research.
The next breakthrough in artificial intelligence might need to come from a discipline that is completely unrelated to computer science.
There have been efforts to someday enhance human intelligence through human genetic engineering, neural engineering (including brain-computer interfacing), and stem cell science; and there are projects in the field of whole-brain-emulation, in which brains are imaged at microscopic resolution and synaptic activity is recorded optically. A computer program will then be designed to emulate those patterns.
If mankind manages to significantly increase human biological intelligence, efforts toward artificial intelligence will continue. In such a scenario, the two fields will become symbiotic. Efforts to create cyborgs will continue as well.
The most long-term outlook involves artificial emulation of every human ability, quality, and experience. Eventually those of other life forms will be emulated as well. A scenario this far into the future will involve next-generation synthetic biology. The entire genome of a new species of bacteria was designed, synthesized, and incorporated into a recipient bacterial species several years ago, and this trend will continue on to higher life forms. Self-organizing systems, eventually including foglets, will continue to be an important field in the efforts to emulate hives, swarms, and regenerated organs. Novel proteins will eventually be designed and produced from scratch, thanks to artificial chemistry. Life processes will increasingly incorporate substances and systems designed by man.
-Nicholas July 17, 2015
References
Turing, A.M. (1950). Computing machinery and intelligence. Mind, 59, 433-460.