Robotics Reference
In-Depth Information
made up of two elements, for example water (made of hydrogen and oxy-
gen) and iron oxide (which consists of iron and oxygen), in order to teach
it plausible combinations of elements and their proportions. He trained
the filtering network so its output correctly gave the
Moh's score
for each
molecule (the standard measure of hardness), so that this second ANN
could weed out those with a low score. When it was set to work the
machine correctly identified known ultra-hard materials such as boron
nitride and boron carbide, even though it had never seen these materials
during its training. The machine also suggested several untried combi-
nations of the elements boron, beryllium, and carbon, each doped with
small amounts of hydrogen. These results were sufficiently impressive
to enable Thaler to license the Creativity Machine to Advanced Refrac-
tories Technology, Inc., to aid the company in the development of new
ultra-hard materials and high-temperature superconductors.
Thaler's research has a decidedly commercial focus, with the result
that he has not published any technical details in the scientific litera-
ture. Perhaps because of this, his work has not yet received the acclaim
of mainstream AI researchers, some of whom regard Thaler's claims as
being rather bold. Thaler reacts robustly to their disbelief: “But look at
Columbus. He was an outcast for the most part, with a lot of ridicule and
scorn. Look at Galileo, excommunicated by the Pope, possibly about to
lose his life. The community had not made the same leap they had.” [4]
Genetic Algorithms
Living organisms are consummate problem solvers. They exhibit
a versatility that puts the best computer programs to shame. This
observation is especially galling for computer scientists, who may
spend months or years of intellectual effort on an algorithm,
whereas organisms come by their abilities through the apparently
undirected mechanism of evolution and natural selection. [5]
Genetic algorithms are sections of computer programs that attempt to
mimic the process of natural evolution. Their goal is to evolve a solution
to a problem. The algorithm starts by devising a population of possible
solutions, and then subjecting each member of this population to op-
erations akin to the processes of natural selection, including mating and
mutation. As a result of these processes, and in accordance with Darwin's
principle—“survival of the fittest”—the population of possible solutions
gradually becomes “fitter”, until, hopefully, one of the members of the
