Posts Tagged ‘reproducing artificial intelligence’
Bacteria running supercomputers?
Apparently, you should not worry when using future computers that will be run by “smart” bacteria.
Bacteria are different from microbes because their life span is pretty short, they don’t occupy much space, and can develop new faculties to compete for nourishment. That last characteristic of bacteria of being excellent in the competition struggle is worrisome to me: I have seen its effect among mankind and feel that it would be wiser for me to purchase a traditional and less performing computer.
How smart bacteria can develop to be?
Laura Grabowski, at the university of Texas-Pan American of Edinburg (USA), is experimenting with how far bacteria can become intelligent. She placed a colony of bacteria in an environment poor in food. A hundred generations afterward (mind you that bacteria do not live long) somehow a single bacteria decided to descend to the lower box rich in food. In an environment of plenty, a new colony of bacteria expanded greatly; with new faculties.
The smarter bacteria can learn to follow computer instructions such as direction to finding food and a weird instruction “Redo what you have done the latest time“. I think that Robert Pennock of Michigan State University at Lansing (MSU) went overboard when he said: “Bacteria developed memory quicker than mankind. Following instructions requires a form of intelligence capable of evaluating situations, realizing we had taken the wrong route, and then reconsidering available data” (again, recall that bacteria have short life span, and a million generations of bacteria would not need 100,000 years as mankind). Ask a member of mankind to recall details of yesterday; he won’t remember much of interest: the work is boring as usual and he had no desires what-so-ever to do anything, and thus, cannot remember worthwhile details.
For example, Ryo Taniuchi in the university of Tokyo has taught E. coli bacteria to successfully playing “Sudoku” of 9 columns and 9 lines with 81 types of bacteria. The bacteria were using parallel calculus to filling all the cells simultaneously, a task impossible by man, using simple rules.
The limiting factor is that there is a limit for the quantity of ADN to be inserted in the bacteria genome.
Martyn Amos said: “Take a colony of ants: an individual ant is not useful, but if you get million ants to come together they are capable of very rich and very complex collective behaviors.”
Laura Grabowski stated: “these organisms are in an environment having to face precise obstacles that demand a form of memory to navigate in. At least a short-term memory must have developed to performing orientation problems. In general, researchers prefer to endow computer with complex intelligence; I opted for the alternative of reproducing artificial intelligence by developing faculties with simple organisms that had only the faculty of procreation.”
To demonstrate the feasibility of her alternative, Laura Grabowski taught “smarter” bacteria to moving toward light sources. These bacteria were introduced into the robot Roomba (a vacuum cleaner) and followed an algorithm of instructions guiding bacteria toward lighted sources.
Smart bacteria are called Avidians in reference to the computer Avida of MSU were these organism live and auto-replicate according to computer instructions. “Avidians are wonderful evolving domestic animals” said Ben Kerr of University of Washington at Seattle.