Intelligent Machines

According to Marvin Minsky, intelligent machines will preserve maps of human cognitive processes for all eternity. Minsky, a professor of electrical engineering and computer science at the Massachusetts Institute of Technology, spoke on the future of intelligent machines on May 25 at the Beckman Center Auditorium as part of the Chancellor’s Distinguished Fellows Series.
‘Ultimately, we should be able to loan entire cognitive processes into one of these machines the size of a cubic centimeter and live forever,’ Minsky said. ‘If you build something, unless it’s a car, you probably know how to fix it and replace its parts. Immortality would be easy to obtain.’
Intelligent machines would also help prevent human mortality by performing dangerous missions, an idea proposed in many science fiction stories.
‘I grew up in the world of science fiction and someone had a set of stories when it was possible to duplicate people and send them to do risky jobs,’ Minsky said.
Artificial intelligence can not only perform many tasks but also solve problems regarding immigration.
‘In Japan, there are fewer young people so they have guest workers,’ Minsky said. ‘Older people think it’s ruining the culture. If we had robots, that problem might go away unless the robots developed cultures that you didn’t like.’
Although robots can improve the quality of human life, Minsky argued that only robots that perform trivial tasks are being made, instead of more useful robots.
‘All over the world, people are building the same robots. People should be working on something more useful,’ Minsky said. ‘Could we make a clever machine other than those that play soccer? Some people say, ‘No, maybe we can build a machine that can do an ordinary thing.”
Although it is generally thought that computers can only solve problems that are logically formulated, Minsky believes that they are capable of solving analogies.
‘There are few cases of people working on solving problems by analogy,’ Minsky said. ‘Twenty people are making progress while 40,000 are working on statistical learning, neuron networks, problem solving and genetic algorithms. All the effort and money is going into fads that will pass away. Only 20 to 30 people are working on how to get computers to have ordinary reasoning, which is where the future will lie.’
Minsky believes that the development of artificial intelligence has benefited society through programs in games and mathematical proofs.
‘By 1969, people found a systematic way to integrate every program that had a formal integral,’ Minsky said. ‘The American Math Society had a collection of how to do integrals [but] no one needed them anymore because Mathematica could solve any integral.’
According to Minsky, the best method to fix bugs is to start over again, write many good programs and put them all in the machine. He applies this idea of panalogy into every day life.
‘My strategy is if there is something you want to explain, you should have at least three explanations,’ Minsky said. ‘Force yourself to make three or four different explanations and combinations of those will explain different phenomena.’
The absence of bodies has not prevented machines from understanding words. Minsky blames the lack of proper programming.
‘You can understand words without body. The reason why computers don’t understand words, is no one has made a semantic dictionary,’ Minsky said. ‘Common words have a dozen meanings. You need to have representations of different things and links between similar definitions.’
Minsky argued against the popular conception of emotion. He believes that emotions are impediments to thinking.
‘People think of emotions as beyond the reach of machines,’ Minsky said. ‘Emotions are generally simplified ways to think. Most emotions turn off your [cognitive resources].’
Education limits children’s learning methods. Minsky argues that rather than teaching a uniform method of learning, schools should embrace different learning methods.
‘I think the problem is when the kid does something, you cannot see what they are doing on the reflective level,’ Minsky said. ‘Is there any place in education, excluding sports, where you think two thoughts at the same time? How do you teach teachers to recognize new ways to think and not mark down for them?’
Lauren Reyes, a second-year biological sciences major, found the lecture insightful.
‘It was interesting to hear what Minsky thought about the cognitive processes,’ Reyes said. ‘I really liked when he broke apart the conception of emotions.’
Aaron Abajian, a second-year computer science and engineering and math major, found the potential capabilities of artificial intelligent amazing.
‘It’s cool that compact computers will be able to complete tasks,’ Abajian said.