ICanHasDataInput
IBM has created a supercomputer with enough processing power for “real-time cortical simulation that achieved more than 1 billion spiking neurons, as well as 10 trillion individual learning synapses” exceeding the capacity of a cat’s brain. The technology is predicted to come as close as ever to replicating the human brain by 2019. This, of course, is the foreshadowing of many marvelous possibilities such as the creation of sentient artificial intelligence or the ability to save a person’s mind within a computer system. All of these issues are clearly subject to debate and these arguments are too plenty and substantial for this blog entry but are definitely worthy of perusing.
One issue is especially interesting to me however. That is, the creative power of man over a potentially intelligent being. I think it is worthy of considering the ethical repercussions of being part of a creator-creation relationship. These supercomputers, assuming they are truly intelligent, are intended to be a useful gadget of mankind. Would this then be a slave-master relationship? Is it ethically allowable to create an intelligent agent, biologically present or not, with a specific intended use? First, and foremost, it should be explored whether a supercomputer such as the one built by IBM could actually be “intelligent”. Achieving a brain-like neuron system replication is clearly within reach but this offers little information on whether a computer could have the same mind-like states of a human. The capacity for memory, learning, and decision making is not necessarily tied to the experience of these traits. Perhaps experience is not essential for being a being in the same scope as humans. Other characteristics are also important including free will or emotion. A neural system reproduction is capable of learning and memory but it is not clear as to whether the computer will be able to transcend the user input of the programmer in its output.
These problems can be highlighted by examining the human brain and how a person makes decisions. When faced with a decision, such as what food to eat, a person will typically base his decision on past experience. For instance, if I were to find a mushroom that, in the past, has made me ill, I will probably avoid it. For a computer, decisions can, too, be based on past occurrences but will, ultimately, depend on the foundation of programming that the computer began with. This programming will be the basis for how the computer learns and how it interprets information. So, would a computer, even a supercomputer of near-human capabilities in neural networking, be free from the programmer’s directions?
To return to my original idea of the relationship between the creation and the creator, it is interesting to consider what it means to create an intelligent agent with intended use in mind prior to its existence. In some ways, people have done just this for quite some time with some-what intelligent animals for a wide variety of useful purposes. However, the dilemma of artificial intelligence is a more curious case since this would be the creation of a an entirely new type of being – a being that rises above and beyond the physical form and only depends on a stream of information.