In a rapidly changing age, it is difficult to focus on anything but the immediate future. My generation is arguably the first to be totally immersed by technology. Literate in the vernacular of personal computing we are more than previous generations more comfortable with changing technologies. Following this pattern, you can expect that subsequent generations will be even more technology and media savvy and will expect their technology to keep up with them: make their lives more comfortable and convenience, which is after all what well made technology should do. Because we have grown up expecting prompt technological growth, it is of no surprise to me that Bill Joy mentions in his article "Why the Future Doesn't Need Us," that many major technology changes will be subtle and incremental, that it will not be jarring to us. I discuss futurism, I agree more with Ray Kurzweil's argument about 'The Singularity,' roughly that technology will only encourage longer life spans, encourage creativity and greater ease of life for most people.
While examining technology advances of the past, I am reaffirmed that I should not fear technology, but rather fear what we may do with technology. The practicality of the steam engine provided thousands with jobs world-wide and for the first time in history made transcontinental travel relatively easy and convenient. Technology that does what it should, serve humanity in the best possible way, making our lives easier. However, you should also examine innovation for war's sake. What might these upcoming advances in technology hold for humanity's ability to hurt? While I am a big fan of science fiction, and fear the opportunity for imposing technological integration/alteration, such as the borg in the 'Star Trek' universe. I must also persist to believe that some of humanity will endure to seek justice and an ethical way of life, much like the crew of the Enterprise. We subsequently should not fear technology (as rapid as those advances may be), we should only fear what we may do with technology.
Sunday, November 19, 2006
Friday, November 10, 2006
Questioning the Turing Test
I think the Turing Test is NOT a reliable way to test intelligence, as intelligence is so much more than being able to respond to what someone is saying. As I learned when I was working with Python, it is easy to program a computer to respond to a set of commands. While it is impressive that some contemporary programs are able to adapt to different conversational prompts, the ability to communicate exclusively does not dictate how we measure intelligence. Several factors must be taken into consideration.
What classifies intelligence? I think the primary litmus test must be considering what makes humans intelligent. What makes us superior to animals? Besides the presence of opposable thumbs, the ability to create makes us intellectually superior to most animals. There will always be arguments to this assertion (like that elephant who paints). However, en masse, most animals have not proven the ability to create in the same way humans have. In addition, I believe some animals have displayed a less intellectual level of creativity and imagination. For example, I have four large dogs at home in Arizona. If you watch any of them sleep, you can see them act out their dreams: paws twitch, snouts growl and whimper, etc. As far as I know, no artificial intelligence has demonstrated an ability to create, dream, etc, an essential part of measuring intelligence in my opinion.
There is also something to be said for emotional intelligence. Sure the Turing Test measures a program's ability to respond and communicate with other humans, but can these programs emote? You could argue that there are plenty of bachelor's that cannot emote so surely this cannot be a measurement of intelligence. However all of humanity has emotional potential, the diversity of experiences in our lives is what determines how we maximize this potential. Computer programs on the other hand, do not have this kind of potential and therefore do not have emotional intelligence. Both emotional and creative intelligence must be acknowledged in determine what programs do or do not have intelligence.
What classifies intelligence? I think the primary litmus test must be considering what makes humans intelligent. What makes us superior to animals? Besides the presence of opposable thumbs, the ability to create makes us intellectually superior to most animals. There will always be arguments to this assertion (like that elephant who paints). However, en masse, most animals have not proven the ability to create in the same way humans have. In addition, I believe some animals have displayed a less intellectual level of creativity and imagination. For example, I have four large dogs at home in Arizona. If you watch any of them sleep, you can see them act out their dreams: paws twitch, snouts growl and whimper, etc. As far as I know, no artificial intelligence has demonstrated an ability to create, dream, etc, an essential part of measuring intelligence in my opinion.
There is also something to be said for emotional intelligence. Sure the Turing Test measures a program's ability to respond and communicate with other humans, but can these programs emote? You could argue that there are plenty of bachelor's that cannot emote so surely this cannot be a measurement of intelligence. However all of humanity has emotional potential, the diversity of experiences in our lives is what determines how we maximize this potential. Computer programs on the other hand, do not have this kind of potential and therefore do not have emotional intelligence. Both emotional and creative intelligence must be acknowledged in determine what programs do or do not have intelligence.
Subscribe to:
Posts (Atom)