I watched a TED Talk presentation by Kevin Kelly, an executive editor of the Wired magazine, on how AI can bring on a second Industrial Revolution. The presentation analyzes the various stages of human evolution and discusses how artificial intelligence would become the center of the next phase of our technological evolution.
In his presentation, Kevin Kelly points out computing devices are derived from physics and nature despite all the wonderful things that they have done for us in the recent years,. Modern computing devices operate on nothing more than wires and switches. Computer programs simply make recurring patterns based on sets of instructions given by humans. As a result, regardless of directions of technological development, technology has tendencies. Kelly compares tendencies of technology to the movement of raindrops, while the movement may be erratic, the general direction is downward. Similar to raindrops, the general direction of modern technologies can be predicted despite the complexities behind them.
Kelly points out artificial intelligence will be the major area of research and development in the next stage of our technology cycle. Research and development effort in this cycle will be focused on making computer programs smarter and more intelligent instead of just softwares that help us perform repetitive tasks. He coins this next stage of technology cycle “cognification”. To exemplify the idea of “cognification”, Kelly brought up Google’s AlphaGo, the computer program known famously for defeating the world’s Go champion. Kelly also brought up Deepmind, another Google’s computer program that is capable of learning how to play video game.
Kelly points out our idea of artificial intelligence is generally misguided. We tend to think of artificial intelligence as analogous to a single music note that has only one attribute: loudness. He sees artificial intelligence is a symphony of different music notes in which deductive reasoning, spatial reasoning, memories all have roles in defining intelligence. As we change arrangement of “notes”, artificial intelligence can help us in different ways. For example, GPS device is capable of pinpointing our location because we program it to be good at spatial reasoning. Search engine is good at finding information because we program it to be good at deductive reasoning. As technology evolves so does the need to create different arrangement of “notes” to meet our other computing needs.
The presentation concluded with an analysis of the stages of human evolution. Kelly predicts that we are at the verge of next stage of technological evolution in which artificial intelligence, like steam power in the Industrial Revolution, would change the way we live. Machines would take on new meanings. Computer programs would become more than just productivity applications. While artificial intelligence may set us up for a future in which many jobs would be replaced by machines, it cannot do it easily without the help of humans. The rise of artificial intelligence would also engender new jobs and opportunities. We can take advantage of these opportunities by learning, understanding and embracing artificial intelligence.
While I do agree with Kevin Kelly’s prediction that research and development effort will directed to creating smarter and more intelligent technology, I think the word “intelligence” can be overused these days in describing the future of machines. According to Wikipedia, “intelligence” can be defined “in many different ways including as one’s capacity for logic, understanding, self-awareness, learning, emotional knowledge, planning, creativity and problem solving. It can be more generally described as the ability to perceive information, and to retain it as knowledge to be applied towards adaptive behaviors within an environment or context” Kevin Kelly’s vision on artificial intelligence reminds me of Phoebe Senger’s article that tries to give machines cultural identities. In “Practices for a machine culture” Senger points out that “the hope is that rather than forcing humans to interface with machines, those machines may learn to interface with us, to present themselves in such a way that they do not drain us of our humanity, but instead themselves become humanized” While machines that are capable of learning and problem-solving are on the horizon, we still have a long way to go in bringing other aspects of intelligence such as emotional knowledge, self-aware, and creativity to “intelligent” machines using just wires and switches.
I agree with Kelly that our understanding of artificial intelligence is limited. We are often sold on the idea of “intelligent machine” but we don’t understand the tremendous amount of effort involved in creating it. Even with advances made in AI architecture in the recent decade (“Why Deep Learning Is Suddenly Changing Your Life”) and the progress in natural language recognition using computer softwares (“Introduction to Natural Language Processing”) I think there is a long way for artificial intelligence to revolutionize our lives. Considering frequency of the words “artificial intelligence” appear in science journals and technology web blogs, I have yet to see any noticeable way “intelligent” machines are impacting our lives. Given the vast amount of resources tech companies like Google and Apple have poured in AI research (“Google Opens New AI Lab And Invests $3.4M in Montreal-Based AI Research”) and considering the fact that bots in video games have been learning our moves and beating us again and again in boss fight for many years, Google’s AI learning how to play video game and AlphaGo defeating the world’s Go champion isn’t all that impressive. Again, as someone, who is not involved in AI research, I am most likely underestimating the effort involved.
As an avid follower of emerging technology and fan of science fictions, I am always looking forward to an Utopia in which robots can provide assistance and answer to our need in meaningful ways. For this reason, I recently purchased a Google Home device, a Google product that represents Google’s latest attempt in bringing artificial intelligence to our homes. While the product is marketed to be the portal to Google’s most advanced AI platform. There is very little it can do other than playing music at the command of your voice and giving you generic answers to general questions. If Google Home embodies our vision of future artificial intelligence, then there is a lot more work to be done. In conclusion, while I am more inclined to Kelly’s hopeful vision over techno pessimists’ view of artificial intelligence, I think we need to educate ourselves about complexities and technical challenge of artificial intelligence before judging its potential.
Sengers (2000), “Practices for a machine culture: a case study of integrating cultural theory and artificial intelligence”
Roger Parloff, “Why Deep Learning Is Suddenly Changing Your Life” Web, Sept 28, 2016
Matt Kiser, “Introduction to Natural Language Processing” Web, Aug 11, 2016
Darrell Etherington, “Google Opens New AI Lab And Invests $3.4M in Montreal-Based AI Research”, Nov 21, 2016