Many people tend to get overly excited by recent advancements in AI and tend to generalize the capabilities they see. After all, generalization is a typical human trait. It’s certainly true that Machine Intelligence is now being successfully applied to tasks that not very long ago were viewed as the exclusive domain of humans. In the past couple of years, advancements in Artificial & Machine Intelligence (AI), have resulted in computers handling a variety of specialized intelligence tasks.
Thomas Malone, founding director of the MIT Center for Collective Intelligence, recently published “Superminds: The Surprising Power of People and Computers Thinking Together.” In his book, Tom Malone points out:
“[Human] Intelligence is a very general mental capability that, among other things, involves the ability to reason, plan, solve problems, think abstractly, comprehend complex ideas, learn quickly and learn from experience. It reflects a broader and deeper capability for comprehending our surroundings – ‘catching on,’ ‘making sense’ of things, or ‘figuring out’ what to do.”
The problem is that when people see such specialized intelligence systems at work, they tend to assume that there’s some sort of generalized intelligence at work. For instance, many people have seen the widely publicized outputs from Google’s image labeling system. When a machine labels a picture something like “kids playing frisbee on a sunny day”—we humans tend to assume that the machine has an understanding of what it means to be a kid, what a frisbee is, or what a sunny day feels like. None of that is true. What we are really seeing is a narrow perceptual intelligence that’s making statistical predictions based on exabytes of labeled images that it’s has been extensively trained on.
Machines are good at specialized and narrow intelligence. This specialized intelligence can be applied to address a variety of seemingly sophisticated tasks. But it’s important to remember that what’s really going on behind the curtain is perceptual intelligence and statistical analysis, not human-like intelligence. While it is possible that sometime in the 10+ year future there may be AI-enabled machines that exhibit human-like general intelligence, and which are as smart and adaptable as people, it is impossible to say when or if that will happen.
In “Superminds: The Surprising Power of People and Computers Thinking Together,” Malone writes that while we often overestimate the potential of AI, we often underestimate the potential power of Human+AI combinations. Malone writes:
“As all the people and computers on our planet get more and more closely connected, it’s becoming increasingly useful to think of all the people and computers on the planet as a kind of global brain.”
Today the commercial value of AI lies in using AI in combination with people (Intelligence Augmentation or IA). Intelligence Augmentation involves humans supplying general intelligence, common sense, contextual reasoning and creativity while machines supply vast information, computational power and massively scalable data processing capabilities.