InfoWorld sticks artificial intelligence the top of a list of IT snake oil, but I think they're being extremely unfair in doing so.
The pitch: "Some day we will build a thinking machine. It will be a truly intelligent machine. One that can see and hear and speak. A machine that will be proud of us." -- Thinking Machines Corp., year unknownOnce upon a time, machines were going to do our thinking for us. And then, of course, they'd grow tired of doing humanity's bidding and exterminate us.
The good news? We haven't been offed by the machines (yet). The bad news is that artificial intelligence has yet to fully deliver on its promises [8]. Like, for example, in 1964, when researchers at the newly created Stanford Artificial Intelligence Lab assured the Pentagon that a truly intelligent machine was only about a decade away [9]. Guess what? It's still at least 10 years away.
This attack on AI betrays a fundamental ignorance about the field. In fact, the biggest failures in the AI community have been over-promising and controlling expectations. As the InfoWorld article acknowledges, there have been a host of successful AI technologies developed in the past decades:
Technologies that seem mundane to us now would have looked a lot like artificial intelligence 30 or 40 years ago, notes Doug Mow, senior VP of marketing at Virtusa, an IT systems consultancy.The ability of your bank's financial software to detect potentially fraudulent activity on your accounts or alter your credit score when you miss a mortgage payment are just two of many common examples of AI at work, says Mow. Speech and handwriting recognition, business process management, data mining, and medical diagnostics -- they all owe a debt to AI.
Add in: Google, TiVo, Netflix, Facebook, airline route management, military planning, weather forecasting, advertising, and thousands more applications. All these advances have descended directly from the field of artificial intelligence.
The difficulty is that as soon as a technology works it is no longer considered "artificial intelligence" -- it becomes just another algorithm.
(HT: RB.)
This reminds me of how someone once defined "technology" as "types of machines that don't quite work yet". The chair, the fireplace, the windmill, the clock, etc, are technically "technology" but they work so reliably that people don't think of them as such. Cars and pacemakers and microwaves might be in this category as well. When you ask people to name objects they relate to the word "technology", they usually say computers and airplanes and other things whose complicated design often becomes apparent due to less-than-reliable functionality.
(Which in turn reminds me of people who say they work in "technology" or in the "technology industry" to mean they work on developing computer hardware, software, websites, etc. And most people tend to understand this meaning when they hear that word. But then I go "I'm helping to design a carbon-fiber jet that runs on bleedless engines over here! And I'm helping to develop new fracture-mechanics equations for its parts! How is that not 'technology'?"...