The Machines are Coming, or Are They?

I was reading this 2007 New York Times article by Jason Pontin about crowdsourcing technology. He specifically addresses the area of organizing and locating relevant information in our increasingly digital society. Jeff Bezos, CEO and founder of leads again with Amazon Mechanical Turk, a service that matches human workers to companies in need of tedious functions that we are, as yet, not able to delegate to our computer helpers. Although the technology hasn’t been fully embraced yet, I wouldn’t write it off. The full article is an excellent read, but here are my favorite bits.

Artificial Artificial Intelligence

In every half decent science fiction story you’ve got the machines taking over in one way or another. We are certainly headed in that direction with Artificial Intelligence (AI), but Bezos has termed this new model Artificial AI.

“Normally, a human makes a request of a computer, and the computer does the computation of the task,” he said. “But artificial artificial intelligences like Mechanical Turk invert all that. The computer has a task that is easy for a human but extraordinarily hard for the computer. So instead of calling a computer service to perform the function, it calls a human.”

As of 2007, Mechanical Turk had already hooked up with over 100,000 workers from all over the world. These private contractors earn pennies when they complete any number of the simple actions called HITs (Human Intelligence Tasks). Two examples in the article are identifying duplicate web pages and matching images to descriptions.

Cost vs. Benefit

In the article, Pontin identifies two significant criticisms of Artificial AI.

1. It’s only as good as it’s strongest members – ie. smart computers can do about as well as dumb people.

2. It takes advantage of workers in poor countries – ie. it creates evil digital sweatshop conditions.

He also states that we probably have at least another 25 years before computers are more powerful than human brains, according to the most optimistic artificial intelligence experts.

So where’s the tipping point?
Forget the evolution of the technology. That’s coming. When we’re talking pennies an hour for human brain power vs. the cost of running the machines on limited resources, what is more cost effective?

  • An interesting twist on AI. I’ve been taught that computers/robots are best at simple repetitive tasks. Humans are good at complex tasks. Having a computer system make a quick decision and source the task to humans…brilliant!