AI: Of, By, and For the One Percenters

September 28, 2019

I read “At Tech’s Leading Edge, Worry About a Concentration of Power.” You can too if you pay the Gray Lady or have a dead tree version of the estimable newspaper.

The main point of the write up is that doing smart software with machine learning and lots of data is expensive. Therefore, if a person struggles to pay the rent, smart software is going to be out of reach.

Sure, Amazon offers deals, but the fees for big time machine learning can be beyond the reach of the average country club member. Even a pro athlete with a history of interesting tweets may not be able to handle the invoices from Google, Microsoft, and other cloud vendors.

The newspaper observes against these somewhat poorly kept smart software secrets:

Computer scientists say A.I. research is becoming increasingly expensive, requiring complex calculations done by giant data centers, leaving fewer people with easy access to the computing firepower necessary to develop the technology behind futuristic products like self-driving cars or digital assistants that can see, talk and reason.

Is there a fix?

Well, sort of. The New York Times pointed to foundation support; for example:

At the Allen Institute in Seattle, Mr. Etzioni [former professor and online expert] said, the team will pursue techniques to improve the efficiency of artificial intelligence technology. “This is a big push for us,” he said. But Mr. Etzioni emphasized that what he was calling green A.I. should be seen as “an opportunity for additional ingenuity, not a restraint” — or a replacement for deep learning, which relies on vast computing power, and which he calls red A.I.

Net net: Smart software requires big bucks, big brains, big computing, and big effort. Can innovations emerge from a lab like the one beleaguered Tesla operated?

Maybe, just not probable. When big outfits “help”, the opportunity for “borrowing” may be tempting. In an ethics free zone, who wins?

The one percent. What’s different this time?

Stephen E Arnold, September 28, 2019


Comments are closed.

  • Archives

  • Recent Posts

  • Meta