Has Artificial Intelligence Gone Too Far?
An artificial intelligence fooled a human being during a phone conversation in which the machine chatted convincingly with the receptionist of a hair salon while making an appointment for its owner. The apparently all-male crowd of Google workers who listened to this exchange thought it was great, to judge by their cries and claps of delight. But not everyone agrees that the rapid advance of AI is a good thing.
Should We Put The Brakes on AI?
I wonder how loudly they would have cheered if the hoodwinked human was a man, and the AI was impersonating a corporate recruiter doing an initial phone screening. How would you feel about being interviewed by a machine for your dream job, especially if you didn’t find out about the ruse until the ordeal was over?
Yeah, Google CEO Sundar Pichai knew how to choose his non-threatening scenario of this demonstration of Duplex, the company’s prototype of a “conversational AI” that can mimic human speech patterns and mannerisms well enough to deceive a real human. (See my article Should Robots Have To Identify Themselves? for the background on Duplex, and a demonstration video.)
A hair appointment, a job interview, a remote psychotherapy session… at what point do we say, “That’s going too far” with AI? Do we draw the line before or after we arm AI with deadly weapons and set it loose in a war zone? How about Robocop, an AI that walks among civilians meting out “justice” with sometimes deadly force?
The job of a judge is not really judgment but to apply the law and precedent to a given set of facts. Why can’t an AI sit in judgment on a human being? For that matter, why does a jury need to consist of human beings?
Artificial intelligence is being used for self-driving cars, stock trading, manufacturing, healthcare, and many other areas. Should we let AI take over all of the jobs it can or should we reserve some occupations for humans only, just because humans need something useful to do? Where are teenagers supposed to learn discipline, responsibility, and “people skills” if all of the entry-level burger-slinging and cashier jobs are taken by AI?
Can a human infant be raised entirely by machines? That was the origin of “Magnus, Robot Fighter 4000 A. D.,” a comic book hero born in the 1960s and resurrected in the 1990s and again in 2010. Already, AI companion bots for the elderly are being tried in Japan. It would be natural to let robots care for infants, too. Of course, they would have to be the very finest robots, full of love and compassion; but an AI could fake those convincingly, perhaps even more so than a tired, frustrated human mother.
Will AI Have Unexpected Benefits?
Some speculate that AI will open up new employment opportunities for humans who are displaced by software. This happened with the advent of the automobile. Jobs related to horse-drawn carriages were lost, but new ones in design, manufacturing and repair were created. I'm not so sure that the same thing will happen with AI, because the goal there is ever-increasing optimization of processes.
Max Versace, writing in Forbes Magazine, makes the argument that AI and robotics will eliminate only those jobs that are the most unpleasant or dangerous, such as the assembly-line worker, the coal miner and the taxi driver. Robots, he says, will enable a factory worker to become much more productive, perhaps even leading to a four-hour workday and more leisure time.
Guess what, Max? Those workers will get paid for a four-hour workday, too. And their robot buddies won't be taking them out for a round of beers when they leave work at noon. You might feel differently if AI-powered software starts writing magazine articles. (Automated journalism is already happening in the news industry.)
The present forebodes the future. Right now, we are being lulled into complacent acceptance of AI in whatever capacities its creators want to put it, from “personal assistant” to nuclear weapons launcher. Yeah, Mr. Pichai and his fellow AI salesmen know how to choose their scenarios.
Even Elon Musk, whose futuristic endeavors have benefitted from AI technology, says we should be cautious (even afraid) if AI is allowed to develop with no guidelines or oversight. Vanity Fair's article Elon Musk’s Billion-Dollar Crusade To Stop The A.I. Apocalypse is an eye-opening read. We, the People, need to shrug off the glamour of AI and seriously ponder how far we are willing to let them go, while we are still in a position to control them.
By “them” I don’t mean artificial intelligences. I mean their human creators and enablers. Your thoughts on this topic are welcome. Post your comment or question below...
This article was posted by Bob Rankin on 10 Aug 2018
|For Fun: Buy Bob a Snickers.|
Geekly Update - 08 August 2018
The Top Twenty
Does Chromebook Deserve Laptop Respect?
There's more reader feedback... See all 26 comments for this article.
Post your Comments, Questions or Suggestions
Free Tech Support -- Ask Bob Rankin
Subscribe to AskBobRankin Updates: Free Newsletter
Copyright © 2005
- Bob Rankin - All Rights Reserved
Article information: AskBobRankin -- Has Artificial Intelligence Gone Too Far? (Posted: 10 Aug 2018)
Copyright © 2005 - Bob Rankin - All Rights Reserved