Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
It's the middle ground: AI sufficient to do the next tier of jobs, jobs that do not involve welding, turning a wrench or nailing shingles. This AI will be pretty much a moron at everything else but its developed task (accounting, analysis, programming, network management, customer service) - but then, we have a world full of people who are pretty much morons at all but one or two things.
I think jobs requiring manual dexterity are a bigger challenge than replacing knowledge/mental workers.
(And I find it a little amusing that you consider autonomous vehicles to be low-end robots.
They're "low end", because the tasks incumbent upon them are well-defined input-output processes. There are things to sense (the road, other vehicles, obstructions, possible pedestrians), things to do (modulate throttle/braking/steering, signal a turn), and a defined "mission" (deliver goods/passengers from Point A to Point B). Also, a self-driving car is an evolutionary development on what's already been done for decades in automatic controls... fuel injection and engine-management, anti-lock brakes and so forth.
Much harder would be to automate, say, dentistry. It's just drilling-and-filling... but... how to interact with the patient? How to make a diagnosis? How to inject Novocaine via a syringe, in the right place, without hurting the patient? And so forth. Strictly speaking, this job doesn't require "creativity". There's no research to be done, no new knowledge to propose, discover or verify. There is a defined mission (fix cavities), things to sense, things to do. But it's much, much harder than to safely drive a truck in city-traffic.
And, to reiterate, nowhere have I implied that there's anything special, desirable or difficult about making a machine structurally/functionally resemble a human. The AI dentist could look like a CNC milling machine -
a box into which the patient is placed, which changes tools for the various tasks of a dental procedure.
Indeed, dentistry is a bit like machining (cutting of metal in a factory), isn't it?
Well we ever see an AI dentist? Eventually... probably.
"AI will change everything"? "The implications of AI"? "Why AI will make most of us irrelevant"? "The coming genocide"? "We're focked if we don't get our ***** together, and soon"?
Right, that's been my thread of contributions, here and in other, uh, threads... but this topic has turned into the usual six-posters-on-different-tracks mess.
If you want to discuss the future of AI and humanity, a new topic with a more focused title might keep out both the steampunk crowd and the, uh, chicken littles.
Quote:
I don't know why I chose robotics. I think it must have been popular way back them. I suppose words mean different things to me than they do to most people.
Completely irrelevant to this topic, that's exactly it. I don't think you can have a meaningful discussion of the weather or what's for dinner, much less complex world issues, without agreement (or at least mutual understanding) of the terms being used. Here, I think "robotics" led a good topic down some completely detached side roads. Same for most economic terms - everyone has their definition; far too many assume that's a universal definition. Away goes a discussion that ends up straight from the pen of Lewis Carroll.
They're "low end", because the tasks incumbent upon them are well-defined input-output processes.
So think most proponents, because most proponents are engineers trained in... input-output processes.
Just as one isolated aside - what about the instance of a completely unavoidable collision involving several vehicles? Just like any good driver, Tesla's AIdrive has done everything right... but some drunken or distracted fool hasn't. Five cars including the drunk and the Tesla are about to tangle. How does "low level input/output processing" make an optimum judgment in such a case?
Quote:
Much harder would be to automate, say, dentistry. It's just drilling-and-filling... but... how to interact with the patient? How to make a diagnosis? How to inject Novocaine via a syringe, in the right place, without hurting the patient? And so forth.
At least from my perspective, you've both contradicted yourself and mashed together several elements that don't necessarily need to be there.
First, an AI dentistbot would not need to make a diagnosis. The patient would be handed over to the DentalBox™ after a competent human did so. (Just like pathologists sit in labs and hand out one judgment after another, based only on imaging and test results.)
Second, the bot would not have to "interact with the patient" by telling him/her long stories about sailing that weekend (and would not ask questions just as the patient's mouth is crammed with gear!) The tech for a soothing, intelligently directed voice interface exists now. (I've even had CS-bots joke with me.)
Third... if you think an autonomous vehicle is here/possible, then things like analyzing visual cues, realtime imaging, and the diagnostic input, then implementing simple, feedback-sensitive processes should be no trouble for a dentist-bot. It's just smaller-scale and one-on-one. You do realize that there are medical bots in increasingly wide use that do just that? Not with complete autonomy, but carrying this to a streetside box that random passersby step into for tooth pain is a bit... science fictiony.
Fast food places are already replacing workers with self-ordering kiosks. I think it is Amazon that's launching a test store that needs only one employee wandering around; everything is automated, including checkout.
There are millions of jobs that don't require much more than self-service machines to replace the human. As Yang said in the article, "five to 10 years before truckers lose their jobs" - that's one million people. Now add cashiers and retail workers, as another example, who are more than 6% of all the employed in the US. It won't require super, sci-fi-worthy AI to displace many millions of workers.
The greater question to my mind is, in any sci-fi scenario, what happens when humans become not only redundant, but potentially a liability? Unemployed people don't pay taxes, and corporations aren't interested in supporting them, either, which is, fundamentally, Yang's solution.
While I agree with your post, humans are not good at seeing what new fields would emerge, that are not even in infancy stage yet. There will be some.
Google, Amazon, Facebook..were not even a thought just a few decades ago.
Will it fully offset automation losses? I doubt it, but again knowing how big an employer fields will be that do not exist now is not possible accurately.
While I agree with your post, humans are not good at seeing what new fields would emerge, that are not even in infancy stage yet. There will be some.
But AI will be good at seeing what will emerge.....
The problem is that the niches where humans have enough superiority to justify paying them a good wage are dwindling. And may disappear entirely.
Please register to post and access all features of our very popular forum. It is free and quick. Over $68,000 in prizes has already been given out to active posters on our forum. Additional giveaways are planned.
Detailed information about all U.S. cities, counties, and zip codes on our site: City-data.com.