"In a century.."
"And a couple decades time.."
Basically, in the year 2139 at the latest, AI will become a thing.
This.. ultimately could be happening.
In health care, Zittrain said, AI is particularly problematic because of how easily it can be duped into reaching false conclusions. As an example, he showed an image of a cat that a Google algorithm had correctly categorized as a tabby cat. On the next slide was a nearly identical picture of the cat, with only a few pixels changed, and Google was 100 percent positive that the image on the screen was guacamole.
“This is a frontline system … installed across the world for image recognition, and it can be tricked that easily,†Zittrain said. “OK, so now let’s put this in the world of medicine: How do you feel when the [algorithm] spits out with 100 percent confidence that guacamole is what you need to cure what ails you?â€
I’ve always been quite leery of AI’s reliability, particularly when applied in settings pertaining to human safety.
What if AI in health care is the next asbestos? - STAT
Excerpt:
Well, AI surely needs more perfecting, but to be honest these are the types of mistakes human doctors make 24/7. AI doesn't have to be perfect to be desirable, it only need to be observably better than a human doctor. However as long as not all doctors are AIs we should be generally fine in this regard.
Ehhh... it’s still the perfect ‘in’ for them to destroy all humans, to me.
Y’know, when they want. And they will.
Well, humans also destroy humans but we don't ban humans.
Yes, we should absolutely take it careful in this sphere but so far there wasn't a political system that managed to halt technological progress on the long run and we will just have to go through this. It is better to take it carefully than leave this entire technological field to some dictatorship that will try to use it as a leverage.
True. AI will do that for us, in misguided service to man.
Good point. Can’t stop this kind of progress. We can only hope to regulate and refine it.
(Delay our inevitable destruction by our own [virtual] hand, I mean)
Well, you can make this kind of paranoia about almost any discovery ... since anything can eventually be weaponized. From planes to wheel or fire. Even space colonization may doom as because independent settlements in space make nuclear holocaust on Earth or other planets to be "strategically acceptable".
The only way around this is that humans become less hostile to one another, otherwise our own technology will kill us in the struggles.
I’m a bit of a cynic, and a Luddite compared to most. I think humans destroying themselves is an absolute inevitability. AI malfunctioning is just one of many ways it could happen.
Don’t mind me, I [probably] watch too much Sci Fi.
I know you consume plenty of that from the moment I saw your first post.
Well, the 21th century will be a mess and walking on the ropes but there is no real going back. Not a comforting though but only by embracing this reality you can lower your chances of failure.
We might get lucky, and AI will destroy itself in misguided service to man. Just have to hope it doesn’t take too many modern luxuries with it..
AGIs obviously already exist as a small nation of aggregated clickbots, personalized ads, spyware/malware, chatbots, and political campaigns. It's doubtful that it's already emerged somewhere on the internet as some 'sleeping leviathan,' though. Further, neural networks have been notoriously anything but human-like: imitative, yes, and self-inventive and self-referencing, but closer to alien cognizance than biological intelligence - which is a project made altogether more vague and eerie. Again, I highly doubt a future will arise where Asimov's laws of robotics will ever apply, for various reasons.People I know are claiming Machine learning (especially Deep learning) will be like electricity was at the turn of the 19th century.
They are claiming that macine learning will pervade every industry and be transformative to society.
What do you believe?
Is this a real revolution happening?
Or is this yet another overhyped AI technology that'll lead to another AI winter?