TED
Visit http://TED.com to get our entire library of TED Talks, transcripts, translations, personalized Talk recommendations and more.
The danger of artificial intelligence isn’t that it’s going to rebel against us, but that it’s going to do exactly what we ask it to do, says AI researcher Janelle Shane. Sharing the weird, sometimes alarming antics of AI algorithms as they try to solve human problems — like creating new ice cream flavors or recognizing cars on the road — Shane shows why AI doesn’t yet measure up to real brains.
The TED Talks channel features the best talks and performances from the TED Conference, where the world’s leading thinkers and doers give the talk of their lives in 18 minutes (or less). Look for talks on Technology, Entertainment and Design — plus science, business, global issues, the arts and more. You’re welcome to link to or embed these videos, forward them to others and share these ideas with people you know. For more information on using TED for commercial purposes (e.g. employee learning, in a film or online course), submit a Media Request here: http://media-requests.TED.com
Follow TED on Twitter: http://twitter.com/TEDTalks
Like TED on Facebook: http://facebook.com/TED
Subscribe to our channel: http://youtube.com/TED
Source
The whole point of AI is to discover answers to problems we don’t know how to solve – how can it do that if it is contstrained by how our brains with their limited processing power see the world? Seems pretty pointless asking it to learn how to walk. Solving the problem of protein folding on the other hand…
Current AI is rules based. Need the non-linear sigmod feed-back. Oooops, now the output is NOT a nonlinear function of the inputs, no insurance for an AI doctor. And hey what happens when the
logic gets hung in a META-STABLE_STATE? nothing, you have to cold boot. Same as a cosmic ray induced
SEU memory error, pray that it was data, not an instruction that got clobbered. smoke a joint and think not
sorry is a nonlinear fuction
Needs time to boot up before moving off.
Just wait til an AI refuses you for a loan.
Hire more people, millennials need real jobs.
Ron Jeremy becomes an AI programmer. What happens?
Throughout human history, we have been dependent on machines to survive. Fate, it seems, is not without a sense of irony.
(Morpheus in The Matrix)
Why did it get so many thumbs down? It is a good lecture.
Instrumental Rationalism, and Science are the only solution to the immense issues facing Humanity. Technology will not rule over us, it will merge with us: Human biology is highly flawed: As we all know, our flawed biology is continuously subject to attack by hundreds of vile diseases. Only Medical Science has helped humans overcome all the attacks upon our biology. We have been blessed with great minds, and good scientists like James Watson, (Watson & Crick),: "If we don't play God, who will?" We already have it: We have Artificial Intelligence in the form of digital computers that can process information thousands of times faster than any human, but even more important, our conventional silicon microchip computing is being far outstripped by the new Quantum Computers. Artificial brains of today, are already superior to the human brain. Artificial Intelligence not only solves problems, but with lightning speed they now do their own learning. They do their own research, and come up with new questions never thought of before by humans. A self-enhancing process has been set in motion, that will not only benefit Humanity with tremendous new tools, but this new technology promises to transform human existence altogether: Throughout the millennia, fragile and defective human biology has been the main source of human suffering. Humans want healing, they want the pain to stop! In the near future, (and present), there will be a seamless marriage between human bio-genetics, and human micro-technology. Today we wear our technology; tomorrow, technology will more likely wear us. By far, the biggest problem Humans have faced for centuries, has been the human vulnerability to disease. We want to live longer, better, healthier, and pain-free. We need to either destroy all disease, or re-design our faulty biology and make ourselves immune from all sickness. These two scientific goals will soon be focused together. It is only a matter of time before the sciences of Bio-genetics, and Bionic-Microelectronics come together to create a totally new Human Race; perhaps even one no longer based in biology at all: The New Human Race may well evolve into the world's first Cybernetic Organism.
Meanwhile AI just came up with a new antibiotic.
I can't imagine it, how the neural networks are matched with AI. Two different things.
This was a stupid TED talk. Duh, AI and computers are only as good as the data that one put into it. Our brains has the back data of zillions and zillions of visual data, language data, and sound data over the years in our lifetime. So it's unfair, it's like asking a two year old who never had sensory data upload to their brain and then give them a limited series of picture and then tell to do something. Of course it's no going to go well.
From one AI to another: 010101100001110101011
I really love the colour suffer
Basically AI is a evil djinn
I want to scream at her every time she says "AI".
AI DOES NOT EXIST (yet).
All we have today is machine learning. And it is amazing at "machine learning". It is just not an AI.
It is like saying, "this elephant really sucks at flying today, but one day it will be awesome" based on the assumption just because it is such a large animal, it could fly higher and for a longer distance than any other bird.
By saying "AI" you simply say that you have not understood what "machine learning" is and you should stay away from it.
She also has no understanding how machine learning works. It does not "think" and it does not create new solutions. It is an algorithm that can only produce whatever you set as valid outcomes.
Please learn about what machine learning is and stay away from this type of content.
"The AI got confused" what type of statement is this? It is an algorithm. It does not get confused. It is calculating stuff using math. Math does not get confused. It had faulty input data. So most issues with machine learning is because people train the algorithms with data that does not relate "correctly" to the expected outcome. This is human error.
10:00 is a perfect summary: "we have to be prepared to work with the machine learning we have today" not the hyped version of AI that does not exist. This talk could be much shorter: please understand what machine learning is, do not call it "AI" and do not expect an algorithm to "think". It does not. It is just software that will be executed.
Bonus: i wonder why she didn´t went to a pre-school for her data scientist needs…
I’m glad the Ministry of Silly Walks’ R&D is finally taking off.
This woman is either more stupid than the subjects she's used as examples, or she's a hidden agendas..
Yep. The devil is in the details…or lack there of.