Geoffrey Hinton Quotes

I have always been convinced that the only way to get artificial intelligence to work is to do the computation in a way similar to the human brain. That is the goal I have been pursuing. We are making progress, though we still have lots to learn about how the brain actually works.

The paradigm for intelligence was logical reasoning, and the idea of what an internal representation would look like was it would be some kind of symbolic structure. That has completely changed with these big neural nets.

This is a magical time for computer programming. We are moving towards the artifical intelligence holy grail of programming the computer to simulate the human brain.

The thing that's going to make artificial intelligence so powerful is its ability to learn, and the way AI learns is to look at human culture.

Moore's Law-based technology is so much easier than neuroscience. The brain works in such a different way from the way a computer does.

There's a long history of artificial intelligence programs that try to mimic what the brain is doing, but they've all fallen short.

I've always believed that human learning is the result of relatively simple rules combined with massive amounts of hardware and massive amounts of data.

A deep-learning system doesn't have any explanatory power.

I think what's really amazing is that given the scale of the web and getting the compute power we have today, we're starting to see things that appear intelligent but actually aren't semantically intelligent.

If you want to solve very complex problems, you will have to end up letting machines work out a lot of the details for themselves, and in ways that we don't understand what they are doing.

In fact, there was general agreement that minds can exist on nonbiological substrates and that algorithms are of central importance to the existence of minds.

My mission at Google is to develop natural language understanding with a team and in collaboration with other researchers at Google.

It seems like everything that we see perceived in the brain before we actually use our own eyes, that everything we see is coming through computers or machines and then is being input in our brain cells. So that really worries me.

The computer is not, in our opinion, a good model of the mind, but it is as the trumpet is to the orchestra - you really need it. And so, we have very massive simulations in computers because the problem is, of course, very complex.

We could construct a machine that is more intelligent than we can understand. It's possible Google is that kind of thing already. It scales so fast.

Artificial intelligence will reach human levels by around 2029. Follow that out further to, say, 2045, we will have multiplied the intelligence, the human biological machine intelligence of our civilization a billion-fold.

I don't think there's anything unique about human intelligence.

Not only have computers changed the way we think, they've also discovered what makes humans think - or think we're thinking. At least enough to predict and even influence it.

Whatever you are studying right now, if you are not getting up to speed on deep learning, neural networks, etc., you lose. We are going through the process where software will automate software, automation will automate automation.

Data is going to be dynamic... and, combined with the performance of computing that allows for artificial intelligence to be applied, is going to bring all kinds of insights and additives to enhance everyday life.

None of us today know how to get computers to learn with the speed and flexibility of a child.

The deep learning techniques, while relatively easy to learn, are quite foreign to traditional engineering modalities. It takes a different mindset and a relaxation of the presumption of control. The practitioners are like magi, sequestered from the rest of a typical engineering process.

A new, a vast, and a powerful language is developed for the future use of analysis, in which to wield its truths so that these may become of more speedy and accurate practical application for the purposes of mankind than the means hitherto in our possession have rendered possible.

Most of the value of deep learning today is in narrow domains where you can get a lot of data. Here's one example of something it cannot do: have a meaningful conversation.

Watson augments human decision-making because it isn't governed by human boundaries. It draws together all this information and forms hypotheses, millions of them, and then tests them with all the data it can find. It learns over time what data is reliable, and that's part of its learning process.