To solve all the problems that I cannot solve on your own.
AI is to solve the science of automatic problem. In particular, the automatic Learning of problem solving.
you are likely to in the foreseeable future, safer than people.
most of The PWs is still 23 out of 24 hours of standing around today. Fleets of robot taxis, first of all, especially in cities, will be far more efficient constantly on the move. From customer to customer, and the offer automatically adjusted to the local Smartphone-based demand. The most common trips will be short. Therefore, you will drive for cost reasons, the majority of Taxis are electric, and with fairly small batteries. Taking a Taxi will not be going to the next charging station when it’s hungry. This is the main problem of today’s E-Mobile, their in spite of heavy batteries low maximum range deal. The rare long taxi journeys are undertaken by the relatively few Autonomous petrol engines in the fleet. E-Mobile are much simpler in construction than gasoline, and therefore almost twice as durable. Therefore, you will only need to build half as many cars. What is drastically shrinking, the number of simultaneously moving cars, but the demand for Parking. This could revolutionize urban planning.
Already, but a significant part of the value of a car will be in his AI Software.
Some people say that it would work in KI, especially data, therefore, large Chinese and American companies with access to many of the users would dominate the data and soon the AI. This is in many respects far-fetched. How to learn for a Baby to be smart? Not the Download of data by a search engine or a social network. No, it learns by it creates its own data through self-invented experiment with toys etc, It learns the consequences of his actions predict. It improving predictive model of the world and of physics, in order to become a better problem-solver uses his all the time. We already know how to build the KIs, which is also a little how babies learn, through what I have called a few decades ago “artificial curiosity”. This will be important for the next wave of KI, sometimes called the “fourth industrial Revolution”. It’s smart robots and other machines that make their data by their own Actions. The next, much more powerful KI wave will revolutionize the entire industrial production. And if you do it right, Europe will play the Central role.
Nnaisense in Lugano – build because of the new super brain? Yes. We want to create the first General-purpose AI that can learn quickly what you can. On the way, we have closed contracts with some of the most famous companies in the world, including in the Smartphone area. Some time ago Nnaisense won against more than 400 competitors in the world of the “Learning to run”competition. Applications of our AI are also found in the automatic portfolio management and in the control of industrial processes. We build, by the way, no Hardware – our AI Software learns instead to control the Hardware of our partners. In a Demo project our artificial brain learned, for example, without a teacher, model cars from Audi’s to Park. In another recent Demo project, the KI of Nnaisense learn to control sophisticated pneumatic robot hands of the company Festo.
(smiles) no, it is still months, if not years, to KIs will surpass humans in any appreciable way.
Jürgen Schmidhuber (56), born in Munich, and is scientific Director of the Dalle Molle Institute for Research in Artificial Intelligence (IDSIA), Manno TI, Professor of artificial intelligence (AI) at the University of Lugano, Professor at Supsi, as well as co-founder and chief scientist of the Nnaisense SA in Lugano.