On Artificial Intelligence AI so that new ideas have to learn to think of wisdom
Liwan Hong 2019-11-1
Many scientists puzzled encountered problems are solved revelation in a dream, such as the discovery of the benzene ring, I think AI has recently wisdom of the problem, last night had a dream, the dream of children go to school in the sea.
An old saying: There is a road ground for the tracks, Boundless Learning bitter for the boat. The dream is to prompt the computer should be allowed to go to school for computer classes, let it acquire knowledge, so be wise.
Humanlike artificial intelligence process should be able to learn and apply knowledge to think, this is the wisdom of having AI now, AI is standard on the model a lot of training to complete the identification data through features such as play, not a thought, obviously not has wisdom, AI has been a long time no new breakthrough theory and methods. To this end, we need a new idea, another way to let the AI has wisdom, to learn and think.
Knowledge expression vector text, children learn to think is accomplished through language, therefore, allow the computer to learn and use knowledge is the way to break through the language. Expert system inference engine is the idea of early artificial intelligence, there has been a major breakthrough, which is the direction of a great value.
Enable computers to obtain the knowledge base of knowledge, reasoning by inference engine, which is a logical way, the result can be a logical explanation, and now the AI are based on probabilities, explanatory poor. AI can be based on the current results, improve knowledge representation and the ability to learn, give full play to the power of knowledge, make more intelligent AI.
Knowledge is power, computer baby needs to go to school as a child, he learned, the more knowledgeable, more wise. The text is a bridge, currently NLP speech recognition technology to achieve good results in the translation, the use of the word vector representation terms, it is worth learning from. Word vector can be used to express knowledge, establish rules of inference, so the machine has a strong analytical reasoning ability, can solve many problems in terms of having the actual production value.
Unity is strength, people Qi remove the mountain, success requires everyone to work together, work together to cooperate. Experts in various fields of expertise can be extracted, AI scientists to express this knowledge is capable of storing computer-calculated way to build a large integrated knowledge base, including social and natural sciences, and then create an inference engine based on the characteristics of expression of knowledge, let computer knowledge can make full use of the knowledge base reasoning, thinking human simulation to solve problems and create value. Can be used operations research, genetic algorithms, annealing, big data and artificial intelligence technology to achieve optimal decisions, improve the quality of decision making. So that the computer AI, like people, can learn and think in order to truly have the wisdom, limitless future. The computer can learn all of human knowledge, to become big achievers than Einstein, Qian and other scientists also learned, imagine, this will bring much benefit mankind?
The computer in the reasoning, in addition to forward, reverse and reasoning, but also reinforcement learning RL, against AI technology to improve learning and other learning GAN thinking, and thus become wiser, more powerful, the future is no ceiling .
In short, by allowing the computer knowledge to think that this is an important direction of strategic value AI has practical significance, we work together, not difficult to implement, can build a fairly AI expert wisdom in a year or two system, this is the AI's human right.
