Geoffrey Hinton: Give up the reverse transmission, our artificial intelligence needs to start again _ Artificial intelligence

Source: Internet
Author: User
In 1986, Geoffrey Hinton collaborated with someone else on a paper ("Experiments on Learning by-back propagation").
In recent years, the midpoint of this paper propagation has become the key point for AI to develop rapidly today.
But just recently, Hinton said the back propagation method should be discarded, and instead explore another new path to the AI.
Hinton said he now "deep doubt" back propagation, it is this method to accelerate the AI field in recent years most of the progress, including images, NLP, voice and other fields.
Hinton said: "My point is to abandon everything and start over again."
Hinton that in order to achieve real progress, a new approach must be created.
Max Planck once said: Science goes one step at a funeral. The future depends on the graduate students who are very skeptical of everything I say.

"I guess this means abandoning the reverse, which is not the way the brain works, and obviously we don't need to label all the data," Hinton said. ”
Hinton's point of view has aroused strong repercussions on Twitter.
Li Feifei commented: "Reverse communication is very important, it is like a Rolls-Royce engine on the plane, although the aircraft can not be as smart and flexible as a bird." Agree with Hinton about the idea that no tool is permanent, even reverse communication or deep learning is no exception. Therefore, the continuation of basic research is extremely important. ”
Gary Marcus commented: "Deep learning is not perfect, we need to reinvent the line." Hinton corroborated the remark I had told 20 years ago. ”

Earlier, Hinton and his colleagues Sara Sabour, Nicholas Frosst completed the paper "Dynamic Routing Between". They put forward the concept of capsule (capsule), the capsule is Hinton for future artificial intelligence form exploration.
It is expected that Hinton capsule technology can promote the new journey of artificial intelligence in unsupervised learning field.

Contact Us

The content source of this page is from Internet, which doesn't represent Alibaba Cloud's opinion; products and services mentioned on that page don't have any relationship with Alibaba Cloud. If the content of the page makes you feel confusing, please write us an email, we will handle the problem within 5 days after receiving your email.

If you find any instances of plagiarism from the community, please send an email to: info-contact@alibabacloud.com and provide relevant evidence. A staff member will contact you within 5 working days.

A Free Trial That Lets You Build Big!

Start building with 50+ products and up to 12 months usage for Elastic Compute Service

  • Sales Support

    1 on 1 presale consultation

  • After-Sales Support

    24/7 Technical Support 6 Free Tickets per Quarter Faster Response

  • Alibaba Cloud offers highly flexible support services tailored to meet your exact needs.