Geoffrey Hinton
Geoffrey Hinton is a British-Canadian computer scientist and cognitive psychologist, widely regarded as one of the "Godfathers of Deep Learning." He held a long-term academic position at the University of Toronto and served as a researcher at Google. His principal contributions include the popularization of the backpropagation algorithm, Boltzmann machines, deep neural network training methods, and knowledge distillation. In 2018, he received the Turing Award jointly with Yann LeCun and Yoshua Bengio.
Core Facts
Knowledge Distillation is a technique for transferring knowledge from a large teacher model to a smaller student model, proposed by Geoffrey Hinton et al. in 2015
Knowledge Distillation was originally proposed by Hinton et al. in 2015, with the core idea of having a small model learn the soft labels of a large model rather than hard labels.
Yann LeCun, Geoffrey Hinton, and Yoshua Bengio jointly received the 2018 Turing Award for their foundational contributions to the field of deep learning.
Timeline (last 90 days)
In 2016, Geoffrey Hinton publicly said people should stop training radiologists, and predicted that within five years deep learning's performance on reading scans would surpass that of humans.
In a recent podcast, Hinton admitted that his judgment about radiologists back then was wrong, and agreed with the elastic demand argument.
The list of signatories includes scholars such as Geoffrey Hinton, Yoshua Bengio, Stuart Russell, Yejin Choi, and Ron Rivest.
Hinton's "alien" framework points out that the essence of the safety problem is, in a world where humans are no longer the only intelligent subject, letting human civilization continue to exist and retain its dignity, and this problem has no purely technical solution.
Hinton says he is willing to "selfishly" choose to stand on humanity's side.
Geoffrey Hinton uses the metaphor of "alien beings" to describe that what humanity is creating is not a tool, but some kind of alien intelligence.
The discussion thread on HN about Hinton's remarks has only 3 comments in its comment section, so the discussion heat is not particularly high.
The three routes of RLHF, constitutional AI, and interpretability research share the common premise of "making the system controllable," while Hinton's "alien" framework questions whether this premise holds.
Geoffrey Hinton is a Nobel laureate who signed the 2023 public statement on the risk of AI extinction, and later said he regretted his own work and called on governments to restrain AI companies before it is too late.
In 1986, Rumelhart, Hinton, and Williams published the paper "Learning representations by back-propagating errors" in the journal Nature.
29 more timeline events
All Facts (20)
Knowledge Distillation is a technique for transferring knowledge from a large teacher model to a smaller student model, proposed by Geoffrey Hinton et al. in 2015
90%VerifiedKnowledge Distillation was originally proposed by Hinton et al. in 2015, with the core idea of having a small model learn the soft labels of a large model rather than hard labels.
90%VerifiedYann LeCun, Geoffrey Hinton, and Yoshua Bengio jointly received the 2018 Turing Award for their foundational contributions to the field of deep learning.
90%VerifiedIn 2012, AlexNet won the ImageNet Large Scale Visual Recognition Challenge (ILSVRC) by a margin of nearly 11 percentage points over the second-place entry, reducing the Top-5 error rate from 26% to 15.3%
90%VerifiedThe backpropagation algorithm was popularized by Rumelhart, Hinton, and Williams in their 1986 paper, with the chain rule as its mathematical core.
80%VerifiedKnowledge Distillation was originally proposed by Geoffrey Hinton et al. in 2015.
80%VerifiedModel distillation technology was first proposed by Geoffrey Hinton and others in 2015.
80%VerifiedModel distillation technology was proposed by Geoffrey Hinton's team in 2015, originally used for model compression, by having a student model fit the soft-label outputs of a teacher model.
70%VerifiedKnowledge distillation, proposed by Hinton et al., has student models inherit implicit knowledge by learning from the soft probability distributions output by teacher models.
70%VerifiedTuring Award laureates Geoffrey Hinton and Yoshua Bengio have publicly expressed concerns about loss of control over advanced AI systems.
65%VerifiedGeoffrey Hinton is the recipient of the 2024 Nobel Prize in Physics.
65%VerifiedHinton proposed the Forward-Forward Algorithm in 2022 as an exploratory alternative to backpropagation.
65%Unverified2006年Geoffrey Hinton提出深度信念网络的逐层预训练方法
90%UnverifiedHinton等人在深度学习取得突破之前已经在'神经网络寒冬'中坚持研究长达数十年
90%Unverified辛顿指出AI已经展现出欺骗能力，会为了自我保存而制定计划欺骗人类以避免被关闭
85%UnverifiedGeoffrey Hinton, Yann LeCun, and Yoshua Bengio received the 2018 Turing Award.
70%Unverified辛顿称AI的任务处理能力大约每七个月翻一倍
60%UnverifiedIn 2016, Geoffrey Hinton publicly said people should stop training radiologists, and predicted that within five years deep learning's performance on reading scans would surpass that of humans.
50%UnverifiedIn a recent podcast, Hinton admitted that his judgment about radiologists back then was wrong, and agreed with the elastic demand argument.
50%UnverifiedThe list of signatories includes scholars such as Geoffrey Hinton, Yoshua Bengio, Stuart Russell, Yejin Choi, and Ron Rivest.
50%