In the field of AI, Turing Award winners Yoshua Bengio, Geoffrey Hinton, and Yann LeCun have long been regarded as the pioneers of deep learning. However, they have recently come under scrutiny for issues related to academic integrity. Renowned AI expert Jürgen Schmidhuber has accused the trio in an 88-page report of repeatedly plagiarizing and republishing his and his team’s work without proper citation.

Schmidhuber’s Bombshell Allegations

Schmidhuber, a pioneer and foundational figure in AI, has conducted extensive research across various core areas of artificial intelligence. In his report, he lists 17 specific priority disputes, covering key concepts and methods in deep learning, from Generative Adversarial Networks (GANs) to the vanishing gradient problem. He claims that these award winners not only failed to cite his original work in their initial publications but, more disturbingly, did not correct this oversight in subsequent publications.

One of the most notable controversies in the report concerns Generative Adversarial Networks (GANs). Schmidhuber asserts that the GAN paper published by Bengio and others in 2014 is essentially a simple application of his 1990 artificial curiosity principle, yet it lacks citation of the original work. Even in later publications, they failed to rectify this issue.

Another significant dispute involves the vanishing gradient problem. Schmidhuber states that his student, Sepp Hochreiter, first identified and analyzed this problem in 1991, but Bengio’s related paper in 1994 did not cite…