Theoretical Limitations News Today : Breaking News, Live Updates & Top Stories | Vimarsana

Whose Creators They Failed to Credit

In 1969, Marvin Minsky and Seymour Papert published Perceptrons: An Introduction to Computational Geometry. In it, they showed that a single-layer perceptron cannot compute the XOR function. The main argument relies on linear separability: Perceptrons are linear classifiers, which essentially means drawing a line to separate input that would result in 1 versus 0. You can do it in the OR and AND case, but not XOR. Of course, we’re way past that now, neural networks with one hidden layer can solve that problem.

1990: gradient descent learns subgoals. 1991: multiple time scales and levels of abstraction. 1997: world models learn predictable abstract representations...

Conclusion (~1,700 words). All backed up by over 200 references (~6,500 words). We must stop crediting the wrong people for inventions made by others. Instead let's heed the recent call in the journal Nature: "Let 2020 be the year in which we value those who ensure that science is self-correcting"[SV20]. Like those who know me can testify, finding and citing original sources of scientific and technological innovations is important to me, whether they are mine or other people's [DL1][DL2][HIN][NASC1-9]. The present page is offered as a resource for computer scientists who share this inclin...