The 20 Most Influential Papers in AI


Each year, thousands of research papers are published in the field of artificial intelligence, machine learning, and computer vision. But only a handfull have made transformed the field of AI is significant ways. Below is a curated list of the 20 most influential papers in the field of AI in chronological order.

1. Computing Machinery and Intelligence

Alan Turing - 1950


Why it is influential:


This paper introduced the concept of the Turing test, which is still widely used as a benchmark for evaluating the intelligence of AI systems.

2. Perceptrons

Marvin Minsky and Seymour Papert - 1969


Why it is influential:


This paper analyzed the limitations of the perceptron algorithm for pattern recognition and helped motivate the development of more powerful neural network models.

3. The Unreasonable Effectiveness of Mathematics in the Natural Sciences

Eugene Wigner - 1960


Why it is influential:


This paper explored the idea that mathematics is an incredibly powerful tool for understanding the natural world and laid the foundation for the use of mathematical methods in AI

4. A Theory of the Learnable

Leslie Valiant - 1984


Why it is influential:


This paper introduced the computational complexity theory of learning, providing a mathematical foundation for the analysis of machine learning algorithms.

5. A Connectionist Model of Visual Processing

David Rumelhart, James McClelland, and the PDP Research Group - 1986


Why it is influential:


This book introduced the parallel distributed processing (PDP) framework for neural networks and demonstrated its ability to learn visual recognition tasks.

6. Backpropagation Applied to Handwritten Zip Code Recognition

Yann LeCun, Bernhard Boser, John S. Denker, Donnie Henderson, Richard E. Howard, Wayne Hubbard, and Lawrence D. Jackel - 1989


Why it is influential:


This paper introduced the use of backpropagation, a widely used neural network training algorithm, for handwritten digit recognition.

7. Support-vector networks

Corinna Cortes and Vladimir Vapnik - 1995


Why it is influential:


This paper introduced the support vector machine (SVM) algorithm, which is a powerful method for classification and regression tasks.

8. A Few Useful Things to Know About Machine Learning

Pedro Domingos - 1997


Why it is influential:


This paper provides an accessible introduction to the basic principles of machine learning and is often recommended as a starting point for newcomers to the field.

9. A Fast Learning Algorithm for Deep Belief Nets

Geoffrey Hinton, Simon Osindero, and Yee-Whye Teh - 2006


Why it is influential:


This paper introduced the deep belief network (DBN) model, which is a type of neural network that is capable of unsupervised learning. This paper was influential in the development of deep learning.

10. Deep Sparse Rectifier Neural Networks

Xavier Glorot, Antoine Bordes, and Yoshua Bengio - 2011


Why it is influential:


This paper introduced the rectified linear unit (ReLU) activation function, which is now widely used in deep neural networks.

11. ImageNet Classification with Deep Convolutional Neural Networks

Alex Krizhevsky, Ilya Sutskever, and Geoffrey Hinton - 2012


Why it is influential:


This paper introduced the AlexNet architecture, which was the first deep convolutional neural network to achieve state-of-the-art performance on the ImageNet image recognition benchmark.

12. Playing Atari with Deep Reinforcement Learning

Volodymyr Mnih et al. - 2013


Why it is influential:


This paper introduced a breakthrough in deep reinforcement learning by showing that a neural network could learn to play Atari games at a superhuman level using only raw pixel data as input.

13. Deep Residual Learning for Image Recognition

Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun - 2016


Why it is influential:


This paper introduced residual neural networks, which have become the state-of-the-art architecture for many computer vision tasks.

14. BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding

Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova - 2018


Why it is influential:


This paper introduced the BERT model, a pre-trained language model that has achieved state-of-the-art performance on many natural language processing tasks.

15. Generative Pretraining from Pixels

Alec Radford, Luke Metz, and Soumith Chintala - 2016


Why it is influential:


This paper introduced the Deep Convolutional Generative Adversarial Network (DCGAN), a generative model that can generate realistic images.

16. AlphaGo Zero

David Silver, Julian Schrittwieser, Karen Simonyan, Ioannis Antonoglou, Aja Huang, Arthur Guez, Thomas Hubert, Lucas Baker, Matthew Lai, Adrian Bolton, Yutian Chen, Timothy Lillicrap, Fan Hui, Laurent Sifre, George van den Driessche, Thore Graepel, and Demis Hassabis - 2017


Why it is influential:


This paper introduced AlphaGo Zero, an AI system that can learn to play the game of Go from scratch, achieving superhuman performance without human input.

17. Attention Is All You Need

Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, Łukasz Kaiser, and Illia Polosukhin - 2017


Why it is influential:


This paper introduced the Transformer model, a neural network architecture that can process sequences of variable length with parallel computation, which has been used for a variety of natural language processing tasks.

18. Neural Ordinary Differential Equations

Ricky T. Q. Chen, Yulia Rubanova, Jesse Bettencourt, and David K. Duvenaud - 2018


Why it is influential:


This paper introduced the Neural Ordinary Differential Equation (ODE) model, a continuous-time model that can learn the dynamics of a system from its observations, which has applications in physics, control, and other domains.

19. Unsupervised Learning of Visual Features by Contrasting Cluster Assignments

Mathilde Caron, Piotr Bojanowski, Armand Joulin, and Matthijs Douze - 2018


Why it is influential:


This paper introduced the SimCLR framework for unsupervised learning of visual representations, which has achieved state-of-the-art performance on several image classification tasks.

20. GPT-2: Language Models are Unsupervised Multitask Learners

Alec Radford, Jeffrey Wu, Rewon Child, David Luan, Dario Amodei, and Ilya Sutskever - 2019


Why it is influential:


This paper introduced the GPT-2 language model, which can generate coherent text in a variety of styles and topics, and has been used for a range of natural language processing tasks.