Fundamental research on artificial intelligence wins Nobel Prize in Physics (2024)
How can we create a smarter, more efficient computer? This is the question many scientists and engineers have tried to answer in recent decades. In the search for an intelligence model, the most familiar and common network system came up as a promising candidate: the human brain. Numerous studies on the brain motivate a practicable approach to creating artificial networks with computer programming.
One of the first programming designs was proposed by John J. Hopfield. The network he designed and published in his seminal paper in 1982, called the Hopfield network, is a fundamental model of the brain that set the basis for future artificial intelligence designs. Here, each neuron stores numbers or information and they are all connected together. The connections between neurons have different strengths and determine how the value from one propagates to another. Following this fundamental research, Geoffrey Hinton presented his work of “convolutional neural networks” (CNN) in 2012. The CNN is a more complex, now widely used, multi-layered neural network with fewer connections between neurons. Because of the fewer connections and multiple layers, it can train images faster and classify test images with high success rates. For contributing these fundamental designs of neural networks, John Hopfield and Geoffrey Hinton won the 2024 Nobel Prize in Physics.
AI research has achieved exceptionally rapid development over the past decade. It has grown into a proficient data-classifying engine that influences daily lives and science. For instance, it is now a part of search engines and suggests customized content everywhere. AI can also classify quantum mechanical properties in materials and identify meaningful data from massive datasets. As manifested through this year’s Nobel Prize, interest and imagination towards AI continues to grow and scientists are motivated to push the boundaries even further.
Managing Correspondent: Seunghyun (Elizabeth) Park
Press Article: Why the Nobel Prize in Physics Went to Al Research (IEEE Spectrum)
Original Journal Articles: ImageNet Classification with Deep Convolutional Neural Networks (Advances in neural information processing systems)
Image Credit: pixabay/GerdAltmann