Godfather of AI Geoffrey Hinton Gets Physics Nobel Prize as AI Research Comes into Spotlight

Read Time: 2 minutes

Geoffrey Hinton and John Hopfield, two scientists, have won the Nobel Prize in Physics for their work on machine learning. The British-Canadian professor Hinton, who has been called the "Godfather of AI" at times, expressed his shock. After leaving Google in 2023, he issued a warning about the risks posed by artificial intelligence systems that could outsmart people.

The committee in Sweden revealed the renowned award, which is regarded as the highest honor in science. A cash award of 11 million Swedish kronor $1 million is linked with the prize. Hinton, a computer scientist at the University of Toronto and a professor at Princeton University was recognized for creating.

The groundwork for machine learning supports many of the AI-based goods and services available today. Hinton, meanwhile, has also raised bookings about AI’s future development and severed ties with his old company, Google, to speak more freely on the matter.

Huge Influence of AI

When asked about the possible significance of the technology his study has helped to produce, he claimed artificial intelligence (AI) would have a "huge influence" on our civilizations. In certain ways, it would be like the Industrial Revolution.

However, it will surpass humanity in intellectual capacity rather than physical power. In a phone interview following the news, he stated, "We have no idea what it's like to have things smarter than us." According to Hinton, technology would completely transform industries like healthcare and result in a "huge improvement in productivity."

Mimicking the Brain

Artificial Neural Network ANN machine learning has become abbreviated as AI. This technique is based on the architecture of the brain and was created by Hopfield and Hinton. In contrast to neurons found in the brain, nodes with varying values are found in artificial neural networks.

Artificial nodes impact one another through connections, while neurons in the brain communicate with one another through synapses. The scale of the networks has increased since the 1980s. Today's networks, like the ones that power Chat GPT, can have more than one trillion parameters, in contrast to Hopfield's network of just 30 nodes and fewer than 500 parameters connecting them.

Whistleblower

Hinton was not only a pioneer in AI but also advocated for caution with the technology. He chose to "blow the whistle" and quit Google in May 2023 because he was concerned about the company's increasing level of intelligence.

There are two types of regrets, according to Hinton: those that come from knowing you shouldn't have done something and those that come from doing something you would do again under the same circumstances but which might not work out in the end.

Previous winners of the Nobel Prize in Physics

  • In 2022: Alain Aspect, Austrian Anton Zeilinger, and American John Clauser for studies in quantum mechanics, the field that studies nature at the smallest scales.
  • In 2023: Pierre Agostini, Ferenc Krausz, and Anne L'Huillier for their research on attoseconds, which are incredibly brief light pulses that may be utilized to record and examine fast atomic processes.
  • In 2021: Giorgio Parisi, Klaus Hassel Mann, and Syukuro Manabe received awards for improving our knowledge of complex systems like the Earth's climate.
  • In 2020: Andrea Ghez, Reinhard Genzel, and Sir Roger Penrose were awarded the prize for their research on the properties of black holes.
  • In 2019: Didier Queluz, Michel Mayor, and James Peebles were awarded a prize for their groundbreaking discoveries about the universe.

Neural Networks Are Computer Technologies. How Does This Relate To Physics?

Physics underlay is for both Hopfield networks and Boltzmann machines, which is an extension of them. Hopfield nets used an energy function, and the Boltzmann machine employed statistical physics concepts.

Thus, a lot of the advancement of neural networks at that point was dependent on concepts from physics. However, the modern A.I. models were constructed using a different method known as backpropagation. That isn't as related to physics.