TL;DR 12Oct24 + podcast ... Google's Demis Hassabis wins a Nobel prize for chemistry ... Google's Geoffry Hinton wins a Nobel prize for physics

Last update: Saturday 10/12/24 

Dr. Demis Hassabis (chemistry)

Dr. Geoffrey Hinton (physics}

Welcome to our 12Oct24 TL;DR summaries by ChatGPT (and the editor of this blog) + podcast of the past week's top 2 stories on our "Useful AI News" page. Last week, our top stories were all about OpenAI. This week our top stories feature Google, more specifically, the Nobel prizes that its current and former employees just received for their AI models. 

As usual, the podcast audio is a recording of the Descript app reading the TL;DR summaries in an AI imitation of the voice of the editor of this blog ... 1) Google's Demis Hassabis won a Nobel prize for chemistry and (2) Google's Geoffry Hinton won a Nobel prize for physics.

Audio podcast ... 6 min 

 

If audio fails to start, or gets stuck, try reloading the page
... Podcast still in progress ..

TL;DR link ➡ HERE

Nowadays the most important AI models are either specialized or generative.  Hassabis' chemistry prize for specialized models was unarguable, indeed, inevitable; but Hinton's physics prize for precursors to his later work on generative models was surprising because it was an illogical ... but politically astute selection by the Nobel committees

A. ChatGPT's TL;DR summaries of Top 2 stories 

1. Hassabis | 2. Hinton 

1) "Google AI scientists win Nobel Prize in chemistry", Jess Weatherbed, The Verge, 9/9/24 *** 

-- This story also covered by Science, Reuters, VentureBeat, NY Times, ... and Nobel Prize + Nobel Prize interview with Hasabis (video)

-- Deep Mind's solution of the protein folding problem was first published in 2019
"One of the Biggest Problems in Biology Has Finally Been Solved", Tanya Lewis, Scientific American, 10/31/22 
-- This story also covered by VentureBeat, Science, 

PROMPT
The Nobel Committee awarded a 2024 prize in chemistry to Demis Hassabis. Below find a copy of the committee's press release. Please summarize the committee's rationale for selecting Hassabis for this award

ChatGPT's RESPONSE

The Nobel Committee awarded the 2024 Chemistry Prize to Demis Hassabis for his groundbreaking work in predicting the three-dimensional structure of proteins using artificial intelligence. Hassabis, co-founder of DeepMind, applied advanced AI models, such as AlphaFold, to solve a problem that had puzzled chemists for over 50 years: how to accurately predict the structure of proteins based solely on their amino acid sequences. His work led to the successful prediction of almost all known protein structures, which is a major scientific breakthrough with vast potential implications for biology, medicine, and chemistry. This achievement has revolutionized how scientists understand and manipulate proteins, moving closer to mastering the chemical tools of life.

Hassabis' motivation

In an interview on Ezra Klein's audio podcast (7/11/23) Hasssabis provided a riveting account of the decades long route he followed in his development of the AlphaFold model.  His burning ambition after completing his undergraduate studies in computer science was to develop AI tools for solving the most complex problems in science. Upon obtaining his doctoral degree, he founded DeepMind (subsequently purchased by Google) where he and his colleagues developed machine learning models that "taught themselves" how to play ever more difficult games, beginning with the simplest Atari video games. Only when the latest models defeated the world's champion in the ancient game of Go, a far more complex game than chess, did he turn his attention to the protein folding problem. 

Other tools for science developed by DeepMind

A partial list of other Ai tools developed by DeepMind to assist scientists in their research

• "DeepMind’s New AI Can Predict Genetic Diseases", Amit Katwala, Wired, 9/18/23 
  -- This story also covered by Wall Street Journal, BBC News, Financial Times, Washington Post, 
  
  
• "Courtesy of AI: Weather forecasts for the hour, the week and the century", Devin Coldewey, TechCrunch, 11/14/23 .
  -- This story also covered by Gizmodo, Financial Times, MIT Tech Review, Ars Technica, Wired, ZDNet, 
  -- These articles describe research published in Nature and Science
  
  
• "AI Tool Pinpoints Genetic Mutations That Cause Disease", Even Callaway & Nature Magazine, Scientific American, 11/21/23 
  -- This story also covered by Wall Street Journal, Wired, 
  
  
• "Google’s DeepMind finds 2.2M crystal structures in materials science win", Michael Peel, Ars Technica, 11/29/23 ... "The trove of theoretically stable but experimentally unrealized combinations identified using an AI tool known as GNoME is more than 45 times larger than the number of such substances unearthed in the history of science, according to a paper published in Nature on Wednesday."
  -- This story also covered by Wired, VentureBeat, Nature, Science, MIT Tech Review,

BackToTop

2) "Geoffrey Hinton, AI pioneer and figurehead of doomerism, wins Nobel Prize" (physics), Will Heaven, MIT Tech Review, 10/8/94 

-- This story also covered by  NY Times, Ars Technica, FastCompany, Washington Post, ... and Nobel Prize
-- Note: Dr. Hinton is a former Google employee

PROMPT

The Nobel Committee awarded its 2024 prize in physics to AI researchers Geoffrey Hinton and John Hopfield. Below find a copy of the committee's press release. Please summarize the committee's rationale for selecting Hinton for this award

ChatGPT's RESPONSE

The Nobel Committee awarded Geoffrey Hinton the 2024 Physics Prize for his pioneering work in artificial neural networks, specifically for developing the Boltzmann machine. Hinton’s contribution was based on applying tools from statistical physics to create a network capable of recognizing patterns and classifying data, such as identifying specific elements in images. His work built upon the foundation of John Hopfield’s network and introduced a new approach that significantly advanced machine learning. Hinton’s innovations have had a profound impact on the explosive growth of AI and its wide-ranging applications, including image recognition and material science.

Objections

Many observers questioned the legitimacy of this award. Here are two examples 

• "The win is already turning heads on social media because it seems unusual that research in a computer science field like machine learning might win a Nobel Prize for physics. 'And the 2024 Nobel Prize in Physics does not go to physics...' tweeted German physicist Sabine Hossenfelder this morning" ... Ars Technica
  
  
• "Professor Dame Wendy Hall, a computer scientist and advisor on AI to the United Nations, told Reuters that, while the recipients’ work deserved recognition, the lack of a Nobel prize for mathematics or computer science had distorted the outcome." ... FastCompany
  

Whereas Hassabis' AlphaFold model used AI procedures to solve a complex problem in biochemistry, Hinton's undeniably important achievements merely borrowed a modeling technique developed by physicists. His achievements revolutionized AI, but they made no contribution to physics. So what's going on here?

As per the comment by Dame Hall, there is no Nobel prize for computer science. That's why the Turing Award was created. But the Turing Award does not have the world-wide renown of the Nobel Prize. According to Hinton and other prominent "doomers", our pursuit of general AI may lead to human extinction. Generative AI has made the most rapid progress towards general AI in the last few years. 

Unfortunately for the Nobel Committees, Demis Hassabis wears two hats. He still leads the Google staffers who developed AlphaFold and other important specialized models ... AND  ... he also leads the Google staffers who are developing GenAI models as fast as possible. Indeed, Hinton resigned from Google last year because he felt that Google was developing GenAi models way, way, way too fast.

• "‘The Godfather of A.I.’ Leaves Google and Warns of Danger Ahead", Cade Metz, NY Times, 5/1/23 
  -- This story also covered by, MIT Tech Review, MIT EmTech interview video 

Perhaps the Nobel Committees were concerned that providing a well deserved prize to Hassabis for his specialized AI work in biochemistry might be misconstrued by large segments of the public as also providing the Foundation's endorsement for his generative AI efforts. Sooooo ... the Committees assumed a politically "neutral" posture by making a dubious award to Hinton, an award whose added prestige he will surely use to convey his existential concerns about generative AI to a much wider audience.

• "This AI Pioneer Thinks AI Is Dumber Than a Cat", Christopher Mims, Wall Street Journal, 10/11/14
  -- Note: Yann LeCun, a leading anti-doomer, is also called a "Godfather of AI" and received the Turing award in 2018 (the highest prize in computer science) together with Hinton for their contributions to deep learning.

BackToTop