The Royal Swedish Academy of Sciences selections for the Nobel Prizes in Physics and Chemistry have left many scratching their heads. Many in their respective fields are quite upset.
It’s not that the winners’ accomplishments were not impressive. They were incredible.
It’s just that they were awarded prizes in fields in which they had no background.
Scientists spend decades on research in their fields of physics and chemistry, and this week “outsiders” took home the prize.
No wonder they’re so upset.
From where did these interlopers come from? How did they pull off such a coup?
I doubt it will surprise anyone that they come from the world of computer science, specifically experts in deep learning – a form of artificial intelligence (AI).
The co-founder and CEO of DeepMind, Demis Hassabis, has been awarded the Nobel Prize in chemistry – a field in which he has no background.
Google acquired DeepMind in 2014 for $650 million. This was a large sum at the time for an AI research organization that was bleeding cash. It’s hard to believe that was a decade ago when very few were thinking about artificial intelligence.
It turned out to be an unbelievably smart acquisition as the team at DeepMind has produced a stream of breakthroughs in deep learning technology that is unmatched, specifically in developing neural networks to achieve impressive tasks.
Initially, it didn’t seem that much of the industry was taking DeepMind’s work that seriously. It focused its efforts on developing neural networks to play and master video games and board games.
The team started small with games like Breakout, Pong, and Space Invaders, and simpler board games like chess and shogi.
But those were just the training grounds… low-hanging fruit to prepare for far more complex tasks.
By 2016, DeepMind released AlphaGo, which decisively beat the world champion of a very complex game – Go – that has more possible positions in a single match than there are atoms in the universe.
AlphaGo demonstrated the remarkable capabilities of a deep neural network to recognize patterns and to “think” to win.
But even that was just scratching the surface.
By 2020, the team at DeepMind focused on one of the grand challenges of life sciences – figuring out how proteins fold.
DeepMind released AlphaFold. Though very few understood the significance at the time, this was a remarkable achievement that has already revolutionized the field of drug discovery and changed the world in a single step.
In the summer of 2022, AlphaFold 2 was released with mindboggling capabilities, specifically the accurate predictions of how more than 200 million proteins fold. This represents nearly every protein known to science – from around 1 million different species.
Understanding a protein’s structure is critical to the drug discovery process. It allows scientists to determine protein interactions within the human body, as well as how a protein might interact with molecular compounds.
If that weren’t enough, earlier this year DeepMind released AlphaFold 3, which can accurately predict the structure of all known proteins, DNA, RNA, and ligands.
It’s like having the key to all biology.
The significance of AlphaFold 3 still hasn’t been felt as the biotech industry has only just recently been empowered with these incredible tools, but the implications for what this means to the scientific community are immense.
Given DeepMind’s extraordinary accomplishments, Hassabis was promoted to oversee the merged Deep Mind and Google Brain teams into one Google AI research organization.
What’s so remarkable is that Hassabis has no background in life sciences, computational biology, or chemistry. And yet, he now has a Nobel Prize.
Geoffrey Hinton, another unexpected Nobel Prize winner, was “flabbergasted” by the announcement that he had won the Nobel Prize in Physics.
Hinton, like Hassabis, is a computer scientist who specializes in deep learning. He is one of the three godfathers of AI and was also awarded the 2018 Turing Award for his contributions to breakthroughs in artificial intelligence.
Named for the mathematician and pioneer of modern computer science Alan Turing, the Turing Award has the significance of something like a Nobel Prize for computer science.
With that said, like Hassabis, Hinton has no background in physics, which is why he was so surprised by the news.
Hinton even said that he “dropped out of physics after my first year in university because I couldn’t do the complicated math.”
Hinton developed the Boltzmann Machine deep learning model. It’s a kind of neural network capable of both classifying images and generating new images or data based on the information it was trained on.
Source: The Royal Swedish Academy of Sciences
That might sound familiar, something along the lines of what we now know as generative AI. Hinton’s early work was just the beginning of an explosion of machine learning and eventually generative AI technology.
The chair of the Nobel Committee for Physics made a telling comment with the announcement…
The laureates’ work has already been of the greatest benefit. In physics, we use artificial neural networks in a vast range of areas, such as developing new materials with specific properties.
Hinton’s contributions to artificial intelligence – his work in the development of machine learning in particular – have literally changed the landscape of research in physics, and many other fields for that matter.
It’s impossible to overstate the significance of his research to the scientific community. And that’s why he’s being honored with this Nobel Prize…
We often hear comments along the lines of, “Sure, AI is developing quickly and ChatGPT is really neat, but it’ll be a long time before people start using the technology mainstream.
Of course, this is a complete misunderstanding. It’s nonsense.
Artificial intelligence is widely used in chemistry, life sciences, drug discovery, autonomous driving, robotics, supply chain management, and a nearly endless number of applications.
And amazingly, it’s still very early in terms of adoption.
This is precisely why there is such a rush to build out so many AI-specific data centers. They are needed to manage the exploding demand for computer processing for both AI training and inference.
And of course, they are necessary to achieve artificial general intelligence (AGI), which will enable AI to be used for autonomous self-directed research. The “self” being the AI.
One of the single most important pieces of advice that I give to any student in high school or college, or someone early on in their career, is to double major in computer science or learn it through postgraduate studies.
AI plus your field of choice is the future. AI will be leveraged in all fields.
And these latest developments do raise a peculiar question. If Hassabis can win a Nobel in Chemistry with no background in chemistry or life sciences, and Hinton can win a Nobel in Physics with no background in physics…
…can an AGI capable of self-directed research win a Nobel Prize?
The Bleeding Edge is the only free newsletter that delivers daily insights and information from the high-tech world as well as topics and trends relevant to investments.
The Bleeding Edge is the only free newsletter that delivers daily insights and information from the high-tech world as well as topics and trends relevant to investments.
