This year’s Nobel Prizes in physics and chemistry highlighted the growing importance of artificial intelligence in science and life in general.

AI may have begun with Alan Turing in the 1950s, but the field really took off with the creation of powerful computers, fast chips and machine language, the last of which endowed computers with the ability to find patterns in enormous data sets and even come up with tweaked or entirely novel algorithms to analyze data and solve questions humans struggle to solve.

That was the case with Demis Hassabis, John Jumper and Google’s groundbreaking and highly successful program AlphaFold 2.

As the Nobel committee put it, they “cracked the code” for how linear strings of amino acids arranged themselves into three-dimensional shapes, often bringing amino acids, otherwise distant from one another in the linear chain, much closer together to perform the functions the proteins evolved to accomplish. 

AlphaFold 2 is iterative: The data is run through successive analyses, each successive version an improvement on the last.

The program drew on lessons learned from a database of over 200,000 proteins whose relationships between linear sequences of amino acids and their related three-dimensional shapes had been laboriously determined by other means such as X-ray crystallography and provided clues to nature’s rules that govern the relationships between linear amino acid sequences and the 3D shapes of proteins.

Yet other lessons were garnered from evolutionarily related proteins and at the atomic level, how combinations of specific atoms influence the shapes of combinations of atoms to form molecules. 

In short, Jumper’s team learned from as many sources as possible to find with atomic-level precision in predicting how the order and selection of amino acids determines the 3D shape of complicated proteins. 

AlphaFold 2 completely changed the landscape for determining the shapes of proteins.

More important would be to learn how specific 3D shapes correlate with specific functions — answers that will revolutionize the design of vaccines and other pharmaceuticals.

It is a perfect example of how AI can solve formidable stumbling blocks in biology and science more broadly.

The team’s success reminds us that coding lies at the heart of life itself. After all, where did the code for the sequence of amino acids come from in the first place?

From the order of the four bases in genes that read (transcribed) in the nucleus, creates single-stranded copies of the genes called messenger RNAs (mRNA).

The sequence of bases in the mRNA is then read (translated) in structures called ribosomes in the cytoplasm as successive triplets of bases: Each triplet coding for a specific amino acid to be added to the growing chain of amino acids in the protein.

It could be argued that life is all about coding — DNA to mRNA and mRNA to protein.

With such simple coding systems at the root of life, where did all the complexity come from that differentiates humans from bacteria, tiny nematode worms and house flies, which employ the same atoms, molecules and coding systems? 

For the nervous system at least, the answer lies in the far greater number of nerve cells and their connections in humans, which, like larger and faster computers, are capable of analyzing larger data sets and endow the brain with extraordinary specializations of which speech is a prime example. 

Regions of the neocortex analogous to human areas of the brain involved in speech exist in other primates but don’t possess the requisite complexity and size (translated as computational power) to match human speech areas or areas of the brain involved in music, mathematics and all else that make humans unique among species. 

But there’s a longer story in play.

Life probably began more than 3.7 billion years ago with single strands of RNA capable of generating copies of itself, coding for amino acids and acting as enzymes to foster and direct chemical reactions.

Sometime in the next hundreds of millions of years simple cells probably emerged akin to bacteria and archaea.

But it took as much as two billion years for complex single cells to evolve and only in the last half billion years did multicellular, then complex multicellular organisms evolve including in the last 200,000 years modern humans.

Along the way, there were several major and many more minor extinctions, which killed off millions of species but also created conditions for some species to take hold such as the earliest primates over 50,000 years ago. 

The most important factors governing evolution were lots of time, chance and underlying biological coding and information systems.

Here we are, the cleverest species so far, but at the molecular level, not much different than countless other far simpler species over the last 3.7 billion years. 

On the grand scale of the universe, increasing complexity was a feature of the evolution of the universe as a whole, the birth, life and death of stars, the creation of increasingly complex atoms and much else in the universe.

Here too, in some form, that increasing complexity was based on coding, and information, which if you think about, is what AI is all about — coding and information.

We created AI and now AI helps us understand ourselves, our world and our universe. 

That’s why AI, biology and physics have so much in common.

Dr. William Brown is a professor of neurology at McMaster University and co-founder of the InfoHealth series at the Niagara-on-the-Lake Public Library.