A brave new world – if we don’t blow it up first

Discover the journey from 'the last man who knew everything' to the era of Artificial Intelligence and AGI research.

We sometimes speak of ‘the last man who knew everything’. It is a Western conceit, for it is unlikely that Samuel Taylor Coleridge, often anointed thus, knew of books or events beyond his hemisphere. Aristotle, Leonardo da Vinci and Goethe have also been similarly cited.

But even in Aristotle’s time (384-322 BC), it would have been difficult for one man to know everything. Two millennia later, it’s impossible for human intelligence to gather, store, find patterns in, and develop insights into the huge amount of data in the world. For that, we need Artificial Intelligence.

We turn thus to ‘the cleverest man on earth’. In recent years, this has been the British scientist Demis Hassabis, 48, who won the Nobel Prize for chemistry this week.

Hassabis wrote his own computer games at the age of eight, was captain of Britain’s under-11 chess team at nine and became the No. 2 ranked 13-year-old player in the world; he created Theme Park, one of the first video games to incorporate AI at 17; had a double first in computer science from Cambridge at 20, did pioneering academic work in neuroscience before founding DeepMind, a company Google bought for 625 million dollars. Now CEO of Google DeepMind, he is at the forefront of AI and AGI (Artificial General Intelligence) research.

In 2016, Hassabis laid out his ambition: to solve intelligence and then use that to solve everything else. This includes energy, climate crisis, financial systems, medical issues and drug design for intractable conditions like Alzheimer’s and Parkinson’s.

AI has been the Nobel theme this year. The physics prize was awarded to John Hopfield and Geoffrey Hinton for establishing the field of machine learning. Chemistry went to Hassabis, John Jumper and David Baker for their work on AlphaFold, an AI system that predicts the 3D structure of proteins from their amino acid sequences. The database is available free.

Proteins have complex 3D structures. Figuring out just one can take several years and millions of dollars. In 2020, AlphaFold began to predict protein structures in minutes.