Inside the (Deep) Mind of Demis Hassabis

He wrote computer games at 8; became a chess prodigy at 14; created Theme Park, one of the first video games to incorporate AI, at 17; took a double first in computer science from Cambridge at 20; founded a groundbreaking video-game company at 22. Who, but Demis Hassabis, could have achieved all of that? Of course, the unstoppable genius went on to create the neuroscience-inspired AI firm, DeepMind in 2010.

Hassabis compares the AI research company to the Apollo programme and Manhattan Project for both the breathtaking scale of its ambition and the quality of the minds he is assembling at an ever-increasing rate.  “Although we work on making machines smart, we wanted to keep humanity at the centre of what we’re doing here,” the polymath told TIME in a rare interview. Hassabis referred to the DeepMind building as a “cathedral to knowledge”, which he sees as an ode to intelligence. 

Since the company was put in place, in a little over a decade, DeepMind has made significant strides by developing algorithms that have outmatched the world’s top players in complex games like chess and Go. In 2021, the company achieved yet another groundbreaking accomplishment by solving the complex puzzle of protein folding with AlphaFold, a remarkable feat that has stunned the scientific community. 

He has been described as a “visionary” by those who work alongside him and Tim Berners-Lee, the inventor of the World Wide Web, once rightly called him “one of the smartest human beings on the planet”.  

Nicknamed the “superhero of artificial intelligence“, Hassabis learned how to programme on his ZX Spectrum 48k, which he bought from some of his chess match winnings and taught it to play the board game Othello. 


Download our Mobile App


(Source: britishchessnews.com/ )

Growing up in north London, the child of a Greek Cypriot father and a Chinese Singaporean mother, he is the eldest of three siblings and also the only one to not take an artistic route in the family. The 46-year-old has always been obsessed with the idea of intelligence and everything around it. As an adult, he strikes a somewhat diminutive figure, a Guardian journalist wrote describing him, but his intellectual presence fills the room. 

He is a man with a mission: to explore the big questions that have captivated philosophers and physicists for centuries and sees AI as the key to unlocking their secrets. “I thought building AI would be the fastest route to answer some of those questions,” he remarked. 

The Smartest Kid Next Door

From a young age, Hassabis realised that in some fundamental sense there were only two subjects really worth studying – physics and neuroscience. 

After graduating from Cambridge, Hassabis returned to Bullfrog to help Peter Molyneux, a video game designer, build his most-popular game to date: Theme Park, a simulation game giving the player a God’s-eye view of an expanding fairground business. A bestseller, it won the industry’s Golden Joystick Award and spawned a host of management simulation games.

In 1998, Hassabis established Elixir Studios, his game company before and then changed his mind to study for a PhD in neuroscience. For him, “the mind is the most intriguing object in the universe”, and his quest to understand its workings ultimately led him to the field of artificial intelligence. Reflecting on his unusual career path, he has acknowledged that his journey may seem haphazard to an outsider, but he insists that he “used every single scrap of that experience” to get where he is. 

(Source: achievement.org/ )

Hassabis is fond of the strategy game called Polytopia. The objective of the strategy game is to grow a small village into a dominating empire through technological advances; which resonates with Hassabis. Drawing parallels in the real world, he believes that technological advancements will be the driving force for human civilization despite its risks.   

But unlike the game, he recognises the complexity of the real world and the burden of capitalism. 

In a world of “radical abundance”, he believes that there should be no room for inequality. He is optimistic that the wealth from AGI could be harnessed for the greater good, and that a brighter, more equal future was possible. He advocates the ideas of universal basic income and services, where the government pays for basic living standards.  

Front-Row Seat in AI

At the recent NVIDIA GTC, Hassabis said, “DeepMind has been lucky enough to be at the front-row seat for what’s been happening”. Since the company’s outset in 2010, no doubt it has managed to scale above everyone’s expectations. In 2016, the AlphaGo program defeated the legendary Go champion Lee Sedol in a match in Seoul. 

Two years later, the company aced protein folding with AlphaFold and in 2021 they put the power in the hands of people. AlphaFold quickly became an essential tool for the scientific community and has helped overcome severe challenges in society like developing a malaria vaccine, fighting diseases before the onset and so on.

Last year, the company built AlphaTensor, an AI system (based on a 3D board game) researchers shed light on a 50-year-old fundamental mathematics question of finding the fastest way to multiply two matrices.

The team also collaborated with physicists from the Swiss Plasma Centre at EPFL in Ecublens, Switzerland, to develop an AI method to control the plasmas inside a nuclear fusion reactor. The study helps further nuclear fusion research and could also help quicken the arrival of a cheaper, cleaner, and unlimited source of energy. 

A billion years of PhD time would have taken to do all of this experimentally, Hassabis said, talking about what Google’s research spine has managed to achieve computationally in less than a year. 

Read the 4 AI Questions That DeepMind Answered in 2022. 

A Cautionary Voice

Beneath the optimism, Hassabis is concerned about the trajectory of the technology. He has warned that the very tools that AI is capable of creating could pose a serious threat to human civilization, if not handled carefully.

At several instances, he has urged DeepMind’s competitors to take a step back and consider the potential risks before moving forward. In a nod to the culture of “move fast and break things”, Hassabis advocates for a more measured approach to the development of AI. He recognizes that the stakes are high and that society as a whole is essentially acting as guinea pigs, testing the boundaries of what is possible with this powerful technology.

As the world hurtles towards an uncertain future, Hassabis’ cautionary voice serves as a reminder that the risks of AI must be taken seriously.


Read more here: Source link