Booth Library hosted a screening of “The Thinking Game” on Tuesday night that highlighted the history behind DeepMind, an artificial intelligence research lab. 

The event was hosted by Kirstin Duffin who represents iSTEM, a mentoring program bringing together students from underrepresented areas of study in the natural sciences, social sciences, technology, engineering, mathematics and computer science fields. 

The screening also featured free popcorn and pizza courtesy of the College of Liberal Arts and Sciences. 

“The Thinking Game” is a 2024 documentary following Demis Hassabis’ life-long research on artificial general intelligence, also known as AGI, to “solve intelligence.” 

What started out as a fascination for the human mind being able to process playing difficult games like chess slowly evolved into incredible leaps in the world of AI. 

Hassabis started out as a professional chess player in Europe before entering a contest to win a job at the game development studio Bullfrog.  

After winning the contest, Hassabis developed the popular game “Theme Park” at the age of 17 before attending and graduating Cambridge University with a double first in computer science. A double first is achieved when a student earns first-class honors in both parts of a subject. 

Following his graduation, development and research into AGI began and the company DeepMind was formed. 

Google took an interest in the company and acquired it in 2014 for $650 million USD, according to The Guardian. 

Come last year, DeepMind was combined with Google AI’s Google Brain, a research team with an exclusive focus on AI and machine learning.

Under Google’s corporate umbrella, DeepMind has achieved outstanding feats such as beating the world’s best go player Ke Jie in 2017. 

The game was going so poorly in Jie’s favor that the Chinese government cut the match’s feed, as seen in the film. 

DeepMind also created an AI called AlphaFold that successfully solved the protein folding problem in 2020. 

Protein folding looks at amino acids to determine what 3D model of protein it could make. Before DeepMind’s AlphaFold AI successfully solved the problem, the accurate creation of these 3D models would take years. 

To achieve these feats, DeepMind utilized reinforcement learning to train the AI. Reinforcement learning is a machine learning technique that helps software, in this case AI, make decisions to reach the best or desired outcome. 

Following the screening, assistant professor of computer science at EIU Md Farhadur Reza talked about the usefulness and dangers of AI. 

“AI is interdisciplinary,” said Reza, referring to the fact that AI requires two areas of knowledge to perform successfully. 

As mentioned in “The Thinking Game,” Reza also highlighted how AI uses the same techniques as people, but its application is what makes it different. 

This sentiment is also highlighted in the film, as Hassabis is portrayed as a figure parallel to J. Robert Oppenheimer’s work on the Manhattan Project and the subsequent results of that introduction. 

During the segment of the film discussing this parallel, Hassabis appeared unconcerned saying, “It’s always easier to land on the Moon if someone’s already landed there.” 

Luke Brewer can be reached at 581-2812 or at lsbrewer@eiu.edu.