Hassabis at the forefront of the AI race
A chess prodigy turned computer theorist, a designer of computer games, a neuroscientist, and one of the world’s leading AI authorities, Demis Hassabis has become a central figure at the Mobile World Congress. Google DeepMind, the AI lab he co-founded and now leads, stands as one of the two major players driving the rapid competition for leadership in this emerging technology, alongside OpenAI. OpenAI, backed by Microsoft and creator of ChatGPT, has popularized AI on a massive scale. Hassabis notes that much of the sector’s current momentum comes from breakthroughs that originated inside Google, including the transformer architecture that powers ChatGPT, first proposed in 2017 by Google researchers.
For years Google quietly advanced its AI work, delaying public releases due to earlier missteps. That dynamic shifted in late 2022 when OpenAI opened access to ChatGPT, catapulting the chatbot to historic success and placing Sam Altman’s startup at the vanguard of the race. Since then, Google has introduced Gemini, a multimodal AI capable of generating text, images, audio, and video in response to user prompts. Just recently the company paused a graphic feature after the system produced historically inaccurate images, such as racially charged Nazi figures. Hassabis stated that the system was not performing as intended.
Scientific milestones
The first day of the conference offered a review of the career of the founder of Google DeepMind. In the bustling venue of Gran Via at Fira de Barcelona, Hassabis described AI as the mission of his life. Known early as a chess prodigy, he redirected that passion into exploring the frontiers of intelligence. His work on neural networks drew the attention of Alphabet, Google’s parent company, which bought his startup in 2014 for roughly half a billion dollars.
Two years later, the DeepMind name made headlines worldwide when AlphaGo defeated the world champion in Go, a highly complex board game that relies on strategic calculation. The moment echoed IBM’s Deep Blue victory over Garry Kasparov in chess in 1997. “That breakthrough likely marked the start of the current AI boom,” Hassabis remarked.
Yet Google DeepMind’s achievements extend far beyond that historic moment. Since 2018, the team has developed AlphaFold, an AI system capable of predicting the three‑dimensional structure of proteins with remarkable accuracy. These discoveries have accelerated scientific research and deepened the understanding of human biology. “Knowing the structure of a protein lets us design drugs that precisely target the affected area, making therapies more efficient,” Hassabis noted. “I hope AlphaFold can shorten drug discovery from ten years to a matter of months.”
AlphaFold’s influence is broad. It is already used to predict how proteins fold, to forecast how biological systems respond to changes, and to tackle difficult mathematical problems. More than a million researchers have reportedly engaged with the tool, reflecting a shift in how science is conducted today.