Who Won the Nobel Prize in 2024, Besides Daron Acemoğlu?
This week, much has been written and discussed about the Nobel Prize awarded to our esteemed professor Daron Acemoğlu and the role of institutions in development. If you allow me, I’d like to share a few thoughts on Demis Hassabis, one of this year’s Nobel Prize in Chemistry recipients. Hassabis, John Jumper, and David Baker were recognized for contributing to our understanding of proteins. Hassabis is one of the world’s leading AI pioneers. In 2010, he founded DeepMind, the world’s most significant AI company, long before OpenAI, the creator of ChatGPT, existed. Let’s take a closer look at who Demis Hassabis is and the institutional structure that led him to the Nobel Prize.
Hassabis was born in London in 1976 to a Cypriot Greek father. He studied at Cambridge University in England. When it was announced that Daron Acemoğlu had won this year’s Nobel, I was at an event in Istanbul with my dear friend Vuk Jeremić, the former foreign minister of Serbia. As everyone praised Acemoğlu’s intelligence and work ethic, Vuk said, “Yes, I also went to Cambridge to become a scientist, but my dorm roommate was Demis Hassabis. After getting to know him, I realized I would never become a scientist and switched to politics.” Hassabis originally wanted to develop computer games that simulated life. However, his games were so technically advanced that many failed commercially. As we know, both games and politics need to appeal to a broad audience, and sometimes, being too bright can become a disadvantage.
When Hassabis was unsuccessful in the gaming industry, he founded DeepMind in 2010, intending to develop an AI system modeled on the human brain. Two essential inputs are needed to develop practical AI systems: human expertise and computing power. Both are expensive. Initially, DeepMind financed its operations through investments from Silicon Valley billionaires like Elon Musk and Peter Thiel, whom I’ve written about before. However, as the personal investments proved insufficient, a new path had to be found.
In 2014, Google acquired DeepMind, allowing its employees to earn competitive salaries and access Google’s unlimited processing power. Hassabis had one condition for sale: that DeepMind maintains its independence within Google and that an “ethics board” be established to oversee the direction of AI development. As we know, large corporations often make such promises but rarely follow through. Hassabis’ ethics board only met once, and ironically, one of its members was Elon Musk. Nowadays, Musk’s involvement in an “ethics board” seems somewhat ironic, doesn’t it?
In its early years under Google, DeepMind enjoyed some autonomy and produced groundbreaking work. For example, in 2016, DeepMind developed AlphaGo, the AI that defeated the Chinese Go champion, triggering an obsession with AI within the Chinese Communist Party. However, being part of Google came with its challenges. When DeepMind started working with data from the UK’s National Health Service, the public protested, claiming, “Our citizens’ data is being sold to Google!” As a result, these projects were shelved. Google told DeepMind, “Since 90% of our revenue comes from advertising, you might as well focus on developing AI for the ad business.” At the same time, Hassabis was working on a legal framework to separate DeepMind from Google; Larry Page, the Google co-founder who made critical promises during the acquisition, retired. Sundar Pichai, the new CEO, merged DeepMind with the newly formed Google Brain. Hassabis completed his Nobel-winning work on protein structures in 2020 while still under Google’s umbrella.
While DeepMind continued its academic research at its London office and contributed to Google’s product development, another AI company emerged in Silicon Valley in 2016: OpenAI, which we now know as ChatGPT. When it was founded, OpenAI was a non-profit organization. However, as funds needed to cover salaries and processors grew, the company created a for-profit entity within the non-profit structure and sold shares to investors. Microsoft eventually purchased 49% of this for-profit venture. By the end of 2022, when the board expressed discomfort with this setup, it was dismissed, leaving Sam Altman as the sole leader. Last month, Altman closed a new funding round, valuing OpenAI at $157 billion, and announced that the company would now fully operate as a for-profit entity.
Money changes everything. Like in many other fields, achieving your ideals in AI requires substantial funding, and to access that capital, you often must adjust your ideals. Once you acquire significant wealth, you tend to change as well. One of Daron Acemoğlu’s most critical recent ideas is that “the institutional structure determines the trajectory of technology.” In other words, today’s institutional framework will shape whether AI will further enrich large corporations or help spread its benefits broadly across society. The stories behind this year’s two Nobel Prizes suggest that, for now, the first option seems to be winning out.
Book Recommendation: Supremacy: AI, ChatGPT, and the Race that Will Change the World by Parmy Olson.
This article is a translated version of “2024’te Daron Acemoğlu dışında kim Nobel aldı” which was initially published in Economic Daily (Nasıl Bir Ekonomi Gazetesi) on October 18, 2024.
