Stay informed with free updates
Simply sign up for the Artificial Intelligence myFT Digest, delivered straight to your inbox.
“While (Sam) Altman measured the success of his investments and the people using his products in numbers, (Demis) Hassabis chased awards,” says Palmy Olson in her book about the co-founders of DeepMind and OpenAI. Supremacy”. “[Hassabis]often told his staff that he wanted DeepMind to win three to five Nobel Prizes over the next 10 years.”
This week, just hours after Olsson won the 2024 Financial Times-Schroders Award for Excellence in Business Books, Hassabis won the Nobel Prize in Chemistry in Stockholm for his work on AI systems that can predict structures. Made the first edition obsolete. of all known proteins.
That Hassabis is already one step closer to his ambitious goal of winning a Nobel Prize shows how quickly AI is transforming the world. Olson’s difficult task was to publish a book about this rapidly advancing technology that would stand the test of time. In an interview this week, she said she wanted to depict “the battle for control[of the AI]and…” . . For proper supervision.” Rula Khalaf, an FT editor who chairs the book award’s jury, said that Olson was a partner in the “development of artificial intelligence” between Hassabis, who has a spiritual mindset and is obsessed with games, and Altman, who is obsessed with numbers. It perfectly depicts the race as a thrilling race.”
Despite the differences between Hassabis and Altman, Olson said she was also intrigued by their similarities. Both men believed that AGI (the point at which AI exceeds human cognitive ability) “will solve many of our current social ills and problems.” Both men shared concerns about the lack of regulation of AI and over-control by companies. “Both companies tried to separate the technology a little bit and put governance structures in place to give them proper oversight, but neither was able to do that,” she says.
While Google currently manages DeepMind, Microsoft backs OpenAI, which launched ChatGPT two years ago to accelerate the use of generative AI. The founders’ “very utopian, almost humanitarian ideals kind of faded into the background as they became more aligned with two very large technology companies,” Olson said. She intended Supremacy in part to be a warning about the need for proper regulation of emerging technology oligopolies.
But given the rapid evolution of AI, is it too late to impose regulatory and ethical constraints on it? Olson, the Bloomberg technology columnist, thinks otherwise. She says there’s still time to influence “how technology companies design their algorithms” to make them safer and less biased.
Legislation like the EU’s AI law, which sets out a strict regulatory regime, will impose guardrails. But Olson also points out that companies that procure generative AI systems will put restraints on technology suppliers. “There’s a lot of experimentation going on, but it’s not really costing a lot of money to put these AI systems into practical use because there are concerns about hallucinations… and bias,” she said. Masu. “Companies like banks and health systems have their own regulatory regimes that they need to ensure they comply with before offering AI-powered products and services to their customers.”
When this year’s Book Prize launched, past winners were asked what they would add to their books if given the chance to write a new chapter. Although Supremacy was only published in September, Olson said future editions will recognize the importance of last month’s election of President Donald Trump and the proximity of entrepreneur Elon Musk to the next president of the United States. I recognize that it may be necessary.
Musk, along with Altman, was a co-founder and early backer of OpenAI, before splitting off to form his own startup, xAI, which also cultivated a rival to Gemini called ChatGPT and Google’s Grok. “I’m honestly surprised by the rapid growth in Grok’s ability to raise capital,” Olson said. But Musk’s presence in the Trump administration could “throw a spanner in the works” instead of accelerating a light regulatory regime for AI, she warned. “You have to remember that Musk is the destroyer of AI. He started OpenAI in part because he was so worried about Google controlling AGI.”
Olson’s concerns about the weaponization of AI for purely commercial gain are clear. Some AI chatbot startups are already facilitating emotional connections between users and bots. She is concerned about advertising-driven models that could encourage an addictive cycle of chatbot use. “Vendors will find it difficult to think about user well-being because their business model relies on users using their apps for as long as possible.”
But Olson says he’s not predicting an AI dystopia. When the FT prompted ChatGPT to generate an “unexpected question” for her, the bot asked: “If AI were to write the definitive history of humanity 100 years from now, what perspectives and biases would it have? And how would this history differ from the history written by humans today? ”
“If humans really try,” Olson said, “if humans really try,” and a more just society evolves, “anything that AI writes could be possible. . . . It will be a reflection of that.” I’m optimistic and I hope it’s actually a very exciting read.”
