During a recent event, Dario Amodei, the CEO and co-founder of Anthropic, shared a thought-provoking insight: his initial expectation that artificial intelligence would bring significant transformations to the biotech field might actually take a decade to materialize.
In a 2024 essay titled “Machines of Loving Grace,” Amodei had proposed that AI, particularly large language models like his company's Claude, could enable researchers to achieve a decade's worth of advancements in just one year—essentially compressing a century's progress into a mere ten years. However, he has now acknowledged that we have not yet reached that point.
“I don’t believe we can achieve a ten-year progress rate in a single year at this moment, for various reasons,” Amodei explained. He highlighted that the current models are not at their full potential, researchers still need time to learn how to effectively use these AI tools, and necessary changes in infrastructure and regulatory frameworks will take time to implement.
This discussion took place at an Anthropic event where the company, established as a public-benefit corporation with a mission to enhance societal welfare, introduced a new tool tailored for biologists and pharmaceutical companies named Claude Science. I had the opportunity to conduct a live interview with Amodei during this gathering, with complete freedom to choose my questions.
