Demis Hassabis, co-founder of DeepMind and current president of Google DeepMind, warned during an interview with Reid Hoffman on the 'Masters of Scale' podcast about the risks associated with recursive self-improvement in AI systems, which he refers to as potentially leading to uncontrollable outcomes. He emphasized the need for human oversight in AI development and highlighted the importance of separating intelligence from consciousness. Hassabis discussed his vision for using AI as a transformative tool for science and healthcare, citing AlphaFold as a major success in applying AI to complex scientific problems. He also mentioned future goals such as expanding AI applications into biochemistry to predict drug safety and efficacy before clinical trials.
Bias read (Center): The article presents a balanced overview of Hassabis’s views on AI without overtly favoring any perspective, focusing on technical discussions rather than political implications.





![US Minister under pressure: Nvidia affair reaches the US government [premium]](https://images.weserv.nl/?url=image.kurier.at%2Fimages%2Foriginal%2F9377758%2F46-235690262.jpg%3FversionId%3D2a285eac5b32&w=3840&q=75&output=webp&we)
