OpenAI is preparing to release an open-weight AI model that will intentionally be one generation behind the frontier models to align with democratic values and to avoid accelerating competitors like China, according to OpenAI Chief Product Officer Kevin Weil. Weil also predicted that soon computers will surpass humans in coding ability, marking a turning point in AI capabilities. Meanwhile, Anthropic CEO Dario Amodei has openly acknowledged the limited understanding of how AI systems function, describing this lack of insight as unprecedented in technological history. Amodei is advocating for the development of an "MRI for AI" by 2027, a tool intended to demystify the inner workings of AI models and enhance transparency. Industry experts, including GV Managing Partner Tom Hulme, recognize OpenAI, Google, and Anthropic as the primary leaders in foundation AI models, with other companies expected to face challenges in this space. OpenAI CEO Sam Altman has emphasized the goal of building a "brain for the world," highlighting the ambition behind these AI advancements.
Anthropic CEO admits that we have no idea how AI works "This lack of understanding” "is essentially unprecedented in the history of technology" - Dario Amodei https://t.co/4gZBespqz1
Anthropic CEO Dario Amodei just dropped a bombshell: even the leading minds don’t fully understand how AI works. In his recent essay, he explained that a detailed 'MRI on AI' is in the works to finally uncover the inner workings of these mysterious systems. The idea is to https://t.co/kKSYy5Hifx
Anthropic CEO wants to open the black box of AI models by 2027 | TechCrunch