Anthropic AI: Claude Fable 5 for Drug Discovery

AI company Anthropic has launched a new large language model, Claude Fable 5, with a focus on life sciences. The company claims it offers a significant lead in drug discovery and scientific research. What's interesting is this autonomous AI is the first based on Anthropic's top-tier Mythos 5 model. This model was previously deemed too risky for public release due to its ability to outperform humans in some hacking tasks. However, the launch faced criticism from AI researchers. This is because it initially used "safeguards" to reroute users asking questions in sensitive areas like biological research to an older, less capable AI model. Researchers were concerned this downgrade would happen invisibly. Anthropic has since backtracked on this plan and will now alert users when their requests are rerouted. The company stated that 95% of cases were handled by Claude Fable 5 without rerouting. The company highlights Claude Fable 5's capabilities in drug design, claiming it accelerated aspects of the process by about ten times. It also matched or surpassed human operators in tasks like choosing binding sites and running protein design tools. The platform produced promising drug candidates for nine of 14 protein targets. It also generated novel hypotheses in molecular biology, with one hypothesis already validated in lab testing. The bottom line is that this new AI model could significantly impact the speed and efficiency of scientific and medical research.