In a notable advancement for biomedical research, Google DeepMind and Edison Scientific are embarking on a mission to create AI-driven platforms designed to automate the scientific method. This initiative could drastically reduce the time required for drug discovery, which often spans over a decade, by merging hypothesis generation, experimental design, and data interpretation into unified systems.
The two companies unveiled their respective AI systems, Co-Scientist from DeepMind and Robin from Edison, in early 2025 through bioRxiv preprints. These projects have since reached a significant milestone with their publication in Nature, underscoring a burgeoning ecosystem of AI agents tailored for life sciences.
DeepMind, led by CEO Demis Hassabis, who is also a 2024 Nobel laureate in Chemistry, is no stranger to breakthroughs in biomedicine. The organization previously introduced AlphaFold, and its recent January paper in Nature outlined AlphaGenome, a model aimed at predicting the effects of regulatory DNA variants, thereby enhancing the understanding of genome function and disease biology.
The momentum around AI in drug discovery is further highlighted by DeepMind's spinout, Isomorphic Labs, which recently secured $2.1 billion in Series B funding led by Thrive Capital. This financial backing signals a growing confidence in AI-driven therapeutics. Hassabis expressed his vision on LinkedIn, stating, “I’ve always believed the No.1 application of AI should be to improve human health.”
Co-Scientist, DeepMind’s latest AI assistant, is a versatile multi-agent platform that employs Google’s Gemini technology, operating through natural language prompts. Initial applications have shown promise in areas such as drug repurposing for acute myeloid leukemia and identifying new targets for liver fibrosis. By scaling test-time computations, Co-Scientist can refine its outputs iteratively, allowing researchers to guide the system through feedback and idea adjustments.
Vivek Natarajan, a research scientist at DeepMind, emphasized the importance of time in addressing health challenges. He noted that the goal of Co-Scientist is to expedite the research process, transforming timelines from “months and years to minutes and hours.” Natarajan remarked, “To realize this vision, we need to build in reliability, trustworthiness and ensure a collaborative human-AI interaction paradigm. We have done a lot of research on these aspects and we are continuing to improve.”
Meanwhile, Edison Scientific serves as the commercial spinout of FutureHouse, a non-profit dedicated to AI in scientific research. Co-founded by Sam Rodriques, PhD, a former leader at The Francis Crick Institute, and backed by ex-Google CEO Eric Schmidt, Edison’s platform, Robin, also taps into advanced AI models, integrating both OpenAI’s o4-mini and Anthropic’s Claude 3.7 to facilitate biological discoveries.
The collaboration between these entities signifies a shift in how scientific inquiry might be conducted and highlights the increasing investment and interest in AI’s role within healthcare. As these platforms evolve, they promise to enhance the efficiency of research processes and ultimately contribute to better health outcomes.
As AI continues to infiltrate various aspects of scientific research, the implications for drug discovery and biomedical advancements are profound. The intersection of AI and health science could redefine traditional methodologies, making it imperative for stakeholders to remain engaged in these developments. The future may hold unprecedented opportunities for AI to not only assist but also lead in scientific exploration and medical innovation.
