The Forefront of 'Cell Rejuvenation' Research Opened Up by AI
On May 19, 2026, Google DeepMind announced that its AI research partner, 'Co-Scientist,' successfully discovered multiple new gene candidates capable of reversing cell aging and achieved cell rejuvenation in actual experiments. The analysis of vast amounts of data, which previously took several months, was reportedly completed in just a few days, marking a significant achievement that strongly signals the advent of the 'AI for Science' era, where AI transforms the scientific discovery process itself. This breakthrough is particularly noteworthy for Japan, the world's most rapidly aging advanced nation, as a major step towards extending healthy lifespans and treating age-related diseases.
Technical Details: The Mechanism of the Collaborative AI Agent 'Co-Scientist'
At the core of Co-Scientist is a 'multi-agent AI system' built upon Gemini. Rather than a single massive AI, it's a mechanism where multiple AI agents, each with specialized roles, collaborate and act like virtual co-researchers. When a researcher sets a research goal in natural language, a 'Generation Agent' first leverages web searches to survey vast numbers of papers and construct initial hypotheses. Next, a 'Reflection Agent' rigorously verifies the validity and quality of these hypotheses, much like a peer reviewer. Furthermore, a 'Ranking Agent' evaluates and ranks multiple research proposals using the Elo rating system, known for assessing chess player strength, and an 'Evolution Agent' combines promising hypotheses to refine them into more sophisticated and innovative ideas. By rapidly repeating this cycle of 'generation, deliberation, and evolution,' Co-Scientist discovers cross-disciplinary insights that humans might overlook, thereby accelerating research and development.
Impact and Outlook for Engineers
The advent of Co-Scientist holds significant implications for engineers in Japan. Previously, specialized fields like biology and medicine presented high barriers to entry for AI engineers. However, with the proliferation of tools like Co-Scientist, which support AI engineers in acquiring specialized knowledge and validating hypotheses, collaboration between interdisciplinary experts and AI engineers will accelerate further. 'AI for Science' is a major trend continuing from the success of AlphaFold in protein structure prediction, and going forward, AI utilization will become essential in various scientific fields such as drug discovery, materials development, and energy issues. This signifies the opening of new career paths for AI engineers. The value of engineers who can learn life science domain knowledge and contribute to such cutting-edge research and development is expected to increase significantly in the future.
📦