- Google’s AI-CO scientist, built on Gemini 2.0, collaborates with researchers on discoveries
- It uses specialized agents to generate, evaluate and refine scientific hypotheses
- Researchers can interact naturally, give ideas or feedback to guide AI research
Artificial intelligence has already had a major impact on scientific research by speeding up discoveries, improving accuracy and dealing with large data sets that would be almost impossible for humans to analyze effectively. AI-driven algorithms can help with the discovery of new drugs, optimize energy storage materials and help model climate change.
A number of projects have been set up to make AI more useful and more reliable in scientific surroundings. We have previously written about the concept of “Exocortex” aimed at giving a bridge between the human mind and a network of AI agents, and recently developed an Australian research team a generative AI tool called LLM4SD (large language model for scientific discovery), designed to accelerate scientific breakthroughs.
Now, Google is also launching a similar initiative aimed at transforming AI into a co-scientist who can accelerate scientific discoveries. The tech giant explains, “The AI-C-C-scientist is a multi-agent AI system designed to act as a collaborative tool for researchers.”
Implementation of specialized scientific agents
The AI-CO scientist is built on Google’s Gemini 2.0 and is the result of collaboration between Google Research, Google Deepmind and Google Cloud Ai teams. It is designed to “mirror the reasoning process that supports the scientific method.” Google says its system is intended to “reveal new, original knowledge and to formulate detectably new research hypotheses and suggestions, build on prior evidence and tailored to specific research goals.”
The system will use a variety of specialized agents – generation, reflection, location, evolution, proximity and meta -review – that can iteratively generate, evaluate and refine hypotheses. Google says researchers will be able to interact with the system in any way best fits their needs. This will include the delivery of their own seed ideas or feedback about generated output in natural language.
“The AI-CO scientist also uses tools such as web search and specialized AI models to improve the grounding and quality of generated hypotheses,” Google says.
The company does not plan to hurry to hurry, that the company offers access to the system for research organizations through a trusted test program.
