- Google’s AI co-scientific, based on Gemini 2.0, collaborates with researchers for discoveries
- Use specialized agents to generate, evaluate and refine scientific hypotheses
- Scientists can naturally interact, providing ideas or comments to guide AI research
Artificial intelligence has already had a great impact on scientific research by accelerating discoveries, improving the precision and management of vast data sets that would be almost impossible for humans to analyze efficiently. Algorithms can help in the discovery of new medications, optimize energy storage materials and help model climate change.
Several projects have been established to make AI more useful and more reliable in a scientific environment. We have previously written about the concept of “exocortness”, whose objective is to provide a bridge between the human mind and a network of AI agents, and more recently, an Australian research team developed a generative the AI tool called LLM4SD (large language model for scientific discovery), designed to accelerate scientists.
Now, Google is also launching a similar initiative, whose objective is to turn AI into a co-scientific that can accelerate scientific discoveries. The technological giant explains: “The co-scientist of AI is a system of multiple agents that aims to function as a collaboration tool for scientists.”
Deploy specialized scientific agents
The co-scientist of AI is based on Google Gemini 2.0 and is the result of the collaboration between Google Research, Google Deepmind and Google Cloud AI. It is designed to “reflect the reasoning process that supports the scientific method.” Google says that your system is intended to “discover a new and original knowledge and formulate demonstrably novel research hypotheses and proposals, based on previous evidence and adapt to specific research objectives.”
The system will use a series of specialized agents (generation, reflection, classification, evolution, proximity and review target, which can generate, evaluate and refine iterative hypotheses. Google says that scientists can interact with the system in any way that best suits their needs. This will include providing their own seeds of seeds or comments on the results generated in natural language.
“The co-scientific of AI also uses tools, such as the web search and specialized the models, to improve the basis and quality of the hypotheses generated,” says Google.
Not wishing to hurry its implementation, the company plans to offer access to the system for research organizations through a trusted tester program.
