An international group of scientists has begun work on developing a ChatGPT-like tool to accelerate scientific discovery.
In recent years, scientists have been leveraging artificial intelligence (AI) for the purpose of advancing scientific research and exploration.
AI's capability to analyze extensive datasets, simulate complex phenomena, and aid researchers in modeling and comprehending intricate systems has the potential to be a game-changer in various fields, including but not limited to medicine, astronomy, climate science, and materials research.
This new initiative, known as Polymathic AI, will harness the same technology that powers ChatGPT to create an AI-driven platform intended to support scientific research across various domains.
The system will be trained using a diverse set of numerical data and simulations gathered from research papers, patents, and other scientific publications covering several scientific disciplines.
The goal is to assist scientists in modeling and understanding a broad spectrum of complex phenomena, whether in astronomy, genetics, or climate research.
“This will completely change how people use AI and machine learning in science,” said Polymathic AI principal investigator Shirley Ho, a group leader at the Flatiron Institute’s Center for Computational Astrophysics in New York City, in an official release.
The idea behind this initiative “is similar to how it’s easier to learn a new language when you already know five languages,” added Ho.
The tool may be able to see things missed by humans
It will be able to understand scientific language and propose innovative hypotheses based on current evidence. Furthermore, the AI system will be able to analyze models and data, uncovering hidden patterns and linkages that humans may miss.
“Polymathic AI can show us commonalities and connections between different fields that might have been missed,” said co-investigator Siavash Golkar, a guest researcher at the Flatiron Institute’s Center for Computational Astrophysics.
Golkar elaborates that in the past, scientists used to have a comprehensive grasp of several subjects. This ability enabled them to find underlying aspects with respect to their work in the other sub-fields.
“With each scientific domain becoming more and more specialized, it is increasingly challenging to stay at the forefront of multiple fields. I think this is a place where AI can help us by aggregating information from many disciplines,” mentioned Golkar.
Accuracy top priority
The official release emphasized that ChatGPT has limitations, especially in terms of accuracy, particularly with numbers.
The fundamental goal of this new initiative is to address many of these issues by recognizing numbers as numerical values rather than considering them as equal to letters and punctuation marks.
Additionally, the training data will incorporate authentic scientific datasets that encompass the fundamental principles of cosmic physics.
The team will also ensure the “transparency and openness” of this AI initiative.
Ho mentioned: “We want to make everything public. We want to democratize AI for science in such a way that, in a few years, we’ll be able to serve a pre-trained model to the community that can help improve scientific analyses across a wide variety of problems and domains.”
The Polymathic AI team comprises researchers from a range of institutions, including the Simons Foundation and its Flatiron Institute, New York University, the University of Cambridge, Princeton University, and the Lawrence Berkeley National Laboratory. This diverse team features specialists in fields such as physics, astrophysics, mathematics, artificial intelligence, and neuroscience.
The researchers hope that this AI system can expedite the speed of scientific discovery, resulting in revolutionary advances in a variety of scientific areas.
We can't wait to watch how this initiative develop over the next few months.
Originally published on Interesting Engineering : Original article