US scientists prove that AI can see the future of science
AI can predict what great discovery humans will make in the future.
New study offers new ways to use artificial intelligence (AI) to expand and accelerate scientific discovery. The authors have developed models that not only predict future scientific discoveries, but also generate hypotheses that might not be considered by human scientists in the near future.
The authors explained that improving AI’s predictive abilities could accelerate the pace of scientific progress. In addition, experts believe that AI can identify and bypass the “blind spots” of human understanding, allowing us to look beyond current scientific frontiers.
The researchers trained AI models based on publications of scientific discoveries, which allowed them not only to improve the predictions of future discoveries by 400%, but also to predict with more than 40% accuracy the real scientists who will make these discoveries.
This approach allows you to create a “digital twin” of the scientific system, which can be used to model future research. The method allows to detect weaknesses and shortcomings of the current system of education and science, which hinder the process of discovering new ideas and approaches.
One of the authors of the study, Professor Evans, insists that artificial intelligence should serve the purpose of improvement and expansion, and not merely copying human intellectual capacity. Instead of creating artificial intelligence, Evans calls for looking at AI as a means to radically increase human intelligence.
Evans sees this as a way to expand and improve the human ability to explore uncharted areas of science. AI, in his opinion, should work in addition to human thinking, and not as a substitute for it. Such an addition could lead to significant collective progress in science.