The Present and Future of AI Development: Classical Mathematics Is No Longer Enough

MIT Technology Review has published an intriguing article, 'What's Next for AI and Math.' The entire modern world is built on mathematics, and breakthroughs in this discipline could open the door to truly stunning new technologies. However, DARPA (the U.S. Defense Advanced Research Projects Agency) believes that mathematics is stuck in the past. That is why the agency has launched an initiative called expMath (Exponentiating Mathematics), which aims to accelerate the development of artificial intelligence in the field of mathematics by creating tools that will become 'AI co-authors.' The ultimate goal is to create an AI capable of independently solving the most complex mathematical problems. True, researchers note that this requires unconventional thinking, since higher mathematics is an experimental field that includes trial and error and moments of insight. And despite impressive AI achievements, we are still far from creating truly 'mathematically thinking' systems.

Meanwhile, in Russia, the results of the selection for the third wave of research centers in the field of artificial intelligence have been presented. Each of the seven selected universities and scientific organizations will receive grants of 676 million rubles to conduct research in the areas of strong, trusted, and multi-agent AI. As reported on the government website