Even if we turn to a very recent example of defending a thesis written with the help of AI. Even there, we are talking about the analytical part – to collect and analyze existing data. For example, scientists can thus test a hypothesis, narrow the search area, and perform routine operations. In some studies, you can not grow cultures in a Petri dish, but use a mathematical model, rucing the research time, and leave cultivation to confirm the hypothesis. Or, with the help of train AI, to recognize tumors, to find some non-obvious patterns. But what such technologies cannot do is generate new areas of knowlge or look for fundamentally new solutions. They can analyze the available data, but synthesizing new knowlge from this is still very limit. Will this ever be possible? – Rather, when we come to a real AI capable of thinking independently. Last year, a project came to our foundation that want to answer the question – is it possible to mathematically assess that an artist has fallen into a state of inspiration? Is it possible to evaluate this process, reproduce it in numbers? Unfortunately, the topic turn out to be outside the scope of the area fund at that time. But let’s imagine that such a study turn out. In fact, we would be one step closer to an AI that can create something. Not by a text query compil by a person, but to generate independently what he wants to see. If we transfer this to science, the scientist also finds himself in a certain state when he is engag in scientific activities. He may not call it inspiration, but it is something close to him. When we can understand, model and digitize such processes, then it is possible. But now, again, AI technologies play a supportive role in science. Freaks or geniuses — How to understand what scientific projects to invest in when they are only at the idea stage? –