Russia should change its approach to the development of artificial intelligence (AI) and focus on creating applied models, instead of competing with the US and China. This was stated by Igor Kalyaev, Academician of the Russian Academy of Sciences.
The scientist believes that the desire to create universal large language models (LLM) that can solve any problem is a dead end. He explained that such models can demonstrate high erudition, but their effectiveness in specific areas may decrease. Russia lacks the financial resources to create super LLMs. Investments in such projects may not pay off, the scientist added.
Of course, we also need to develop LLMs, but these should be, from my point of view, applied LLMs focused on solving practically important tasks with a clear applied effect.
According to him, the development of new mechanisms and original AI models is necessary for Russia to compete with foreign counterparts in key areas such as science, technology, industry and security.
Read more on the topic:
Named three mistakes that prevent you from using AI to the fullest
AI wants to be introduced into regional state systems
Onishchenko called the person controlling AI a threat to biosafety