AI's Journey to the World Economic Forum: Discussions on the Uncharted Territories of Artificial Intelligence
In a thought-provoking discussion at the AI Journey Session during the 16+ Eastern Economic Forum, experts and industry leaders in AI development shared their insights on the development of reasoning-capable and self-learning AI models.
Denis Dimitrov, Managing Director of Data Research and Head of the Basic Models Department at Kandinsky, Sberbank, was one of the speakers. He spoke about the current state of GenAI and highlighted the potential of Sber GigaChat (0+) and Kandinsky (0+) models for automating intellectual labor.
Dimitrov traced the evolution of artificial intelligence from its origins to the latest breakthroughs in creating text, images, videos, and sound. He also discussed the role of humans in the new reality where AI is rapidly transforming all aspects of life.
Piotr Anokhin, Leading Researcher at the AIRI Institute and Senior Researcher at the Artificial Intelligence Research Center of Moscow University, focused on self-improving systems and AI's ability to make scientific discoveries. Anokhin outlined significant works in the field of self-improving systems and AI scientific discovery and formulated key challenges that lie ahead.
Andrey Daudrikh, Director of the Department of Social Research and Consulting at the Analytical Center of ВЦИОМ, presented survey data showing that 63% of Russians use AI technologies, primarily through voice assistants and neural networks. Most of them perceive AI as an assistant.
Valentin Khrulev, Head of the "Generative Design" group at the FusionBrain laboratory, AIRI Institute, demonstrated how generative technologies are being used in unexpected fields, such as design and construction. Khrulev discussed the creation of discrete diffusion models that synthesize optimal planning solutions, which was previously a human domain.
In the construction industry, agent systems are being employed, as Khrulev explained.
The session also touched on the transition to reasoning, self-learning AI models, a transformative tool currently being used across various sectors. The potential of these models for automating intellectual labor was a recurring theme throughout the discussion.
In conclusion, the AI Journey Session provided a comprehensive overview of the current state and future prospects of AI, emphasizing its transformative potential across various sectors and the role of humans in this new reality.
