The right AI for every purpose: George Tilesch and Erik Slooten's vision for 2025

There is no question that artificial intelligence will continue to be a defining technology in the world of IT in the future. But exactly where it is headed, what we can expect from AI agents, and how we will have the energy to train the models? George Tilesch, founding president of the PHI Institute for Augmented Intelligence and lead AI expert at EY AI Confidence, as well as Erik Slooten, partner at EY AI Confidence, spoke to our newspaper about these topics.

Can we talk about an AI bubble, and if so, will it burst?

George Tilesch: The term "AI bubble" refers to the excessive media hype surrounding the technology, which often emphasizes two extremes: either that AI development has stalled or that its progress is unstoppable. This fluctuating narrative is a natural result of a cleansing process, where the efficiency of AI applications and technologies gradually crystallizes. Fears related to the AI bubble mostly stem from the fact that many companies overestimate the potential of the technology while lacking the proper expertise or strategy for implementation. Although the idea of a bubble around AI is popular, it is unlikely to burst in 2025, as the foundations of technological development are solid.

Erik Slooten: Generative AI – such as ChatGPT, Bard, and other systems – has already proven to create real value. Large companies in the industry are not only focused on maintaining the technology but also on its continuous development, making a complete collapse unlikely. Industry players' experiences show that the application of AI is not just the privilege of tech giants. Smaller companies are increasingly discovering the potential of AI, for example, in automating customer service systems or data analysis tasks. This is similar to the development of the internet: some companies may fail, but the core technology continues to thrive and evolve.