The tech world is abuzz with "agentic AI," which can perform specific tasks autonomously like booking flights. While AI agents aren't new, with the technology behind Siri dating back to the 1990s, the current emphasis on agentic AI highlights potential risks amplified by its interaction with the real world.
Concerns include data bias that could be introduced and scaled by AI agents, potentially leading to irreversible consequences. However, one of Siri's inventors, Babak Hodjat, argues that the greater danger lies in over-reliance and lack of critical evaluation of AI outputs, urging users to understand the boundaries of trust.
This caution is particularly noted in Europe, where strict regulations like the EU AI Act are in place. Jarek Kutylowski, CEO of AI company DeepL, believes Europe's approach is too cautious and risks stifling innovation, potentially causing the region to be left behind in the global AI race.
He suggests the larger risk is missing out on technological progress and its economic benefits by being overly focused on potential downsides. Kutylowski advocates for a pragmatic embrace of AI rather than attempting to halt its advancement.
