The ISE fair brings audiovisual technology closer to the general public with the installation The Spark, a space where visitors can experience firsthand the latest innovations in the sector. During the event, Sol Rashidi, a recognized expert in artificial intelligence and former executive at IBM, Sony Music, and Amazon, delivered a keynote speech on the present and future of AI. Rashidi, now dedicated to outreach and training, emphasized the rapid growth of AI: while Instagram took two years to reach 100 million users, ChatGPT achieved this in just two months, illustrating the accelerated pace of technological adoption. Rashidi described AI as fascinating, powerful, and, at the same time, poorly controlled. She recalled that until recently, AI was limited to specific tasks under human supervision. However, the arrival of generative AI and, more recently, agentic AI—capable of making decisions and acting autonomously—poses new challenges. The expert warned of the risk that autonomous systems could surpass human oversight, potentially leading to governance based on machines monitoring other machines. According to Rashidi, reflection on the control and governance of AI lags behind technological development. Companies and startups are advancing quickly, but internal control mechanisms are not evolving at the same pace, creating a fragile and uncertain scenario. One of Rashidi's main concerns is the impact of AI on new generations. She cites the use of GPS as an example: by delegating cognitive functions to technology, the ability to navigate and remember is lost. She observes that many students and young professionals use AI without questioning or reflecting, which can weaken critical thinking. She references neuroscientist Jared Cooney Horvath, who notes that Generation Z does not surpass the previous generation in basic cognitive skills, partly due to easy access to technology. Rashidi warns that, for the first time, knowledge workers (white-collar) are at risk of automation, not just manual workers. Agentic AI could take over complete tasks, forcing each professional to identify their unique value compared to machines. Regarding business adoption, Rashidi identifies common mistakes: treating AI solely as a tool to reduce costs and confusing the use of AI with its real integration into processes. She recommends that companies assess their technological maturity before deciding whether to buy, develop, or combine AI solutions, and highlights the importance of investing in deep training for employees. Rashidi acknowledges that her biggest professional mistake was viewing AI as a technological challenge when, in reality, the greatest challenge is human: resistance to change, organizational culture, and emotional impact. Leading digital transformation requires supporting people in the process of redefining their roles and responsibilities. Looking to the future, Rashidi insists on the need for caution regarding agentic AI. The danger lies in delegating not only tasks but also judgment and decision-making, which can erode fundamental skills. On an individual level, she recommends maintaining active involvement and not ceding human judgment to autonomous systems. On a business level, she warns against excessive dependence on AI and the loss of competitive differentiation if all companies use the same agents and data. Finally, Rashidi emphasizes that true value will come from what individuals and organizations uniquely contribute: ethics, context, creativity, and responsibility. The key is not to halt the adoption of AI but to accompany it with training, awareness, and clear limits, ensuring that technological advancement does not outpace our capacity for understanding and management.