Surfshark's Lead System Engineer, Karolis Kaciulis, believes that instead of worrying about a futuristic AI-driven apocalypse, people should focus on the tangible problems that generative AI is already creating today. Recent concerns about the potentially catastrophic risks of artificial intelligence have gained attention, with major technology companies urging caution in AI development and some researchers warning of the possibility of AI leading to human extinction. However, Kaciulis views these warnings as a recurring marketing strategy, suggesting that the dystopian stories being told are more of a distraction than a genuine concern. Kaciulis points out real and present dangers associated with AI, such as deepfake fraud and the environmental impact of data centers. According to Surfshark's research, deepfake-related fraud has caused $2.19 billion in global losses, with $149 million of that amount occurring in the United Kingdom alone. Additionally, the environmental cost of AI is significant, as data centers require vast amounts of energy and water for cooling. Surfshark estimates that a single query to ChatGPT uses about 2 watt-hours of energy on average. Kaciulis also suggests that the AI industry may be highlighting potential future dangers to obscure a more ordinary reality: the progress of AI technology may be slowing down. He argues that it is not clear if newer AI models are genuinely more capable or if they are just better at mimicking human responses. This illusion of intelligence can be misleading, as users may trust these AI tools with personal and sensitive information without realizing they are interacting with systems designed to collect data for corporate use.