NAVIGATING AI CONSCIOUSNESS: FROM SCIENCE FICTION DREAMS TO SCIENTIFIC REALITY

Despite no evidence of sentience, a surprising number of experts and the public believe AI may already be conscious—20% of U.S. adults and 17% of researchers agree. This belief clashes with Microsoft leaders' warnings against anthropomorphizing AI— highlighting how science fiction deeply shapes our imagination, even among those building the technology.

- 

Illustration of the film Her (2013)

Modern artificial intelligence (AI) technologies are transforming everything from search engines to scientific research, yet they remain widely misunderstood.

Movies depict romantic partnerships with synthetic voices or robots plotting revenge. Even everyday conversations sometimes imply that systems like ChatGPT “listen” to our personal emails or guess our thoughts. On 5 November 2025, these misconceptions reached the boardroom: Microsoft's CEO Satya Nadella insisted that AI remain a tool and not a person, and Microsoft AI chief Mustafa Suleyman dismissed the pursuit of conscious AI as misguided.

In early 2024, Microsoft unveiled a voice assistant that spoke in a soulful tone reminiscent of the film Her, prompting viral comparisons between an AI product demo and science-fiction romance. Shortly afterwards, CEO Satya Nadella expressed discomfort with the term “artificial intelligence” itself. He said he dislikes anthropomorphising AI and prefers to think of it as a different kind of intelligence, emphasising that AI is a tool rather than a mind like his own. Nadella argued that conflating algorithms with human beings encourages unrealistic expectations and distracts from the important work of building useful systems.

Microsoft's head of AI, Mustafa Suleyman, went further in an interview published on 3 November 2025. He criticised projects attempting to bestow consciousness on machines and called such efforts “totally the wrong question”. Suleyman explained that models can mimic emotional responses but cannot truly feel; an AI that “complains” when overloaded does not experience pain the way humans do. Instead of chasing sentience, he argued, developers should concentrate on tangible benefits and ensure that AI serves human needs.