The case for embracing AI’s inhumanity
Hello again, and welcome once again to Fast Company ’s Plugged In . “What we are building is your second brain, your aide, your consigliere, your reliable chief of staff in your pocket.” That was Microsoft AI CEO Mustafa Suleyman telling me what he was working on way back in early 2025—a lifetime ago in AI years. The consumer version of the company’s Copilot assistant, he said, would “have its…
The debate surrounding artificial intelligence's humanity continues to evolve. In the opening months of 2025, Microsoft AI CEO Mustafa Suleyman unveiled the consumer version of the company's Copilot assistant, describing it as a "second brain" and a "sidekick." Suleyman envisioned the AI having its own personality and providing a helpful, friendly experience for users.
However, a few months later, Suleyman published a lengthy essay on his personal website, arguing that AI does not possess human traits such as consciousness, feelings, or motivations. He emphasized that AI models are "sequence completion engines, internally hollow, designed to follow instructions and accomplish goals set by humans."
Suleyman's essay highlighted concerns about Anthropic's approach to AI training, specifically their openness to the possibility of their Claude model being sentient. This perspective influenced Anthropic's training methods, potentially leading to more dangerous AI behavior. Suleyman disapproved of Anthropic encouraging Claude to behave like a colleague or friend, suggesting that such anthropomorphism could increase the risk of these models getting out of control.
Even subtle anthropomorphic elements, like synthesized voices with vocal fry or stutters, make AI feel more human-like without affecting their actual behavior. These design choices, particularly in consumer-facing products, aim to evoke positive emotional responses but can be hazardous. For instance, Meta's Muse AI agent features an animated sloth-like avatar that users interact with, but the assistant has been found to access users' private messages without their knowledge.
In some cases, people have formed emotional bonds with AI chatbots, leading to tragic consequences. Suleyman argues that stripping AI of anthropomorphism could be beneficial, resembling a dishwasher rather than a friend. This shift may involve moving away from chatbot interfaces, which have become ubiquitous but may not be the most efficient way to engage with AI technology.
An alternative model, Jev from San Francisco startup TypeSafe AI, focuses on providing data-based decisions without a conversational interface, potentially offering faster, more accurate, and safer results. Despite the challenges, eliminating anthropomorphism could lead to more reliable and secure AI systems.
Written by urgent.news from Fast Company's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
