Mustafa Suleyman, the head of AI at Microsoft and a co-founder of DeepMind, has taken a bold stance in the race toward advanced artificial intelligence, calling the pursuit of artificial superintelligence, AI that vastly surpasses human capabilities, an “anti-goal.” In a recent interview on the Silicon Valley Girl Podcast, Suleyman laid out his vision of a “humanist superintelligence” designed around human interests rather than abstract technological dominance.
According to Suleyman, the idea of an all-powerful AI raises serious alignment challenges. “It would be very hard to contain something like that or align it to our values,” he said. He further warned that efforts to grant AI systems something resembling consciousness or moral status are misguided: “These things don’t suffer. They don’t feel pain. They’re just simulating high-quality conversation.”
Instead of chasing a god-like, superintelligent AI, Suleyman wants Microsoft’s research to focus on a more pragmatic, human-first approach. His vision emphasises tools that empower people, enhance creativity, bolster education, and allow technology to augment, not replace, human experience.
Suleyman’s comments come amid growing debate in Silicon Valley over the future direction of AI. Leaders like OpenAI CEO Sam Altman have publicly embraced the pursuit of artificial general intelligence (AGI) and even superintelligence. Altman has said that superintelligent systems could “massively accelerate scientific discovery … and in turn massively increase abundance and prosperity.” He has also expressed surprise if superintelligence does not emerge by 2030.
Similarly, DeepMind co-author Demis Hassabis has suggested that AGI could arrive “in the next five to 10 years,” envisioning systems deeply integrated into daily life. Yet not everyone agrees. Meta’s top AI scientist, Yann LeCun, has expressed more caution, arguing that we may be “decades” away from genuine AGI. LeCun warned that scaling up compute and data alone doesn’t guarantee smarter systems or deeper understanding.
Suleyman’s views signal a tangible shift in Microsoft’s AI strategy, away from pure capability competition toward long-term responsibility. By calling superintelligence an “anti-goal,” he is explicitly pushing back against an accelerationist narrative that has dominated parts of the AI world.
He wants to see AI deployed as a force for human-centric progress rather than a race to build a system that outthinks and outruns us.
OpenAI launches ChatGPT group chats pilot in Japan, New Zealand, South Korea, and Taiwan
