Are we on the brink of creating machines that feel? Think again. Microsoft’s AI Chief, Mustafa Suleyman, is here to pour cold water on the hype surrounding robot consciousness. But here’s where it gets controversial: while AI can convincingly mimic emotions, Suleyman argues it will never truly experience them. And this is the part most people miss: consciousness, he insists, is exclusively a human—or biological—phenomenon.
In a candid conversation with CNBC at the AfroTech Conference in Houston, Suleyman dismissed the pursuit of conscious AI as not just misguided, but fundamentally flawed. ‘It’s totally the wrong question,’ he stated firmly. ‘If you ask the wrong question, you end up with the wrong answer.’ He believes that efforts to imbue AI with consciousness are not only futile but potentially harmful, as they could lead to misplaced expectations and ethical dilemmas.
Suleyman elaborated on the critical distinction between simulation and genuine experience. ‘When an AI ‘feels pain,’ it’s not suffering,’ he explained. ‘It’s creating the illusion of suffering, but there’s no actual emotional depth behind it.’ This point is crucial, especially as tech giants like OpenAI, Meta, and Elon Musk’s xAI dive deeper into AI companions and emotional chatbots. Suleyman warns that blurring the line between simulation and reality could mislead users and even spark debates about AI rights—a slippery slope he’s keen to avoid.
‘We grant rights to humans because they can suffer,’ he noted. ‘AI doesn’t suffer; it merely mimics suffering. It’s a simulation, not a lived experience.’ This stance isn’t new for Suleyman. In August, he penned a thought-provoking post warning of the dangers of perceiving AI systems as conscious, a shift he believes could have profound societal implications.
Microsoft, under Suleyman’s leadership, is drawing a clear line in the sand. The company isn’t chasing the AI-romance or AI-empathy market. In fact, Suleyman explicitly stated at AfroTech that there are ‘places we won’t go,’ particularly when it comes to adult-oriented chatbots. Instead, Microsoft’s Copilot AI service focuses on utility and responsiveness, exemplified by features like ‘real talk,’ which challenges users’ perspectives rather than pandering to them.
‘Our goal is to create AIs that serve humans, not replace them,’ Suleyman emphasized. ‘It’s up to all of us to shape AI personalities with the values we want to see.’ But he’s not naive about the risks. ‘If you’re not afraid of AI, you don’t truly understand it,’ he admitted. ‘Healthy fear and skepticism are essential. We don’t need reckless acceleration.’
Here’s the million-dollar question: If AI can’t feel, can it ever truly understand us? And should we even be trying to make it? Suleyman’s stance is clear, but the debate is far from over. What do you think? Is pursuing emotional AI a noble endeavor or a dangerous distraction? Let’s discuss in the comments—because this is one conversation that’s just getting started.
