In a striking departure from the industry's relentless push towards more powerful artificial intelligence, Microsoft AI CEO Mustafa Suleyman has issued a stark warning, labelling the pursuit of AI superintelligence as an 'anti-goal' for humanity. This cautionary stance was presented during a recent episode of the 'Silicon Valley Girl Podcast', where Suleyman argued that creating an intelligence vastly superior to our own may not lead to a positive future.
Why Superintelligence is a Dangerous Goal
Suleyman provided a clear definition of the term, describing superintelligence as an AI system capable of self-improvement, setting its own goals, and acting independently of humans. He expressed deep concern over this vision, stating it does not align with a hopeful outlook for humanity. The core of his argument hinges on control, or the lack thereof. "It would be very hard to contain something like that or align it to our values. And so that should be the anti-goal," Suleyman asserted. He clarified the distinction, often blurred, between Artificial General Intelligence (AGI) and superintelligence, noting AGI is a step before the more autonomous and powerful superintelligence.
The Imminent Arrival of AGI and a 'Humanist' AI
While warning against superintelligence, Suleyman aligned himself with a significant prediction made by Google DeepMind CEO Demis Hassabis. He agreed that AGI could be achieved within the next five years, enabling AI to perform at human level in most cognitive tasks. He pointed out that current models already surpass human capabilities in specific areas like summarization, translation, and research. Furthermore, he noted AI is making rapid progress in roles such as project management, marketing, and human resources.
In response to these advancements, Suleyman revealed that Microsoft is focusing on developing a "humanist" version of superintelligence. This approach aims to create an AI that is fundamentally aligned with human interests, acting as a supportive partner rather than an independent entity.
A Profound Shift in Work and Society
Suleyman elaborated on the monumental impact this technology will have, predicting it will "fundamentally change work in the most profound way." He envisions a world where the distance between an idea and its realization will collapse, leading to an explosion of creativity and competition. "People are just going to be thinking new companies into existence, new products, new pieces of poetry," he said, describing a future of democratized access to intelligence.
However, this transformative power comes with a critical need for oversight. Suleyman emphasized the urgent necessity to establish regulations and guardrails to ensure that increasingly autonomous AI agents collaborate with humans safely and effectively, preventing them from operating entirely on their own.
