Brad Smith Advocates for Safety Measures in AI Systems Overseeing Critical Infrastructure, Ensuring Human Control is Always Maintained
Addressing reporters in Britain, Microsoft President Brad Smith refuted claims of an imminent breakthrough in artificial intelligence, particularly the emergence of highly intelligent AI within the next 12 months. He emphasized the improbability of witnessing such advancements in the short term, suggesting it may take several decades for such technology to materialize. Smith stressed the urgency of prioritizing safety measures in the evolving landscape of AI.
Smith stated, “There’s no chance that we will witness this supposed AGI, where computers surpass human capabilities, in the next 12 months. It’s a matter of years, possibly many decades, but I strongly believe that the time to prioritize safety is now.”
Regarding recent developments at OpenAI, where co-founder Sam Altman was briefly removed from the CEO position and later reinstated due to employee and shareholder outcry, Smith clarified that Altman’s removal wasn’t primarily linked to concerns about a potentially risky discovery called Project Q*. This internal initiative at OpenAI aims to advance artificial general intelligence (AGI), defined as autonomous systems outperforming humans in most economically valuable tasks.
When asked about the potential impact of Project Q* on Altman’s removal, Smith stated, “I don’t believe that’s the case at all. There was evidently a divergence between the board and others, but it wasn’t primarily about such a concern.”
Highlighting the need for safety measures in AI systems overseeing critical infrastructure, Smith drew parallels to safety brakes in elevators, circuit breakers for electricity, and emergency brakes for buses. He emphasized the importance of implementing safety brakes to ensure AI systems always remain under human control.
