The head of artificial intelligence at Microsoft said those creating artificial intelligence should make sure it’s uncomplicated for the public to understand – and offered his own analogy that might lend a hand.
Mustafa Suleiman, CEO of Microsoft AIhe said during a talk at the TED 2024 conference that artificial intelligence is the latest wave of creation since the beginning of life on Earth and that “we are in the fastest and most significant wave in history.”
Suleyman said the industry must find appropriate analogies for the future potential of artificial intelligence to “prioritize security” and “ensure this novel wave always serves and empowers humanity.” While the AI community has always referred to AI technology as “tools,” Suleyman said that term does not capture its capabilities.
“To stem this tide, put human agency at its center, and mitigate the inevitable unintended consequences that may arise, we should start thinking about them as a novel kind of digital species,” Suleyman said.
He also said he sees a future in which “everything” – from people to companies to government – is represented by an interactive personality, or “personal artificial intelligence,” that is “infinitely competent,” “factually correct and reliable.”
“If AI achieves even a fraction of its potential” in finding solutions to problems in everything from health care to education to climate change, “the next decade will be the most productive in human history,” Suleyman said.
When asked what keeps him up at night, Suleiman replied that the AI industry faces the risk of falling into a “trap of reluctance to pessimism” when in fact it should “have the courage to face the potential of shadowy scenarios” to fully realize the opportunities artificial intelligence possible profits.
“The good news is that if you look at the last two or three years, there have been very, very few failures,” Suleyman said. “It is very arduous to say clearly what damage LLM has caused. “But that doesn’t mean that this will be the trajectory over the next ten years.”
While Sulejman said he sees humans facing the dangers of autonomous AI models in five to 10 years, he believes the potential dangers should be discussed now.