Listen to the Podcast:
Brad Smith, president and vice chair of Microsoft (MSFT), said on CBS’s “Face the Nation” on Sunday that the government needs to move faster to regulate AI, which has more ability to help people than any other invention before it.
Smith said that its uses are almost “everywhere.” “It’s used in medicine, drug discovery, and diagnosing diseases. It’s also used by the Red Cross and others in disasters to find the most vulnerable people in places where buildings have collapsed,” said the executive.
Smith also said that AI isn’t as “mysterious” as many people think. He also said that AI is getting stronger.
Smith said, “If you have a Roomba at home, it learns how to move around your kitchen by using artificial intelligence to figure out what to bump into and how to avoid it.”
Concerning worries about AI’s power, Smith said that people who lived before any technology would have thought it was dangerous.
Smith said that a safety stop should be put in place.
Smith said that AI will change jobs over the course of years, not months.
“Most of us will have to change the way we work,” Smith said. “This will be a new set of skills that we’ll need to, well, build and get.”
Smith said that to stop things like the fake photo of the explosion near the Pentagon, there needs to be a watermark system or “use the power of AI to detect when that happens.”
“You put in what we call information. It is part of the file, and if it is taken out, we can tell. If there’s a changed version, we make a “hash” of it. “Think of it like a fingerprint, and then we can look for that fingerprint all over the internet,” Smith said, adding that a new way should be found to find a balance between controlling deepfakes, misleading ads, and free speech.
Smith said that since the US presidential election year is coming up and cyber influence operations from other countries are still a threat, the tech industry and states need to work together on an international project.
Smith wants the government to set up a new body to oversee AI systems.
Smith suggested “something that would ensure not only that these models are developed safely, but that they are deployed in, say, large data centers, where they can be protected from cybersecurity, physical security, and national security threats.”
According to Elon Musk and Apple co-founder Steve Wozniak, a six-month moratorium on AI systems more potent than GPT4 is not “the answer,” according to Smith.
“I don’t think China’s going to jump on that bandwagon,” Smith said, “rather than slow down the pace of technology, which I think is extraordinarily difficult.” We’ll move more quickly if we use six months.
Smith proposed an executive order in which the government itself declares that it will only purchase AI services from businesses that are putting AI safety standards into place.
Smith stated, “The world is progressing. Let’s ensure that America at least stays up with the rest of the globe.
