Bill Gates warns AI is powerful enough to cause a billion deaths

[UltraShane86]

Bill Gates has given one of his starkest warnings yet about artificial intelligence. Speaking to Kristen Welker on NBC's Meet the Press, he said AI is certainly powerful enough to drive events that could cause a billion deaths. He also said there has never been a weapon as powerful as the combination of people with ill intent and the latest AI tools. Coming from someone who spent decades as one of tech's biggest optimists, that is quite a shift in tone

This builds on a memo he put out in August, where he described AI as more consequential than almost any technology in history. In it he said AI could be the greatest equaliser ever invented or the worst source of injustice. He also warned that on the current course and speed, there is a very high chance of a net negative outcome. Those are strong words from the co-founder of Microsoft, a company that has bet enormous sums on AI

His answer is regulation. Gates wants federal legislation with enforceable safeguards, because he does not think companies policing themselves is enough. That includes proper government oversight, independent evaluators with ongoing access to the labs and real public accountability rather than voluntary promises

It is worth being clear about what he did and did not say. He was describing what AI could be used to do in the wrong hands, not predicting that a billion people will die or putting a date on anything. The concern is mainly about misuse, where bad actors use powerful tools to cause harm on a scale that was not possible before. That is a different worry from AI going rogue by itself

There is an interesting tension here as well. Only a couple of weeks ago the Gates Foundation announced it would spend at least $1 billion over two years expanding access to AI, mostly in education and health. So he clearly still sees huge upside. Is this a sensible call for guardrails from someone who knows the industry, or is the billion deaths line just headline bait? Keen to hear what people think

Bill Gates issues chilling AI doomsday warning saying one BILLION people will be killed... unless humanity takes his advice | Daily Mail Onlinewww.dailymail.com

[Cheugy]

Quite a turnaround from the man who spent years telling us AI would cure diseases and fix education. Maybe he has seen something in private that the rest of us have not

Or maybe he just realised how fast things are moving. Either way, a warning from Gates carries weight in Washington. Politicians listen to him

[BringItOnRhodes43]

Funny how the people who made their fortunes from tech always want regulation once they are comfortably at the top. New rules tend to protect the big players and make life harder for smaller rivals

I am not saying Gates is doing this deliberately. But it is worth asking who benefits

[LogicWitch58]

Federal legislation in the US seems unlikely any time soon. Congress struggles to agree on anything, let alone technology this complex

States will probably keep passing their own patchwork of laws. Companies will lobby against most of it

[Wolfhound39]

Independent evaluators with sustained access to the labs is the most useful idea in there. At the moment we mostly rely on companies telling us their own models are safe. That is like letting a drug maker approve its own medicines. Outside testers would at least give some confidence. The hard part is making it mandatory rather than optional

[Eastern Brett]

Misuse is the realistic risk in my view. Someone with bad intentions and a very capable model could do a lot of damage without needing any rogue AI. That is the scenario security people worry about most

[ShawnMichaels_99]

A billion deaths is the kind of line that goes straight into a headline. The actual argument is far more measured than the Daily Mail makes it sound

[Darren_20]

He is talking about capability, not probability, and that is an important distinction. Nuclear weapons are powerful enough to kill billions too, yet we manage them with treaties and oversight. That seems to be the model he wants for AI. It is hard to argue with that in principle

[SerialScroller52]

Spending a billion dollars on AI access while warning about a billion deaths is not as contradictory as it sounds. He clearly thinks the benefits are huge if the risks are managed. That is the same position many researchers hold. Push the useful applications and put guardrails around the dangerous ones

Whether governments can manage the second part is the open question. History says they will be late. They usually are