| Hi, it's Drake in New York. If you're concerned about an AI doomsday, you'd probably be better off watching Barbie. But first... Three things you need to know today: • Biden vowed vigilance on AI
• JPMorgan is betting Asia chip stocks will catch up to the US
• AI demand from Wall Street is soaring Late last week, executives from some of the largest tech companies and AI startups went to the White House. They were there to help the Biden administration unveil a set of voluntary industry guidelines to help mitigate the dangers of artificial intelligence, a technology they are racing to develop and commercialize. Also late last week, millions of people went to see a movie about J. Robert Oppenheimer and the Manhattan Project. One of the most fascinating things about the endlessly fascinating Oppenheimer, and one of the main themes of the film, is that the American physicist who, more than anyone else, was responsible for bringing the first atomic bomb into existence then came to believe that America, and the world, should stop developing increasingly more powerful nuclear weapons. Since those weapons could lead to the destruction of civilization, that argument made a lot of sense. But the US government ignored Oppenheimer and, to boot, publicly humiliated him by stripping him of his security clearance. Those debates are the backdrop for what the Biden administration and its tech partners are trying to do now. AI, like nuclear weapons, is a technology that presents potentially species-wide risks. But nuclear history suggests the difficulty of curbing technological advancement. Some of the same arguments Oppenheimer confronted persist today, especially this one: that if we don't do it, someone else will. But the debates are, if anything, trickier now. Whereas nuclear weapons could only be created by massive government research and engineering efforts, AI technology is becoming ever more off-the-shelf. The decisions made by governments and the world's biggest tech companies may not matter so much if rogue actors can push the technology along on their own. Also, nuclear weapons do their terrifyingly destructive job in a predictable way. That was the basis of the insane existential standoff that managed to keep the Cold War cold. The ultimate outcome of turning on an advanced AI is far less knowable. Nor can nuclear weapons go off and make other nuclear weapons on their own — unlike AI, which, once it passes a certain threshold, could in theory start creating more sophisticated versions of itself. In other words, knowing when we've reached the point of no return is far less clear with AI than with the bomb. Even then, one of the tendencies of human beings is that it's hard to stop trying to figure something out, even if we know that thing might be terrible. As Oppenheimer himself famously said of the Manhattan Project, "When you see something that is technically sweet, you go ahead and do it, and argue about what to do about it only after you've had your technical success." —Drake Bennett Arm sells blueprints needed to design microprocessors and licenses technology known as instruction sets that dictate how software programs communicate with those chips. The initial public offering planned for the fall could be the year's biggest; it could also provide an interesting temperature check on the fever over AI.  Watch the Bloomberg Technology TV interview with Credo AI's Navrina Singh about company's AI safeguards commitments. Microsoft's success fighting the US regulators' challenge to its Activision Blizzard acquisition could ease the path to more deals at a time when Wall Street has been confronting a severe merger drought. Amazon plans to build a $120 million satellite-processing facility in Florida to support Project Kuiper, an initiative to sell internet access from space. A likely re-weighting of Sirius XM in the Nasdaq 100 is adding another twist to the stock's wild moves, with the market transfixed by an unprecedented short squeeze. Generative AI has spawned a ton of hype. Here's how to leverage it in your own work. | 
No comments:
Post a Comment