AI Doesn’t Need Giant Data Centers — So Why Are We Building Them? |9|15|26

Everything you’ve heard about the debate over AI and data centers is built upon a false choice. Not only do we not need a centralized cloud-based AI strategy built upon thousands of hyperscale data centers, but it is holding us back from cheaper, more efficient distributed AI technologies. I’m joined by computer scientist and AI expert Jay Valentine, VP of operations at Fractal Computing, who explains why the future of technology is distributed, localized, and edge and sovereign AI — technology that can run on existing computers and hardware while being dramatically cheaper, faster, more private, more secure, and less vulnerable to centralized points of failure. Valentine explains why AI is fundamentally a powerful pattern-recognition technology rather than the sentient superintelligence often portrayed in the media, and he demonstrates how sophisticated applications can be moved away from massive, centralized data centers and onto distributed systems. AI is NOT synonymous with ChatGPT, and if the president only understood that, he’d realize that he’s the one putting us at a disadvantage with China.  Learn more about your ad choices. Visit megaphone.fm/adchoices