Two Astonishing Things (One Of Them, TOTALLY Astonishing) October 4, 2024 DID YOU LIKE SEINFELD? VEEP? If you missed Monday’s Zoom with Nebraska Senate candidate Dan Osborn, consider signing up for this one with Julia Louis-Dreyfus — Sunday 7pm EDT. As before, you can “buy” a free ticket and decide whether and how much to give once you hear the pitch. I, for one, was impressed. But this is not one of the astonishing things. DO YOU LIKE TAYLOR SWIFT? THE 76-ERS? CRUISING TO COZUMEL? Andy Cohen? Stephen Colbert? Shakira? P!nk? Sign up — or get your kids or grandkids to signup — to earn free raffle tickets at FanOut.Vote. That is not one of the astonishing things, either. The first astonishing thing is Jack Smith’s 165-page brief outlining in damning detail how defendant Trump acted to cling to power after losing the 2020 election. No Trump supporters will read it, just as none read the Mueller Report. (Known to movie lovers as, the “I’m not listening!” approach to knowledge acquisition). But choosing not to know something doesn’t make it any less true. Not being a Trump supporter, maybe you’ll breeze through it this weekend. (They are double-spaced pages.) The second — TOTALLY — astonishing thing is this 14-minute audio clip. Two experts talking colloquially about artificial intelligence. What makes it astonishing is that the two experts are not human. It is a conversation about AI between two AI’s. As veteran computer scientist Gregory Miller — who co-founded the OSET Institute and TrustTheVote and has his own, real, podcast — explained it for me: This clip is from a podcast that explains a new audio AI technology that greatly enhances the emerging conversational capabilities of AI agents. There is just one completely wild twist: the podcast participants engaged in this conversation about how this new level of conversational AI tech works… are both AI agents themselves. They’re not real; they don’t exist; they’re figments of two machines’ creation. Both were fed a blog post from computer scientist Simon Willison and subsequently autonomically generated a real-time, on-the-fly conversation with one another. What makes it so ridiculously authentic is the integration of disfluencies. . . language elements such as changes in tone, pauses, “ums” and “ahs” . . . because, well, no one can listen to a pair of robots with mechanical-audio voices talk with one another. It boldly illustrates the deep and disturbing depth of AI deep-fake potential. Sure, AI may cure cancer and Parkinsons and solve the climate crisis. That’s nice. But as it races along, where will actual humans fit in? And what if someone decides to use it for ill? “I fear it is unstoppable and uncontrollable,” I wrote Miller. “Unstoppable, yes,” he agreed, “but not uncontrollable. We need responsive and responsible regulation. And on this point, I cannot overemphasize how urgent it is to have the right technologists at the tables of policy formulation.” My sense is that he should be one of them. BONUS – Humor from Our Neighbor to the North America: Have You Lost Your Mind?