Wait, did Leo just say that solving math was not supposed to be something AI was good at? I love @leo as he has been part of my world for 30 years? (holy hell) but the defensiveness toward any criticism of AI, the industry, etc is getting really old quick. AI is advancing in big ways, but it’s biggest gains in coding and math which actually makes sense, and if it was sold as this from day 1 where we sold it would replace art etc things would be a lot different from a PR standpoint. Texas, yes Texas, just put a stop on data centers until things around power are figured out. Last thing as Leo as part of his rant it does not matter what the “AI boys” say… maybe to Leo it does not which is fair, but it matters when it comes to the public at large and how that impacts the future given the “AI boys” connection to politicians and who/what we all get access to in the future.
edited to add: 1 square of dark chocolate, 7 to 10 almonds, and 1/2 cup of blueberries is a great afternoon snack.
AI was very bad at arithmetic, but that is very different to math, or rather arithmetic is just a small part of math.
But Leo was really annoying today, it was almost as if he was hearing bad words pushed towards his beloved AI, but not actually hearing what was being said. Paris und Jeff were saying that the heads of OpenAI and Anthropic were the ones peddling general AI, and that this was the wrong way to go, that the industry should be making specific models that are good at specific tasks - because they will evolve quicker and give demonstrably better results - and Leo seemed to be ignoring what was being said and saying doom-sayers, who keep changing what AGI actually means, should be ignored, when Jeff was saying, it was the AI leaders themselves that were changing the goal posts.
The show has been through some ups and downs over the years, but Leo’s attitude in recent months has me very close to removing IM from my podcast feed. It used to be Jeff that would annoy me, especially as a German, and his patronising way of telling Leo that he knows how Germans think and what they do. As an ex-pat Brit, who is now a naturlised German citizen, I can say that he was always way off base, at least outside of publishers in Berlin, but even though I respectfully disagreed with Jeff and still respected his opinion on other topics, Leo is getting to the point where he is becoming almost insufferable.
OpenAI’s hacking debacle, Claude publishing malicious code, and AI agents running amok
The fight against deepfake abuse, digital piracy, and protecting online identities with Phillip Shoemaker
Sam Altman’s support for AI regulation and the White House review of AI frameworks
Elon Musk’s prediction that Optimus robots will surpass human surgeons in four years
AI’s expanding cultural footprint, including Billboard hits, Pulitzer disclosures, and Google AI Overviews