Confirmed that when you interrupt it, it shuts up immediately. Ya know, like a tool that is useful instead of a friend that burns tokens for no reason.
In all honesty I think this is an unfair characterization. I've listened to his podcast for a fair number of years and I think you're exaggerating quite a bit.
Mostly the episodes are specific topics with a guest who's area of research/specialty is the topic and which stay pretty well on-topic.
Most of the AMA episodes are in the area's where Carroll specialize: physics and philosophy of science. It -occasionally- veers off into politics or basketball, but it's nowhere close to 51% politics and basketball. I think the most recent AMA did have more basketball than usual, and admittedly more than I like. And I am not a fan of his politics nor do I consider his political positions especially thoughtful or insightful. But those diversions are more occasional distractions than some majority of the AMA content. And in a 3+ hour Q&A format its not surprising that he's asked about those areas where his regular listeners know that he has a particular interest or position.
z3 is also just so thoroughly optimized that even if your formulation of the constraints is inefficient it is faster. it is a great library that lets you solve pretty complicated DP problems with a few dozen lines of code.
In general that's true, but to reason about boolean circuits like in this challenge we only need a SAT solver. Z3 is just used for it's convenient API.
> The model dynamically navigates the video timeline, loading only the content it needs based on the prompt. Up to 88% more token-efficient and ~7% higher quality on long-form content.
The Eurasian Cave Lion was around there back then and was probably the direct model for the Lion-man sculpture. It was roughly 20% larger than modern African lions. Paleolithic cave paintings (e.g., Chauvet) show the male cave lions lacked the heavy neck manes of modern lions.
reply