Season 2 · Episode 181
Why Your AI Is Finally Stopping to Think
Discover how AI shifted from instant reflexes to deep reflection through inference-time compute and hidden reasoning steps.
My Weird Prompts · Daniel Rosehill
January 6, 202631m 23s
Audio is streamed directly from the publisher (dts.podtrac.com) as published in their RSS feed. Play Podcasts does not host this file. Rights-holders can request removal through the copyright & takedown page.
Show Notes
In this episode of My Weird Prompts, Corn and Herman Poppleberry dive deep into the seismic shift occurring in artificial intelligence: the transition from fast, predictive chatbots to slow, deliberate reasoning models. They explore the engineering behind "inference-time compute scaling," explaining how hidden tokens and "System 2" thinking allow models to catch their own errors before they even reach the user. By breaking down complex concepts like Monte Carlo Tree Search and Process Reward Models, the brothers reveal what happens when you crank an AI's "reasoning level" to the max and why the future of tech depends on an AI's ability to show its work. Whether you're a software engineer or just curious about the data center's rising energy costs, this deep dive explains why the most powerful AI isn't necessarily the biggest, but the one that thinks the longest.