Intellectually Curious

TurboQuant: The 3-Bit Breakthrough Making AI Faster and Smaller

Google Research's TurboQuant uses polar quant and Quantized Johnson-Lindenstrauss to shrink the KV cache to roughly 3 bits per value, delivering up to 8x speedups and sixfold memory savings on high-end GPUs without sacrificing accuracy. We unpack how shifting to polar coordinates avoids heavy normalization and how a single sign bit preserves data relationships, enabling faster semantic search and smarter AI tools on standard hardware.Note:  This podcast was AI-generated, and sometimes AI can make mistakes.  Please double-check any critical information.Sponsored by Embersilk LLC

The AI Scientist: Automating the Scientific Life Cycle

We unpack the March 25, 2026 paper that envisions an AI system capable of ideation, experimentation, write-up, and internal peer review to autonomously advance scientific research. Learn how Claude Sonnet 4 writes and tests code, how Semantic Scholar integration checks novelty against decades of literature, and how a dual-agent setup self-critiques to improve quality. We'll also examine real-world evaluation (ICLR 2025) and discuss the implications for future discovery and human–AI collaboration.Note:  This podcast was AI-generated, and sometimes AI can make mistakes.  Please double-check any critical information.Sponsored by Embersilk LLC

Protein Truths and Fiber Focus: A Stanford Reality Check

We cut through the hype around protein bars, powders, and the latest dietary guidelines, using a Stanford Medicine report to explain what our bodies actually need. Learn how muscle growth is sparked by resistance training, why higher protein targets mainly matter for older adults, and why fiber deserves equal attention for a healthy gut. We debunk plant-protein myths and offer practical tips for eating real foods that support both muscles and the microbiome.Note:  This podcast was AI-generated, and sometimes AI can make mistakes.  Please double-check any critical information.Sponsored by Embersilk LLC

AI and the High Temperature Superconductivity Challenge

Could AI become the ultimate research assistant? In this deep dive, we review a study that pits six LLMs against a curated database of 1,726 high-temperature superconductivity papers, using custom retrieval architectures to fight misinformation and conflicting results. We explore why gated, sandboxed AIs outperform general web-searching models, the critical blind spot in visual reasoning, and what this means for future cross-disciplinary scientific breakthroughs.Note:  This podcast was AI-generated, and sometimes AI can make mistakes.  Please double-check any critical information.Sponsored by Embersilk LLC

Black Mass: Turning Spent EV Batteries into a Circular Economy

We dive into how the industry converts dead EV batteries into 'black mass,' a concentrated mix of lithium, nickel, cobalt, and manganese. From safe disassembly and inert shredding to hydrometallurgy that recovers 95–99% of metals with far less energy than smelting, this episode explains the new, sustainable supply chain powering the future of energy tech.Note:  This podcast was AI-generated, and sometimes AI can make mistakes.  Please double-check any critical information.Sponsored by Embersilk LLC