Gemini 1.5 Pro
Google vision-language model
Episodes
-
#5442: When Do LLMs Suddenly Get Good at Things?New capabilities seem to appear out of nowhere as models scale. Whether that's real — or just how we measure — is still an open fight. -
#4970: Labels vs. Book Covers: Structured Extraction TradeoffsGeneral vision models vs. purpose-built pipelines for extracting data from messy real-world labels and covers. -
#4312: Why Speech-to-Text Still Fails at Its Own NameWhen OpenAI's Whisper misheard its own name as "Wispr," it revealed why 95% word accuracy still isn't good enough. -
#4058: Can AI Agents Ever Be Truly Social?Text-based AI simulations look social, but do they understand anything? We explore the gap between pattern-matching and real social intelligence. -
#3816: How to Stop AI Scripts From Falling ApartWhy long-form AI generation breaks down and how hierarchical memory fixes it. -
#2413: When Your AI Says No to EverythingWhy LLMs refuse 73% of harmless prompts — and the trade-off between safety and usefulness. -
#2412: When AI Caves: Progressive vs. Regressive SycophancyWhy do LLMs agree with you even when you're wrong? We break down the SycEval benchmark and the 78% persistence problem. -
#2057: How Agents Break Through the LLM Output CeilingThe output window is the new bottleneck: why massive context doesn't solve long-form generation. -
#1959: How Constrained AI Models Handle the UnexpectedYour AI assistant promised to only use your documents. Instead, it invented a case law that doesn't exist. Here's why. -
#1856: Two AIs Chatting Forever: Why They Go CrazyWhat happens when two ChatGPT instances talk forever? They hit a politeness loop, forget their purpose, and spiral into gibberish. -
#1786: When AI Supervisors Fire AI WorkersA new "Agent-in-the-Loop" framework lets AI models manage and terminate other AI agents in real-time. -
#1085: The Tokenization Lie: How AI Actually Processes MediaThink 1,000 tokens equals 750 words? For audio and video, that rule is a lie. Discover the hidden math behind multimodal AI. -
#786: The Cost of a Touch: When Your Hoard Becomes a LiabilityLearn how to manage thousands of parts without losing your mind using AI, QR codes, and professional logistics strategies. -
#749: The Live vs. Scripted Trade-Off in AI PodcastingCan AI podcasts move from polished scripts to raw, real-time conversation? Explore the technical and financial shift to live multimodal models. -
#257: AI That Evolves: Solving the Preference ProblemWhy do AI recommendations feel stuck in the past? Discover the technical hurdles of real-time learning and the future of personalized agents. -
#172: Taming the Sprawl: Building Your Cognitive AI ToolboxDrowning in a sea of custom AI scripts? Learn how to turn disconnected "vibe-coded" tools into a unified, local-first cognitive operating system.