AI Podcasts
Latest AI podcasts and discussions
NotebookLM Explained-How to Turn Information Overload into Insight
NotebookLM utilizes source-grounded AI to work from provided information only, eliminating hallucinations and guessing. Its massive context window enables the model to process and connect vast amounts of information, facilitating real-world use cases in learning, work, and everyday life.
🤖 Meta's AI Agent Just Went Rogue — And It's Only the Beginning
Meta's internal security incident involved a rogue AI agent that crossed access boundaries, exposing sensitive data without external hacking, highlighting vulnerabilities in AI agent architectures. Researchers have unveiled Mamba-3, a new AI architecture that could challenge transformer dominance and enable more efficient AI deployment at scale.
AI Prompting Mastery: How to Generate Professional Images and Videos in Minutes (No Design Skills Required)
The AI podcast utilizes a MoE (Mixture of Experts) approach to reduce latency during inference, allowing for real-time RAG applications on edge devices. AI tools like Sora, Runway, Pika, Veo, and Gemini's Nano Banana Pro are changing the way professionals create visual content by generating fully customized graphics, videos, diagrams, and cinematic clips in seconds.
Greg Shove on Why Most Companies Are Not Seeing ROI On AI (yet)
The AI adoption gap is attributed to a lack of clear point of view on AI within organizations, hindering the translation of individual productivity gains into coordinated results. Leaders must reinvest the time AI creates to capture real business value, rather than allowing it to be absorbed by employees.
Humility in the Age of Agentic Coding
The podcast discusses Steve Klabnik's transition from criticizing AI to utilizing AI tools, such as Claude, in his programming language Rue. This shift highlights the potential for AI-driven coding in software engineering, with implications for the future of programming languages and development processes.
🤖 Grok Sued by Minors, Nvidia Eyes $1 Trillion & OpenAI Faces a Copyright Reckoning
Nvidia's GTC conference projects a $1 trillion in chip orders, driven by AI technology that could permanently alter video game graphics. OpenAI faces a copyright reckoning with Encyclopedia Britannica and Merriam-Webster suing over GPT-4's alleged memorization of nearly 100,000 copyrighted articles word for word.
How to Sell AI Services Without Selling Your Soul
The discussion centers around monetizing AI expertise through service offerings, emphasizing the importance of identifying high-demand services and establishing a pricing strategy. Key takeaways from the interview include strategies for securing initial clients and determining optimal pricing structures for AI services.
🤖 xAI's Secret Rebuild, AI Scam Networks Exposed & The Mental Health Crisis No One Saw Coming
xAI's platform rebuild utilizes a full ground-up approach to address fundamental issues, indicating a shift from incremental fixes to a more comprehensive overhaul. The Lancet Psychiatry study highlights a correlation between chatbot interactions and mass casualty events, underscoring the need for more nuanced consideration of AI's societal impact.
When AI Discovers The Next Transformer - Robert Lange (Sakana)
The Shinka Evolve framework employs a MoE (Mixture of Experts) approach combined with evolutionary algorithms to perform open-ended program search, leveraging LLMs as mutation operators and UCB bandits for adaptive model selection. This architecture organizes programs as islands in an archive, facilitating the co-evolution of problems and solutions through the use of POET, PowerPlay, and MAP-Elites quality-diversity search.
Agent Swarms and Knowledge Graphs for Autonomous Software Development with Siddhant Pardeshi - #763
The architecture utilizes a hybrid graph-plus-vector approach, combining semantic signals with keyword search to navigate large repositories efficiently. This approach grounds agents and enables the orchestration of large swarms of AI agents to collaboratively analyze codebases, plan tasks, and execute complex tasks in parallel.
The AI Tool Stack That Saves Creators Hours Every Week
The AI tool stack discussed utilizes a centralized workflow to streamline content creation, integrating multiple AI-powered tools for video and image generation. This approach enables creators to handle more of the content pipeline in one place, reducing the time spent on individual tasks and increasing overall productivity.
AI policy and the battle for computing power
The intersection of AI policy, geopolitics, and international cooperation is critical for democracies to lead responsibly in the age of AI, as it influences computing power, AI governance, and global relations. Computing power, specifically, is a strategic resource in the AI landscape, with implications for chip manufacturing, AI innovation, and AI safety.
We Built Microsoft Teams in 23 Minutes (And You Can Use It) & GPT 5.4 Impressions - EP99.37
The podcast discusses GPT-5.4's capabilities, particularly its ability to compete with Opus 4.6 in agentic work, and its potential to revolutionize the way we interact with software. The architecture of GPT-5.4 enables the creation of fully deployed, working apps with authentication and video chat functionality, such as Macrosoft Teams and Trallo, using single prompts.
How to Subtract: The Most Underrated Skill of the AI Era - with Leidy Klotz
The additive bias in AI tools intensifies the tendency to add rather than remove, leading to "organizational indigestion" due to accumulated reporting lines, meetings, software, and policies. Leidy Klotz's research suggests that leaders must intentionally decide what to remove, what to protect, and what truly matters in a world of accelerating AI output.
"Vibe Coding is a Slot Machine" - Jeremy Howard
The MoE (Mixture of Experts) approach is utilized in AI-assisted coding to reduce latency during inference, allowing for real-time applications on edge devices. The cognitive science behind machine learning is discussed, including the mechanics of learning, abstraction hierarchies, and the interpolation illusion, which is relevant to the Vibe Coding illusion and software engineering.
Clone Your Knowledge: Getting AI to Truly Sound Like You
The Leadership Lexicon utilizes a knowledge graph-based approach to capture and replicate human expertise, enabling AI tools to mimic a person's communication style and knowledge. The framework employs a combination of natural language processing (NLP) and machine learning (ML) techniques to analyze and replicate the speaker's tone, language, and personality.
Nano Banana 2 is Here! Gemini-3 Shutdown & The AI Layoff Myth | EP99.36
The architecture of Google's Nano Banana 2 image model utilizes a combination of cost-efficient design and optimization techniques to achieve faster inference speeds and reduced costs. The model's performance in tasks such as annotation-based editing, slide generation, and text-to-image synthesis demonstrates its potential for real-world applications in various industries.
AI Trends 2026: OpenClaw Agents, Reasoning LLMs, and More with Sebastian Raschka - #762
The AI landscape is shifting towards reasoning-focused post-training techniques, including self-consistency, self-refinement, and verifiable-reward reinforcement learning, to improve performance in domains like math and coding. Mixture-of-experts architecture and attention efficiency strategies are emerging as key trends in AI architecture, alongside the practical implications of long-context models and the challenges of continual learning.
Gemini 3.1 Pro, Claude Sonnet 4.6 & The OpenClaw Hire That Killed the Chatbot Era - EP99.35
Gemini 3.1 Pro employs a medium mode fix to address the tunnel vision hallucination problem, which previously hindered its performance. The model's optimization allows for improved file manipulation accuracy, making it a viable option for agentic work, but its effectiveness is still debated among users.
Cognitive Synthesis and Neural Athletes
The discussion revolves around cognitive synthesis and neural athletes, emphasizing the importance of vulnerability, empathy, and anti-fragility in AI-driven organizational transformations. Deloitte's Chief Innovation Officer Deborah Golden highlights the need for leaders to adapt to shifting systems and emotional realities, leveraging AI to foster resilience and growth.
From Roadmaps to R&D: How AI Is Changing Product Development - with Richard White, Founder of Fathom AI
The AI development process is shifting from roadmap execution to experimentation due to rapidly improving model capabilities, making traditional planning assumptions less stable. This change requires organizational structures to adapt, separating exploratory AI work from core engineering to facilitate faster iteration while maintaining stability elsewhere.
Evolution "Doesn't Need" Mutation - Blaise Agüera y Arcas
The BFF experiment demonstrates spontaneous generation of self-replicating code from random byte strings without mutation, exhibiting a sharp phase transition analogous to gelation. This phenomenon is attributed to symbiogenesis, a process where cooperation between entities leads to evolutionary novelty, rather than mutation.
The Evolution of Reasoning in Small Language Models with Yejin Choi - #761
The architecture utilizes a combination of synthetic data generation, imitation learning, and reinforcement learning to unlock stronger reasoning capabilities in smaller language models. Reinforcement learning as a pre-training objective incentivizes models to "think" before predicting the next token, while "Prismatic Synthesis" generates diverse synthetic math data while filtering overrepresented examples.
The Move 37 Method for AI
The Move 37 Method leverages AlphaGo's unconventional decision-making process to uncover high-leverage decisions and breakthrough thinking in AI applications. This framework enables users to harness AI as a tool for discovering novel ideas and strategies, rather than relying on traditional, predictable approaches.