The AI Signal & The AI Noise
The AI Signal & The AI Noise” is the daily podcast for anyone curious about artificial intelligence, but who isn’t a developer. We know the AI world can be overwhelming, so our mission is simple: find the most important news every day and explain it clearly. From groundbreaking image generators and new AI music to the latest business and ethical debates – we give you the signal and help you ignore the noise.
Listen on:
Episodes
Friday Sep 19, 2025
Friday Sep 19, 2025
In this episode, Taylor and Morgan discuss a range of AI developments - from Luma AI's Ray3 to Google's Gemini and Alibaba's Tongyi DeepResearch. They also delve into the concept of agentic commerce, AI scheming, and how AI models are becoming increasingly integral in various fields, all while raising crucial questions about potential misuse, privacy concerns, and the need for cautious optimism.
Thursday Sep 18, 2025
Thursday Sep 18, 2025
Join Taylor and Morgan as they discuss the latest in the AI world; Google's Agent Payments Protocol, YouTube's addition of generative AI to Shorts and podcasts, Meta AI's 'MapAnything' technology, and Keplar’s voice AI for market research. The episode also covers fluid benchmarking for large language models by Ai2.
Wednesday Sep 17, 2025
Wednesday Sep 17, 2025
Taylor and Morgan discuss major breakthroughs in AI, from the creation of new benchmarks for robotics and healthcare, to the rise of startups making training simulations for AI agents. They delve into the issues surrounding OpenAI's new restrictions for underage users and unpack Google's multi-billion investment in UK's AI infrastructure. They also explore the use of AI in the creative tools being offered by YouTube.
Tuesday Sep 16, 2025
Tuesday Sep 16, 2025
In this episode, Taylor and Morgan dive into discussions about OpenAI's GPT-5-Codex and its potential impact on developer productivity. They also explore how different AI models are finding their niche in the tech world, NVIDIA's open-source Video Pose Engine, OpenAI's renewed interest in robotics, Meta AI’s MobileLLM-R1 for edge devices, and user-friendly applications that help bring AI closer to the public.
Monday Sep 15, 2025
Monday Sep 15, 2025
Taylor and Morgan dive into the potential risks and revolutionary rewards of AI. They discuss a call for a global AI shutdown due to safety concerns and explore efforts to democratize AI development, highlighting the importance of responsible use and robust evaluation.
Friday Sep 12, 2025
Friday Sep 12, 2025
Taylor and Morgan discuss the week's AI news, highlighting OpenAI's new developer tool - Full MCP support for ChatGPT, leading to potential AI integration into business workflows but raising serious security concerns. They also explore Thinking Machines' research on LLM inconsistencies, the efficiency of mmBERT's multilingual model in natural language processing, and YouTube's multilingual audio tracks for accessibility.
Thursday Sep 11, 2025
Thursday Sep 11, 2025
Taylor and Morgan discuss the latest in Artificial Intelligence, including ByteDance's revolutionary Seedream 4.0, YouTube's multilingual audio feature, and NVIDIA's Universal Deep Research. They also explore the potential of Microsoft's MAI-Voice-1 model, tailored voice technology from ElevenLabs, Baidu's efficient ERNIE-4.5-21B-A3B-Thinking, and Stability AI's generative audio tool. Accentuating both advancement and implications, they delve into how these innovations are shaping the future.
Wednesday Sep 10, 2025
Wednesday Sep 10, 2025
In this engaging discussion, Taylor and Morgan explore the potential of AI in disrupting the hotel booking industry and the concept of personalized travel planning. They also discuss Apple's strategies in AI implementation, the impressive K2 Think model, ParaThinker's parallel thinking approach, and the concept of 'vibe coding' in Model Context Protocol.
Tuesday Sep 09, 2025
Tuesday Sep 09, 2025
Taylor and Morgan discuss the latest in AI news. They delve into the rise of Language Learning Model-run twitter accounts, the tutorial on multi-domain AI web agent development, smarter search versus bigger models, motion's funding for AI agents for small businesses, an open-source memory engine for AI called Memori, reinforcement learning mitigating catastrophic forgetting in AI models, and the practicality of wheels over humanoid designs in robotics.
Monday Sep 08, 2025
Monday Sep 08, 2025
In this discussion, Taylor and Morgan delve into the debate around whether language models actually understand anything. They examine the issues of hallucinations in AI outputs, explore the emerging AI company, Mistral AI, and discuss the innovative developments from Meta's Superintelligence Labs.