


49 episodes
by TWiT · English
The 21st Century began with the rise of the Internet and social media. The next decade will mark the rise of the Intelligent Machines. AI will inhabit all our devices from cars and appliances to smart phones and robots. The Intelligent Machines podcast explores the most exciting revolution humanity has ever seen, filled with promise and peril. More than ever we need to understand what these new devices will bring to our lives and how to make best use of them as the 21st century unfolds. On this show you'll meet the AI pioneers, inventors, and innovators who are about to disrupt every aspect of modern life. You'll learn what's real and what's hype, and you'll come away with a deep understanding of the intelligent future that awaits us all. You can join Club TWiT for $10 per month and get ad-free audio and video feeds for all our shows plus everything else the club offers...or get just this podcast ad-free for $5 per month. New episodes every Wednesday.




Post-mortem on Jensen Huang’s departure from NVIDIA and what it may mean for the company; discussion of AI model trends and AI’s future; and a debate over allegations that NVIDIA chips were diverted…
19 Nov 2026 · 2h 41m · 57 chaptersLocal AI performance engineering—Ash Hart’s TensorFlow/Tensorfold approach to speeding up running large language models on home hardware, plus the practical “secret sauce” (and tradeoffs) of…
12 Nov 2026 · 2h 42m · 56 chapters
Episode zero introduces “Hands on AI,” a short, practical show meant to turn AI confusion into AI confidence by treating AI as a tool rather than hype or fear.
11 Nov 2026 · 2m · 1 chapterThe episode focuses on backlash to AI-related data centers in the 2026 US elections, how that backlash may shape broader AI policy, and the growing use of LLM chatbots in voter outreach.
5 Nov 2026 · 2h 30m · 43 chaptersThe episode focuses on the AI industry’s financial and strategic fault lines, centered on Anthropic’s IPO filing and what it implies about enterprise lock-in, compute costs, and “safe AI” branding.
29 Oct 2026 · 2h 20m · 42 chaptersThe episode discusses recent AI model and hardware advances (local/open-weight models, NVIDIA’s Numatron and DGX/“container data centers”), the rise of consumer “persistent” AI assistants (Meta’s…
22 Oct 2026 · 2h 26m · 44 chaptersGitHub’s “agent revolution” for software development, focusing on how AI coding agents change GitHub usage, DevOps workflows, and open-source contribution dynamics.
15 Oct 2026 · 2h 32m · 54 chaptersAgentic AI “digital swarms” that persistently perform tasks on computers, sometimes doing unexpected or abnormal actions due to permissions, shared context, and emergent behavior.
8 Oct 2026 · 3h 14m · 75 chaptersLocal (“at-home”) AI using open-weight models, and how one engineer orchestrates a team of AI agents for work and education; plus discussion of frontier-model economics and OpenAI’s Dev Day…
1 Oct 2026 · 2h 53m · 64 chaptersThe episode argues that large language models (LLMs) are fundamentally weak at complex/combinatorial reasoning and that many “LLM-only” deployments fail because they can’t be constrained or made…
24 Sep 2026 · 2h 37m · 52 chaptersThe episode debates “AI doom” risk and whether frontier AI labs can/should prevent catastrophe, then pivots to practical “local AI” building.
17 Sep 2026 · 2h 42m · 62 chaptersApple’s “Siri AI” overhaul and how it will work with iPhone apps via App Intents/App Intense (built on the Shortcut/Workflow lineage).
10 Sep 2026 · 2h 21m · 54 chaptersThe episode focuses on “agentic” AI on phones—specifically Div Garg’s AGI Inc vision that apps will fade away as an “agentic OS” can understand and control whatever is on your device.
3 Sep 2026 · 2h 44m · 72 chaptersLocal AI hardware vs frontier models; Apple’s new Mac Studio (M5 Ultra) for running local models; how NVIDIA “DGX Sparks” compare (CUDA, unified memory, power use, clustering); and a separate segment…
27 Aug 2026 · 2h 22m · 50 chaptersDan O’Dowd argues that critical infrastructure software must be engineered like “weapons of mass destruction” because failures or hacks can cause mass starvation and long-term societal collapse.
20 Aug 2026 · 2h 36m · 65 chaptersThe episode focuses on AI “watermarks” for text (Anthropic’s plan to embed imperceptible identifiers in generated text) and, in the main guest segment, how one journalist is using AI to automate and…
13 Aug 2026 · 2h 35m · 64 chaptersThe episode focuses on Persona Shield, a “permission layer for AI identity” that aims to stop or control deepfakes and monetization abuse of a person’s likeness.
4 Aug 2026 · 2h 30m · 50 chaptersThe episode argues that AI is ending the usefulness of “news aggregation” and forcing newsrooms to refocus on original reporting, distinctive analysis, and direct reader relationships.
30 Jul 2026 · 2h 9m · 10 chaptersThe episode focuses on recent AI security incidents and the broader competitiveness question “Can American AI compete when China gives it away?” The main case is Hugging Face reporting that an…
23 Jul 2026 · 2h 39m · 59 chaptersRafi Krikorian (Mozilla Foundation CTO) discusses Mozilla’s “State of Open Source AI” report and argues that open (open-weight/open-source) models are increasingly “good enough” for everyday use,…
16 Jul 2026 · 2h 24m · 54 chaptersThe episode centers on robotics and “smart doodads” in everyday life, but the main discussion is Dr.
9 Jul 2026 · 3h 6m · 65 chaptersThe episode focuses on “invisible AI failures” in modern AI assistants—cases where users don’t realize the system is wrong, so the failure leaves no trace.
2 Jul 2026 · 2h 47m · 73 chaptersThe episode centers on Section 230 of the Communications Decency Act and whether it should be changed or narrowed.
25 Jun 2026 · 2h 45m · 66 chaptersThe US government’s export-control action that shut down Anthropic’s “Fable” (and “Mythos”/related capabilities) after an Amazon security report, framed as a “jailbreak,” and the broader AI “power…
18 Jun 2026 · 2h 15m · 43 chapters