ChatGPT 
Summary
Key takeaways
- The article curates more than 40 AI podcasts and says every show on the list was verified as active in Q1 2026, with at least one episode published between January 1 and March 31, 2026.
- The shortlist is built around role-fit, not generic popularity. The article emphasizes that the most useful podcasts in 2026 are the ones that help engineers ship systems, understand research, govern deployments, or translate frontier AI into business decisions.
- The 10-show essential shortlist includes Latent Space, Dwarkesh Podcast, Machine Learning Street Talk, The AI Daily Brief, The TWIML AI Podcast, Practical AI, The Cognitive Revolution, No Priors, Interconnects, and The Pragmatic Engineer.
- Latent Space is positioned as the most important podcast for AI engineers building production systems, especially around agents, infrastructure, multimodality, and real deployment questions.
- Machine Learning Street Talk is framed as one of the most technically rigorous options for people who want depth on papers, mechanisms, and research debates rather than simplified summaries.
- Dwarkesh Podcast is presented as a high-signal source for long-form conversations with frontier builders, making it especially useful for people tracking where top labs and researchers are headed.
- The article separates podcasts into clear groups: technical and research-grade AI, AI engineering and applied AI for builders, MLOps and ML platforms, and AI strategy for leaders.
- For engineering leaders, the article argues that podcasts have become one of the fastest ways to absorb primary-source reasoning from researchers, founders, and operators before the same ideas fully appear in papers or mainstream reporting.
- The article repeatedly favors depth, currency, and practical usefulness over hype, and says many generic “AI for business” explainers that felt important in 2024 are no longer the most useful in 2026.
- The overall recommendation is to subscribe by role and need, not to chase one universal “best” show for everyone.
When this applies
This applies when an engineer, CTO, product leader, platform lead, or AI practitioner wants a curated listening list that is aligned with actual work rather than broad consumer interest. It is especially useful when the goal is to follow frontier labs, understand technical papers, learn how AI systems are built and deployed, stay current on AI news, or translate rapid AI changes into team and business decisions. It also applies when someone wants to build a small, high-signal podcast stack instead of browsing randomly across podcast platforms.
When this does not apply
This does not apply as directly when someone needs a hands-on tutorial, vendor comparison, implementation guide, or detailed written benchmark instead of ongoing audio learning. It is also less useful for people who want only short beginner-friendly explainers, because the article strongly prioritizes expert depth, research context, builder relevance, and role-specific value. If the need is strictly for one narrow topic, such as only MLOps or only AI policy, a specialized list may be more efficient than the broader catalog in this article.
Checklist
- Decide whether you need podcasts for research, engineering, MLOps, leadership, or daily news.
- Start with the essential shortlist before exploring the full catalog.
- If you build AI systems in production, prioritize Latent Space.
- If you want frontier-lab interviews and long-form thinking, prioritize Dwarkesh Podcast.
- If you want technical paper-level depth, prioritize Machine Learning Street Talk.
- If you want short daily updates, add The AI Daily Brief.
- If you want applied AI and deployment-oriented discussion, add Practical AI.
- If you lead teams or translate AI into business strategy, add The Cognitive Revolution.
- If you care about open models, RLHF, and post-training, add Interconnects.
- If you run platform or reliability-heavy ML systems, review the MLOps podcast section separately.
- Prefer podcasts that match your actual role instead of subscribing to everything.
- Remove dormant or low-signal shows from your rotation regularly.
- Balance one daily news show with one deep technical show and one strategy show.
- Use podcast listening to track how fast the field is moving, not just to collect opinions.
- Build a listening stack that helps you make better technical or leadership decisions week to week.
Common pitfalls
- Subscribing based on popularity instead of role relevance.
- Listening only to daily AI news and missing deeper technical context.
- Choosing broad “AI for everyone” content when your work needs engineering depth.
- Treating all AI podcasts as interchangeable.
- Following too many shows at once and getting low retention from all of them.
- Ignoring MLOps and production-system podcasts if your team actually ships models.
- Listening only to frontier hype and skipping operator-focused deployment discussions.
- Assuming a strong podcast for founders is automatically the best one for engineers.
- Staying with dormant or lower-signal shows out of habit.
- Looking for one universal best podcast instead of building a focused mix by need.
The best AI technology podcasts in 2026 for engineers and CTOs are Latent Space (AI engineering, weekly), Dwarkesh Podcast (frontier-lab interviews), Machine Learning Street Talk (technical depth), The AI Daily Brief (daily news), and The Cognitive Revolution (strategy for leaders). The full catalog below covers 40+ shows grouped by role — every one verified active in Q1 2026.
If you lead an engineering team, build AI systems in Python, or own a CTO-adjacent role, this article is the working list we use internally at Uvik Software and hand to clients when they ask what to listen to.
Every podcast on this list was verified active in Q1 2026 — meaning it published at least one episode between January 1 and March 31, 2026 at the time of research. Dormant shows, pivoted shows, and hype-heavy shows were cut. What remains is a list we would actually recommend.
Essential shortlist: the 10 AI technology podcasts that matter most in 2026
If you are starting from zero, subscribe to these ten. The rest of the article is an expansion of this shortlist for specific roles and interests.
| # | Podcast | Host(s) | Best for | Frequency |
| 1 | Latent Space | swyx, Alessio Fanelli | AI engineers shipping to production | Weekly |
| 2 | Dwarkesh Podcast | Dwarkesh Patel | Frontier-lab interviews, long-form | Weekly |
| 3 | Machine Learning Street Talk | Tim Scarfe, Keith Duggar | Technical rigor on papers and mechanisms | Fortnightly |
| 4 | The AI Daily Brief | Nathaniel Whittemore (NLW) | Daily AI news at commute length | Daily |
| 5 | The TWIML AI Podcast | Sam Charrington | Research interviews since 2016 | Weekly |
| 6 | Practical AI | Chris Benson, Daniel Whitenack | Applied AI, MLOps, real deployments | Weekly |
| 7 | The Cognitive Revolution | Nathan Labenz | CTOs translating AI to business | Weekly |
| 8 | No Priors | Sarah Guo, Elad Gil | AI founder and investor lens | Weekly |
| 9 | Interconnects | Nathan Lambert | Open models, RLHF, post-training | Weekly |
| 10 | The Pragmatic Engineer | Gergely Orosz | AI × software engineering at Big Tech | Weekly |
Why AI podcasts are different in 2026
Audio is the format where frontier researchers, founders, and operators speak most candidly — often months before the same thinking surfaces in mainstream press or published papers. Agent capability has been doubling roughly every 70 days over the last 12 months, compared to every seven months in 2023–2024. Written media cannot keep pace. The practical consequence for engineering leaders: podcast listening is now one of the fastest ways to absorb primary-source reasoning from the people actually building the systems.
The list below reflects that shift. Shows that were essential in 2024 — generic “AI for business” explainers, basic LLM primers — have been displaced. What matters in 2026 is depth, currency, and role-fit: whether the show helps you ship better agents, reason about a paper, govern a deployment, or translate frontier research to a board.
Technical and research-grade AI
Deep, long-form, technical. When you want to actually understand a paper, a mechanism, or a scaling law debate, not read a summary of one.
Machine Learning Street Talk (MLST)
Hosted by Tim Scarfe and Keith Duggar. Arguably the most technically rigorous AI podcast in the English-speaking world. Current affairs in AI, cognitive science, neuroscience, philosophy of mind. Running since April 2020, latest episode February 2026, roughly fortnightly. Recent episodes: Jeremy Howard (fast.ai) on AI-assisted coding, Max Bennett on 600 million years of brain evolution, Robert Lange (Sakana AI) on Shinka Evolve. No hype, hand-picked technical guests.
Listen: mlst.ai · Apple Podcasts · Spotify
Dwarkesh Podcast
Hosted by Dwarkesh Patel. Deeply researched interviews with the people building frontier AI. Long, meticulous, high signal. Recent guests include Jensen Huang (NVIDIA), Dylan Patel (SemiAnalysis) on compute bottlenecks, and Terry Tao on AI × mathematics. Weekly.
Listen: dwarkesh.com/podcast · Apple Podcasts · Spotify
Lex Fridman Podcast
Hosted by Lex Fridman. Not a pure AI show, but the AI episodes are landmark. Episode #490 (January 31, 2026) with Nathan Lambert and Sebastian Raschka — a four-hour state-of-AI-in-2026 walkthrough covering LLMs, coding, scaling laws, agents, GPUs, and AGI timelines — is a near-canonical reference. Long-form, multi-hour, philosophical.
Listen: lexfridman.com/podcast
The TWIML AI Podcast
Hosted by Sam Charrington. Formerly This Week in Machine Learning and AI. Running since 2016, 700+ episodes, a go-to for ML researchers, data scientists, and engineering leaders. Weekly interviews. Recent: Sebastian Raschka on the 2025-to-2026 LLM landscape shift from raw scaling to reasoning-focused post-training.
Listen: twimlai.com
The Gradient Podcast
Hosted by Daniel Bashir. Produced by The Gradient, a nonprofit run by grad students and researchers. Deeply researched, technical interviews. Lives between arXiv and mainstream reporting — paper authors walking through methodology on their own work. Biweekly.
Listen: thegradientpub.substack.com · Apple Podcasts · Spotify
Interconnects
Hosted by Nathan Lambert, post-training lead at Ai2 and author of the RLHF Book. Audio essays and interviews from inside the frontier AI labs. The best single feed for open-model ecosystem news and RLHF/post-training. Weekly.
Listen: interconnects.ai/podcast · Apple Podcasts · Spotify
Google DeepMind: The Podcast
Hosted by Hannah Fry. Behind-the-scenes from DeepMind itself: AlphaFold, Genie 3 world models, robotics, nature/ecology AI, cybersecurity. Season 2 through 2025, returning in 2026. No hype. Fortnightly.
Listen: deepmind.google/the-podcast · Apple Podcasts
Deep Papers
Produced by Arize AI founders and engineers. Paper-reading podcast on today’s most important AI research. Recent: IBM’s CUGA enterprise agent, Meta’s ARE and Gaia2 frameworks, and hallucination-bounded LLM evaluation work. Semi-monthly.
Listen: Apple Podcasts · Spotify
AI engineering and applied AI for builders
For people shipping AI into production — agents, RAG, evals, inference, infrastructure. If you write code that calls models or deploys them, this is your tier.
Latent Space: The AI Engineer Podcast
Hosted by swyx (Shawn Wang) and Alessio Fanelli. The definitive AI engineer podcast. Interviews with engineers from OpenAI, Anthropic, Databricks, Meta, Sierra, Modular, Answer.ai. Technical depth on foundation models, agents, infra, code generation, multimodality. 175+ episodes, weekly. The team also runs the AI Engineer Summit conferences.
Listen: latent.space/podcast · Apple Podcasts
Practical AI
Hosted by Chris Benson and Daniel Whitenack, part of the Changelog network. AI that actually ships: MLOps, LLMs, edge AI, real deployments. Running since 2018. Recent 2026 episodes: what mattered in AI in 2025 and what to expect in 2026; AI-driven document processing beyond OCR; Ben Buchanan on US–China AI governance. Weekly.
Listen: practicalai.fm · changelog.com/practicalai
How AI Is Built
Hosted by Nicolay Gerold. “Real engineers. Real deployments. Zero hype.” Interviews engineers who actually put AI in production. Recent guests: Charity Majors (Honeycomb) on the observability cost crisis, Kieran Klaassen on building Cora solo with AI agents, Samuel Colvin (Pydantic), Pete Warden on offline AI, Brandon Smith (Chroma) on chunking. Weekly.
Listen: Apple Podcasts · Spotify · YouTube
Vanishing Gradients
Hosted by Hugo Bowne-Anderson. Long-form conversations on agents, evals, multimodal systems, data infrastructure. Guests include Jeremy Howard (fast.ai), Hamel Husain (Parlance Labs), Shreya Shankar (UC Berkeley), Wes McKinney (creator of pandas), Samuel Colvin (Pydantic). Fortnightly.
Listen: hugobowne.substack.com/podcast · Apple Podcasts · Spotify
High Signal
Hosted by Hugo Bowne-Anderson, produced by Delphina. Practical insights for data scientists and AI engineers: context engineering, agent harnesses, AI in production. Twice monthly.
Listen: high-signal.delphina.ai
Gradient Dissent
Produced by Weights & Biases, originally hosted by Lukas Biewald, now Caryn Marooney. Machine learning in production at companies like NVIDIA, Meta, Google, Lyft, OpenAI. Long backlog.
Listen: wandb.ai/site/resources/podcast
The Robot Brains Podcast
Hosted by Pieter Abbeel (UC Berkeley, Covariant co-founder). Robotics and embodied AI — interviews with people building conscious computers, mindful machines, rational robots.
Listen: therobotbrains.ai
The AI Podcast (NVIDIA)
Hosted by Noah Kravitz. Biweekly, real-world AI applications across industries from climate modeling to cancer detection. 3.4M+ listeners. Not coding tips, but breadth and inspiration.
Listen: blogs.nvidia.com/ai-podcast
MLOps, ML platforms, and production systems
For platform teams, ML infrastructure engineers, and anyone responsible for keeping AI running at 3 a.m. This is the most under-listened category relative to its impact on production reliability — if you lead a platform team, prioritize it.
MLOps Community Podcast
Hosted by Demetrios Brinkmann. The podcast of the 9,500-strong MLOps Community. Relaxed conversations about getting AI into production: agentic, traditional ML, LLMs, whatever works. Weekly. Recent 2026 episodes: Mihail Eric (Head of AI at Monaco, Stanford CS lecturer) on the modern software engineer; Rob Ennals (Uber) on working with teams of agents; Vincent Warmerdam on marimo/Molab and LLM-driven notebooks.
Listen: podcast.mlops.community
ML Platform Podcast
Hosted by Piotr Niedźwiedź and Aurimas Griciūnas (neptune.ai). ML platform design choices, stack components, and real-world MLOps challenges discussed with ML platform engineers from production teams. Technical depth on internal platforms.
Pipeline Conversations
Produced by ZenML. Fortnightly discussions on ML, deep learning, and AI with a particular focus on MLOps and production model deployment. Interviews with industry leaders and technology professionals.
The MLOps Podcast
Hosted by Dean Pleban (DagsHub). Each episode features a top data science or ML practitioner sharing best practices for getting models to production.
Listen: YouTube playlist
AI strategy and enterprise AI for leaders
The shows that bridge “what the tech does” and “what the business does with it.” Useful for CTOs and engineering leaders who have to translate AI capability into organizational outcomes.
The Cognitive Revolution
Hosted by Nathan Labenz and Erik Torenberg. Interviews with the builders on the edge of AI and their implications. Labenz is a former AI-company founder with a philosophy background; thoughtful and analytical. Biweekly, effectively weekly with bonus episodes. Recent: Steve Newman (Writely founder) on personal AI toolkits and vibe-coding; AI in the AM roundups.
Listen: cognitiverevolution.ai · Apple Podcasts · Spotify
Eye on A.I.
Hosted by Craig S. Smith, former NYT correspondent. Biweekly interviews that put incremental AI advances into broader context. 320+ episodes since 2018. Recent: Sergey Levine (Physical Intelligence) on robot foundation models; Dan Faulkner (SmartBear) on AI coding failure modes; Amanda Luther (BCG) on the widening AI value gap.
Listen: eye-on.ai/podcast-archive · Apple Podcasts
Me, Myself, and AI
Hosted by Sam Ransbotham, MIT Sloan Management Review. What separates AI success from AI hype. Leaders from Hugging Face, YouTube, Cisco, ServiceNow, Sony, Wendy’s, OpenAI, and the US Department of Labor. Especially good for understanding how enterprises actually deploy and govern AI. Semi-monthly.
Listen: sloanreview.mit.edu/audio-series/me-myself-and-ai
Azeem Azhar’s Exponential View
Hosted by Azeem Azhar. How AI and other exponential technologies are reshaping business and society. Mix of solo analysis and expert guest interviews. Recent 2026 episodes: 2026 outlook on orchestration and AI bubble dynamics; Davos 2026 dispatches; Jared Kaplan (Anthropic) on Claude’s reasoning. Weekly.
Listen: exponentialview.co
Hard Fork
Hosted by Kevin Roose (NYT) and Casey Newton (Platformer). The award-winning NYT tech podcast with a heavy AI focus. Weekly, Fridays. Less technical than others here, more tuned to news and culture — but useful for CTOs who need to understand how AI is landing in the public and policy spheres. Recent: Project Glasswing and Anthropic’s cyber defense initiative; AI job-loss signals at Atlassian, Block, Meta.
Listen: Apple Podcasts
Pioneers of AI
Hosted by Rana el Kaliouby, AI scientist, investor, Affectiva co-founder. Conversations with creators, critics, and thinkers behind AI, framed around how it is changing daily life and the enterprise.
AI founder and investor shows
Where the capital, talent, and strategy conversations happen. Useful for CTOs thinking about competitive dynamics, platform risk, and the state of the market.
No Priors
Hosted by Sarah Guo (Conviction) and Elad Gil. The AI founder/investor podcast. Guests from frontier labs, AI-native startups, and the biggest incumbents. Recent 2026 episodes: Simon Last (Notion) on custom AI agents and coding-agent-authored integrations; Jensen Huang (NVIDIA) on the state of AI; Sridhar Ramaswamy (Snowflake Intelligence); the Trump-era Department of War generative AI rollout. Weekly.
Listen: no-priors.com · Apple Podcasts · YouTube
Training Data
Hosted by Sonya Huang, Pat Grady, Stephanie Zhan, and others at Sequoia Capital. Conversations with AI builders and researchers. 75+ episodes, weekly. Recent: Harrison Chase (LangChain) on long-horizon agents and context engineering; Anna Goldie and Azalia Mirhoseini on AlphaChip; OpenAI’s Sora 2 team on generative video; Eric Ho (Goodfire) on mechanistic interpretability.
Listen: sequoiacap.com/series/training-data · Apple Podcasts · Spotify
AI + a16z
Produced by Andreessen Horowitz. Discussions with leading AI engineers, founders, and a16z general partners on where the technology is heading — from Replit’s growth story to GitHub’s Scott Chacon on reinventing Git for AI agents. Weekly.
Listen: a16z.com/podcasts/ai-a16z
Unsupervised Learning
Hosted by Jacob Effron, Patrick Chase, Jordan Segall, and Erica Brescia (Redpoint Ventures). The sharpest minds in AI on what is real today versus what is coming. Guests include Noam Brown (OpenAI), Sholto Douglas (Anthropic), Ari Morcos (Datalogy AI), Max Jungestål (Legora). Biweekly.
Listen: Apple Podcasts
Generative Now
Hosted by Michael Mignano, Lightspeed. Stories, strategies, and insights from the builders of today’s AI companies — Macroscope, Wispr Flow, Legora, and more. Weekly.
Listen: Apple Podcasts
Possible
Hosted by Reid Hoffman and Aria Finger. Hoffman’s podcast exploring what is possible with AI — optimistic and pragmatic, with guests across technology, policy, and business.
Daily and weekly AI news
Keep up with the firehose without losing your evenings.
The AI Daily Brief
Hosted by Nathaniel Whittemore (NLW). Daily news analysis show. Short enough for a commute, dense enough to catch you up. Covers model releases, enterprise AI deployments, policy, and industry news. Semi-weekly to daily. Recent focus: Anthropic Opus 4.7 and OpenAI Codex launches; Stanford AI Index and PwC data on the widening AI value gap; the emerging monothread workflow pattern.
Listen: Apple Podcasts
Last Week in AI
Hosted by Andrey Kurenkov (Stanford/DeepMind circles) and Jeremie Harris. Biweekly roundup of the most important AI news, research, product releases, and policy moves — with analyst commentary, not just headlines. A staple.
Listen: Apple Podcasts
AI safety, policy, and frontier labs
For engineering leaders who want to understand where AI governance, alignment, and risk are actually going.
80,000 Hours Podcast
Hosted by Rob Wiblin, Luisa Rodriguez, Zershaaneh Qureshi. Long-form interviews on AI safety, policy, biosecurity, governance. Recent 2026 episodes: Rob Wiblin on Claude Mythos System Card findings (April 2026); Richard Moulange on AI biology tools; Samuel Charap on Ukraine/Russia and nuclear policy.
Listen: 80000hours.org/podcast
ChinaTalk
Hosted by Jordan Schneider. US–China tech, AI competition, chips, and industrial policy with frequent AI deep-dives. Cross-posts regularly with Interconnects (the Overfit series with Nathan Lambert). Essential for geopolitical context.
Listen: chinatalk.media
Data engineering × AI intersection
Data Engineering Podcast
Hosted by Tobias Macey. 500+ episodes on data infrastructure, increasingly covering the AI × data engineering overlap: agentic AI in pipelines, 10-to-50x productivity shifts, self-improving AI systems. Weekly. Recent 2026 episodes: Gleb Mezhanskiy (Datafold) on AI-first data engineering; Rowan Cockett on reproducible science data systems.
Listen: dataengineeringpodcast.com
Data Skeptic
Hosted by Kyle Polich. Running since May 2014, 200+ episodes. Critical thinking and the scientific method applied to data science, statistics, ML, and AI claims. Weekly. A long-running antidote to hype.
Super Data Science
Hosted by Jon Krohn. Weekly interviews covering ML, AI, data careers, and tools. Broader audience, but technical depth when the guest warrants it.
Listen: superdatascience.com/podcast
Python and general engineering with heavy AI coverage
Not AI-first podcasts, but consistently high-signal on AI engineering topics. Worth keeping in rotation if you are shipping with Python or leading an engineering team.
The Pragmatic Engineer
Hosted by Gergely Orosz. Software engineering at Big Tech and startups, from the inside. Launched September 2024, 55+ episodes, weekly, 10M+ downloads by end of 2025. AI coverage dominates: Claude Code, Cursor, vibe coding, MCP, AI-fueled interview cheating, how AI is reshaping engineering hiring. Recent 2026 episodes: Steve Yegge on Gas Town and the future of software development; Grady Booch on the third golden age of software engineering.
Listen: newsletter.pragmaticengineer.com/podcast
Talk Python To Me
Hosted by Michael Kennedy. Weekly, 544+ episodes. Increasingly AI-focused in 2025–2026: Samuel Colvin on Monty (Rust-based sandboxed interpreter for LLM-generated code); Sydney Runkle (LangChain) on Deep Agents and agent harnesses; MCP, Pydantic AI, AI-native Python development.
Listen: talkpython.fm
Python Bytes
Hosted by Michael Kennedy and Brian Okken. Weekly 15-to-20-minute Python headlines roundup. 2026 coverage: Simon Willison’s agentic engineering patterns; Raw+DC as the ORM pattern of 2026 for agent-friendly code; uv python upgrade; PEP 686. Short, high signal.
Listen: pythonbytes.fm
Listening stacks by role
The backlog-guilt trap is real. Subscribing to twenty shows is a commitment you will not keep. Pick three or four that match your current reality, then add or remove as quarters pass. Below are role-based starter stacks drawn from this catalog.
For a research-leaning engineering leader
Dwarkesh Podcast + Machine Learning Street Talk + Interconnects. Roughly five to seven hours per week. You will finish the quarter with a genuine grasp of frontier lab direction, RLHF/post-training, and where scaling is hitting walls.
For an AI engineer shipping to production
Latent Space + How AI Is Built + MLOps Community Podcast. Four to six hours per week. Specifically tuned to the realities of evals, agents, context engineering, and the operational failure modes you will actually hit.
For a CTO translating AI to the business
Training Data + The Cognitive Revolution + The AI Daily Brief. Three to five hours per week. Blends founder/investor framing, builder interviews, and daily news — the three inputs a CTO needs to make decisions faster than the org is asking for them.
For a Python-first engineering organization
The Pragmatic Engineer + Talk Python + Practical AI. Four to five hours per week. Covers Python ecosystem shifts, AI-native Python patterns, and applied AI deployment — the practical stack for a Python shop.
For a data-leaning platform team
Data Engineering Podcast + Vanishing Gradients + TWIML AI. Five to seven hours per week. Covers pipelines, agents, evals, and the research signal that feeds platform roadmap decisions.
For governance, safety, and policy focus
80,000 Hours + Interconnects + ChinaTalk. Four to six hours per week. If you are on the board, advising one, or thinking about geopolitical exposure, this is the stack.
Practical listening tip. 1.5x speed works for conversational episodes; drop to 1x for technical discussions involving code, math, or architecture. If an episode has been sitting in your queue untouched for two weeks, delete it — you will not go back.
Episode length and frequency
A quick reference for matching shows to the time you actually have.
| Podcast | Avg length | Frequency | Format |
| Lex Fridman Podcast | 2–4 hrs | Weekly | Interview |
| Dwarkesh Podcast | 60–120 min | Weekly | Interview |
| Machine Learning Street Talk | 90–150 min | Fortnightly | Technical debate |
| TWIML AI | ~60 min | Weekly | Research interview |
| The Gradient | 60–90 min | Biweekly | Paper deep-dive |
| Interconnects | 30–60 min | Weekly | Essay + interview |
| Latent Space | 75 min | Weekly | Engineering interview |
| Practical AI | 50–70 min | Weekly | Applied AI |
| How AI Is Built | 45–60 min | Weekly | Engineering interview |
| Vanishing Gradients | 60–90 min | Fortnightly | Long-form |
| MLOps Community | 45–75 min | Weekly | Practitioner interview |
| The Cognitive Revolution | 90–120 min | Weekly | Executive interview |
| Eye on A.I. | 30–50 min | Biweekly | Journalist interview |
| No Priors | 35–50 min | Weekly | Founder/investor |
| Training Data | 45–60 min | Weekly | Sequoia-hosted |
| The AI Daily Brief | 15–30 min | Daily | News analysis |
| Last Week in AI | 60–90 min | Biweekly | News roundup |
| 80,000 Hours | 90–180 min | Weekly | Policy interview |
| The Pragmatic Engineer | 60–90 min | Weekly | Engineering interview |
| Talk Python To Me | 60–75 min | Weekly | Python + AI |
| Python Bytes | 15–20 min | Weekly | Python headlines |
Where to listen
| App | Standout feature |
| Apple Podcasts | Reliable discovery and recommendations |
| Spotify | Largest catalog; expanding video podcast support |
| YouTube | Full video for Lex Fridman, Dwarkesh, MLST, All-In |
| Overcast | Smart Speed silence trimming (iOS) |
| Pocket Casts | Best cross-platform sync |
Most shows on this list distribute across all major platforms. Video-first shows (Lex Fridman, Dwarkesh, MLST) are increasingly optimized for YouTube first — and the transcripts published alongside episodes are useful when you want to search or re-read a segment instead of re-listening.
Methodology
This list was compiled in April 2026 from three inputs: feed-level verification that each show published at least one episode in Q1 2026; review of publicly available rankings, aggregator lists, and community recommendations across ML engineering, data science, and AI leadership audiences; and direct listening assessment against a consistent rubric of technical depth, guest quality, publication consistency, and specificity of audience-fit.
Inclusion criteria. Shows were included if they met all of the following: (1) English-language and actively publishing in Q1 2026; (2) AI or AI-adjacent as a primary focus, or reliably heavy AI coverage if the show is a general engineering podcast; (3) signal-to-noise test — most recent four episodes demonstrated either technical depth, frontier-lab access, or durable applied-AI insight, rather than hype, repackaged press releases, or generic business commentary; (4) distinct audience-fit — the show serves a specific engineering or leadership use case that at least one of the nine category taxonomies captures.
Exclusion criteria. Shows were excluded if they were dormant, pivoted away from AI in 2025, failed the signal-to-noise test on recent episodes, or duplicated a better show in the same category. The list is deliberately not exhaustive — it is the 40+ podcasts we would actually recommend to a senior engineer or CTO, not a dump of every AI podcast that exists.
Publication frequencies, episode lengths, and URLs are current as of the April 2026 research date. If a link has changed, search the show title on the podcast player of your choice.
About this resource
This list is maintained by Uvik Software, a Python-first senior-engineering partner that has shipped AI and data engineering systems for companies across the US, UK, DACH, Nordics, and Benelux since 2015. If you are building with Python and AI and want a senior team embedded with yours, get in touch.
If you found this useful, share it with a colleague or link it from your own resources — attribution appreciated, not required. We update this list quarterly as shows launch, pivot, or go dormant; the next revision is planned for July 2026.
Frequently Asked Questions
What is the best AI podcast for engineers in 2026?
For engineers shipping production AI, Latent Space: The AI Engineer Podcast is the consensus pick. It is weekly, 75 minutes on average, and interviews senior engineers at OpenAI, Anthropic, Databricks, Meta, Sierra, and similar companies on foundation models, agents, infrastructure, code generation, and multimodality. How AI Is Built and Practical AI are the strongest complements.
What is the best AI podcast for CTOs and engineering leaders?
The Cognitive Revolution is the most-cited choice for CTOs in 2026. Host Nathan Labenz is a former AI-company founder with a philosophy background, and episodes bridge frontier capability and organizational implication in a way generic business-AI shows do not. Pair it with Training Data (Sequoia Capital) for the investor and strategy lens, and The AI Daily Brief for daily news compression.
What AI podcasts do ML researchers actually listen to?
Machine Learning Street Talk, The TWIML AI Podcast, The Gradient, and Interconnects. These are the four podcasts consistently cited by working ML researchers as worth their time, with MLST widely regarded as the most technically rigorous show in the English-speaking AI world.
How many AI podcasts should I subscribe to?
Three to five at most. Beyond that, queue guilt outpaces listening, and the quality of attention collapses. Pick a stack that matches your current role, keep a daily news show for compression, and rotate quarterly. A podcast that has sat unlistened for two weeks should be deleted, not saved for later.
Are there good AI podcasts under 30 minutes?
Yes. The AI Daily Brief (15-to-30 minutes, daily) and Python Bytes (15-to-20 minutes, weekly) are the strongest short-format options for AI and Python-adjacent coverage respectively. Eye on A.I. occasionally runs under 30 minutes as well.
Which AI podcasts cover MLOps and production ML?
MLOps Community Podcast (Demetrios Brinkmann) is the primary, with the widest guest range and strongest community tie. ML Platform Podcast (neptune.ai), Pipeline Conversations (ZenML), and The MLOps Podcast (DagsHub) are more specialized picks — each is run by a vendor, so weight accordingly, but the technical content holds up.
What is the best AI podcast for Python developers?
Talk Python To Me (Michael Kennedy) leads for Python-specific AI coverage — 544+ episodes, weekly, with increasingly deep coverage of Pydantic AI, LangChain, MCP, and AI-native Python patterns in 2025–2026. Python Bytes pairs well for short-format Python-ecosystem headlines. For broader AI engineering coverage that is heavily Python in practice, add Practical AI.
Which AI podcasts are worth listening to for frontier lab insight?
Dwarkesh Podcast (for interviews with frontier-lab leaders), Interconnects (for Ai2 / open-model perspective and frontier ecosystem news), Google DeepMind: The Podcast (for DeepMind directly), and Training Data (Sequoia Capital interviews with lab researchers). Lex Fridman Podcast episode #490 (January 2026) with Nathan Lambert and Sebastian Raschka is a near-canonical four-hour state-of-AI reference.
How often are these podcasts updated?
All 40+ shows on this list were verified active in Q1 2026, meaning each published at least one episode between January and March 2026 at the time of research. Publication cadence ranges from daily (The AI Daily Brief) to fortnightly (Machine Learning Street Talk, Google DeepMind).
Which AI podcasts are best for AI safety and policy?
80,000 Hours Podcast is the depth pick — long-form interviews on AI safety, alignment, governance, and adjacent policy topics. ChinaTalk (Jordan Schneider) is essential for US–China AI, chips, and industrial policy context. Interconnects overlaps in both directions with its Overfit cross-series with ChinaTalk.
