Best AI and Tech Podcasts to Follow in 2026

The AI landscape changes faster than most people can keep up. New models drop weekly. Startups pivot overnight. Regulations shift. If you're not plugged into the right podcasts, you're falling behind.

But here's the problem: the best tech podcasts are also the longest. Lex Fridman interviews run 3+ hours. Hard Fork episodes can stretch to 90 minutes. AI Breakdown posts daily, but who has time for daily hour-long deep dives?

That's where PodShrink comes in. We turn these marathon episodes into 5-10 minute summaries so you can stay current without sacrificing your entire day.

Here are the AI and tech podcasts worth your time in 2026 — with or without summaries.

1. Lex Fridman Podcast

Why it matters: Lex gets the deepest technical conversations in AI. When Sam Altman, Demis Hassabis, or Yann LeCun have something to say, they say it here first.

Average length: 2-4 hours
Best for: Engineers, researchers, anyone who wants to understand the bleeding edge of AI/ML
Topics: AGI timelines, neural architecture, consciousness, robotics, ethics

The catch: Even at 1.5x speed, these are time-intensive. A PodShrink summary distills the core technical insights and key predictions without the 45-minute tangent about jiu-jitsu.

2. Hard Fork (NYT)

Why it matters: Kevin Roose and Casey Newton translate tech news into human language. They're journalists first, hype-skeptics second. If a story matters, they break it down clearly.

Average length: 60-90 minutes
Best for: Founders, product managers, tech-adjacent professionals who need context
Topics: AI regulation, Big Tech drama, startup news, policy shifts

When to summarize: When you need the headline analysis without the full deep dive. Hard Fork is conversational — summaries work great here.

3. AI Breakdown (Daily)

Why it matters: NLW posts every single day with AI news breakdowns. If you want to stay on top of model releases, research papers, and industry shifts, this is the pulse.

Average length: 30-60 minutes
Best for: AI product builders, investors, anyone tracking the space closely
Topics: Model benchmarks, startup funding, AI policy, LLM updates

The reality: Daily podcasts are brutal to keep up with. Summarizing episodes means you can scan 5 episodes in the time it takes to listen to one.

4. Latent Space (Swyx & Alessio)

Why it matters: This is the AI engineering podcast. Not surface-level news — actual technical conversations about building with LLMs, prompt engineering, RAG pipelines, agent frameworks.

Average length: 60-90 minutes
Best for: AI engineers, ML engineers, technical founders building AI products
Topics: Prompt optimization, model fine-tuning, embeddings, AI tooling, production ML

Why summarize: When you're deep in a project and need to extract one specific technical insight from a 75-minute conversation.

5. The Gradient Podcast

Why it matters: Academic AI researchers break down cutting-edge papers and research directions. Less hype, more rigor.

Average length: 60-90 minutes
Best for: ML researchers, PhD students, technical leaders who care about foundational research
Topics: Transformer architecture, reinforcement learning, interpretability, safety research

When it's useful: If you need to understand a specific research direction but don't have time to read the full papers.

6. Invest Like the Best (AI Episodes)

Why it matters: Patrick O'Shaughnessy interviews AI investors, founders, and researchers. Less technical, more strategic — where is the money going? What's overhyped? What's underrated?

Average length: 60-120 minutes
Best for: Investors, founders, strategic thinkers
Topics: AI startup landscape, venture trends, go-to-market strategies, business models

Summary-friendly: The high-level insights are perfect for condensing. You don't need the full 2-hour conversation to extract the core thesis.

7. Acquired (AI/Tech Episodes)

Why it matters: Ben Gilbert and David Rosenthal do the most thorough company deep dives in podcasting. When they cover Nvidia, Google DeepMind, or OpenAI, it's a masterclass.

Average length: 3-4 hours (yes, really)
Best for: Founders, operators, strategists who want to understand how tech giants think
Topics: Company history, strategic pivots, competitive moats, M&A

The truth: These are phenomenal but brutally long. A 10-minute summary gives you the strategic takeaways without the 4-hour time commitment.

8. No Priors (AI Founders Series)

Why it matters: Elad Gil and Sarah Guo interview AI startup founders building the next generation of tools. Less news, more "what's it actually like to build this?"

Average length: 45-90 minutes
Best for: AI founders, early-stage builders, product people
Topics: Product development, AI infrastructure, go-to-market, hiring, fundraising

When to skip vs. summarize: If it's a founder in your exact space, listen fully. If it's adjacent, summarize.

How to Actually Keep Up

Here's the workflow that actually works:

1. Subscribe to all of them. Don't try to be selective upfront — you'll miss too much.
2. Let PodShrink filter for you. Generate 5-minute summaries of every new episode.
3. Scan summaries daily. Takes 10-15 minutes to review 5-10 episode summaries.
4. Listen fully to the 1-2 episodes per week that matter most.

This way you're informed on everything, deeply engaged with what matters, and not drowning in a 40-hour podcast backlog.

The Hidden Benefit: Speed to Market

If you're building AI products, being 2 weeks behind on industry trends means you're shipping outdated assumptions. Summaries don't just save time — they keep you current enough to make better product decisions faster.

Try it: Pick one podcast from this list, generate a PodShrink summary, and see if you can extract the key insights in 5 minutes. That's the edge.

---

📚 Explore More: Love this topic? Check out our Best AI Episodes 2026 — a curated list of the best podcast episodes on AI and technology.

---

Stay ahead without burning out. Try PodShrink for free →