The AI conversation, week by week
18 AI-focused podcasts currently tracked18 AI podcasts

Currently tracking18 AI-focused podcasts

  • AI For Humans
  • AI Inside
  • AI Today Podcast
  • Eye on AI
  • Gradient Dissent
  • Interconnects
  • Last Week in AI
  • Latent Space
  • Machine Learning Street Talk
  • Me, Myself, and AI
  • No Priors
  • Practical AI
  • The AI Daily Brief
  • The AI Native Dev
  • The Artificial Intelligence Show
  • The Cognitive Revolution
  • The TWIML AI Podcast
  • Training Data

VAIBE tracks AI-focused podcasts: what changed, and the moments behind it.

Next edition Oct 5–11 · Expected Tuesday, Oct 13 · 10 AM ET

Safety & governance attention cooled while research rebounded

Across 27 episodes from 15 contributing podcasts, safety & governance fell 5.4 points to 8.6% while research rose 3.6 points to 22.7%.

The week, in context

What changed
Safety & governance fell 5.4 points and Research rose 3.6 points.
Where it showed up
Safety & governance and Research accounted for 31% of the relevant conversation across 27 episodes.
How to read it
Recurrence across podcasts does not mean agreement. The cited moments show the language behind each summary.

In the discussion

safety needs to be front and center in the design for AI and also robotics
The Cognitive RevolutionAt 1:03:27Hear the context
Report this quote

What’s hot this week

Open a topic for the podcast moments behind it.

  1. Safety & governanceWays to reduce harm, set limits, test behavior, and govern how AI is used.Discussion of safety & governance eased compared with recent weeks. It made up 9% of the conversation across 11 podcasts this week.Losing attention9% of this week · 11 podcasts
  2. Adoption Use CasesPractical ways people and organizations are putting AI to work.Discussion of adoption use cases stayed near its recent level. It made up 33% of the conversation across 14 podcasts this week.About the same33% of this week · 14 podcasts
  3. ResearchNew models, training methods, benchmarks, and discoveries from AI labs.Discussion of research grew compared with recent weeks. It made up 23% of the conversation across 11 podcasts this week.Gaining attention23% of this week · 11 podcasts
  4. AgentsAI systems that can plan, use tools, and carry out multi-step tasks.Discussion of agents stayed near its recent level. It made up 15% of the conversation across 12 podcasts this week.About the same15% of this week · 12 podcasts
Showing 4 of 10 topicsShowing all 10 topicsView 6 more topicsShow fewer topics
  1. HardwareThe chips, data centers, energy, and devices that AI depends on.Discussion of hardware grew compared with recent weeks. It made up 8% of the conversation across 9 podcasts this week.Gaining attention8% of this week · 9 podcasts
  2. RegulationLaws, rules, and government decisions about how AI may be built and used.Discussion of regulation stayed near its recent level. It made up 4% of the conversation across 7 podcasts this week.About the same4% of this week · 7 podcasts
  3. Open ModelsAI models whose weights or code are available for others to use and adapt.Discussion of open models stayed near its recent level. It made up 3% of the conversation across 7 podcasts this week.About the same3% of this week · 7 podcasts
  4. EconomicsThe cost of building and running AI, and how money moves through the market.Discussion of economics eased compared with recent weeks. It made up 3% of the conversation across 8 podcasts this week.Losing attention3% of this week · 8 podcasts
  5. InfrastructureThe software and systems needed to build, deploy, and monitor AI.Discussion of infrastructure stayed near its recent level. It made up 1% of the conversation across 9 podcasts this week.About the same1% of this week · 9 podcasts
  6. JobsHow AI is changing work, hiring, roles, and the skills people need.Discussion of jobs stayed near its recent level. It made up 1% of the conversation across 7 podcasts this week.About the same1% of this week · 7 podcasts
Back to all concepts
From Sep 28–Oct 4, 2026

About this Concept

Economics

The cost of building and running AI, and how money moves through the market.

Trend this week

Losing attention

12 of 12 published weeks met the cross-podcast threshold.

A closer read

What changed
Its share fell from 3.8% to 2.5%, with evidence across 8 podcasts.
Where it showed up
It appeared in episodes including “What the Best Business AI Users Are Doing Different” and “AI put Minecraft in Skyrim (and OpenAI had a Dev Day)”.
How to read it
6 cited episodes from 6 podcasts are shown below. The topic appeared across 8 podcasts in the full analysis.

Podcast moments from the week

7 quoted moments · 6 episodes · 6 podcasts

AI For Humans

AI put Minecraft in Skyrim (and OpenAI had a Dev Day)

Play episode ↗
  1. A
    “That would have been the cost of Crafty for one shoot day, though.”
    Play from 39:07

    ConceptEconomics

    Report this moment
  2. B
    “Well, that's right. So that makes it— that's a joke she makes in the— which I didn't write. That's a joke that the character makes in my documentary, is that it costs as much as a day of um catering. Right. Which is true. Right.”
    Play from 39:11

    ConceptEconomics

    Report this moment

Latent Space

Academia is for Ambition — Alex Zhang, MIT

Open episode ↗
  1. B
    “Well, there's a lot in there. I do want to say uh the amount that OpenAI has spent is semi-public. It's uh 10,000 agents in 88 hours. 130 uh billion output tokens, which is estimated to be about $40 million in public pricing.”
    Open episode and seek to 1:04:20

    ConceptEconomics

    Report this moment

The Cognitive Revolution

AI:AM: Was Trump-Xi Anything? What Counts as Utopia? + AWS GPUs Cost 3X & AI Diagnoses Rare Diseases

Open episode ↗
  1. H
    “Even dramatically so. And the retail posted API price hasn't changed, right? Except when they kind of introduce new models. So what are you measuring on a day-by-day basis that allows you to say, Yeah.”
    Open episode and seek to 39:12

    ConceptEconomics

    Report this moment

AI Inside

The Proactive AI Agent Race is On

Play episode ↗
  1. B
    “the section I really want to read is on open weight and Chinese models. Mm-hmm. M And I'm sure they're going to position China as, uh, you know, a big strategic risk, hm but open weight is a risk to them, to their business, to their core business, for sure. And I think they've got to feel that. Um, they, uh, revenue reached $4.59 billion in 2025, which is pretty damned impressive. A quarter of that came from 2 customers.”
    Play from 54:15

    ConceptEconomics

    Report this moment

The Artificial Intelligence Show

#243: GPT-6 Sol and Luna, Opus 5.5, the New Microsoft Copilot, Jensen Huang vs. AI Doomers & Introducing AI Score

Play episode ↗
  1. B
    “the rate at which the capabilities are growing and the costs are dropping is something you should not ignore”
    Play from 21:32

    ConceptEconomics

    Report this moment

Signals to keep watching

Research

Changed direction

Research rebounded

Its share rose from 19.1% last week to 22.7% this week, reversing the previous week's drop.

Share of this week’s podcast conversation

Recent weeks19.2%This week22.7%

Came up more often than in recent weeks.

Podcast moments5 cited moments
  1. The Cognitive Revolution

    One Brain, Any Body: Google DeepMind's Keerthana on Gemini Robotics 2, Cross-Embodiment & Humanoids

    Tell me more about, it's like running into a wall at their sprint and get smashed into pieces. Clearly they are not training them like that on an ongoing basis because they'd run out of physical robots to do the training on awfully quickly. So from that alone, I was like, okay, clearly they must be doing a lot of training in simulation. Tell me more about why these kinds of tasks are amenable to training in simulation and what kinds of tasks are still difficult to train in simulation?
    Open episode and seek to 07:45
    The co— yeah. Contacts are basically, the physics is harder to model in simulation. So where the physics is easier to model, simulation gets in.
    Open episode and seek to 09:35
  2. No Priors

    Frontier Chips for Frontier AI Labs, with Walter Goodwin, Founder/CEO of Fractile

    Well, today we have a model coming out about every uh 2 weeks.
    Play from 16:18
  3. AI For Humans

    AI put Minecraft in Skyrim (and OpenAI had a Dev Day)

    Opus 5.5 really does feel like we have— uh it feels like a next-level model
    Play from 05:06
    Yes, you can iterate and make these better. I wanna make sure I, that, that Soll, it can get there, but it just shows you that, that visually how different the level is of these 2 models. And I think that's something really important. And like, honestly, like I would be slightly worried if I was at OpenAI that this is the thing we have going against Opus 5.5.
    Play from 22:06

Agents

Changed direction

Agents rebounded

Its share rose from 12.4% last week to 14.5% this week, reversing the previous week's drop.

Share of this week’s podcast conversation

Recent weeks15.6%This week14.5%

Came up less often than in recent weeks.

Podcast moments5 cited moments
  1. AI For Humans

    AI put Minecraft in Skyrim (and OpenAI had a Dev Day)

    Opus 555 does, to me, feel like a pretty significant step towards AGI, especially after my AI agent Fig made a 27-minute documentary about AI almost entirely on its own. It scripted it, it generated the host, it edited it, it added the music.
    Play from 01:08
    2 other agents that I've been testing that are completely free. Completely free and outright awesome
    Play from 15:04
  2. Latent Space

    Academia is for Ambition — Alex Zhang, MIT

    under the hood, whatever is running can be some really weird, complicated swarm of agents that ultimately come up with an answer.
    Open episode and seek to 1:02:03
    pointing models, like, or agents at a growing kind of context of information, maybe it's just enough to solve very difficult problems. That um I can buy.
    Open episode and seek to 1:03:49
  3. Machine Learning Street Talk

    How a Voice Agent Learns the Rhythm of Conversation — Shawn Wen

    an autonomous agent, when it sees problems, can just fix things
    Open episode and seek to 46:08

Safety & governance

Losing attention

Safety & governance cooled

Its share fell from 14% last week to 8.6% this week.

Share of this week’s podcast conversation

Recent weeks13.4%This week8.6%

Came up less often than in recent weeks.

Podcast moments5 cited moments
  1. The Cognitive Revolution

    One Brain, Any Body: Google DeepMind's Keerthana on Gemini Robotics 2, Cross-Embodiment & Humanoids

    safety needs to be front and center in the design for AI and also robotics
    Open episode and seek to 1:03:27
    as AI becomes more capable, Like we need to think, spend more flops thinking about safety.
    Open episode and seek to 1:04:46
  2. Last Week in AI

    #258 - Opus 5.5, Sol and Luna, Muse, DeepSeek-V4.1-Flash, Xi

    the delicate equilibrium that we are in is at great risk. Like you can't just keep inventing crazy new technologies every Tuesday and think that stability will win through.
    Play from 1:33:01
  3. AI Inside

    The Proactive AI Agent Race is On

    Who didn't put the guard— the Biden guardrail in place?
    Play from 53:15
  4. Machine Learning Street Talk

    Who Checks a Proof No Human Can Read? — Leo de Moura

    we can put safety guardrails on there, and at least there's only one model to check
    Open episode and seek to 54:36

Adoption Use Cases

Holding steady

Adoption Use Cases held its ground

Its share has stayed near 33.4% of the conversation for three weeks.

Share of this week’s podcast conversation

Recent weeks32.5%This week33.4%

Came up more often than in recent weeks.

Podcast moments5 cited moments
  1. The Cognitive Revolution

    One Brain, Any Body: Google DeepMind's Keerthana on Gemini Robotics 2, Cross-Embodiment & Humanoids

    people buy these things for novelty reasons, but they don't go back to using them
    Open episode and seek to 55:58
  2. AI For Humans

    AI put Minecraft in Skyrim (and OpenAI had a Dev Day)

    There's incredible workflows that uh try to decompile games for you.
    Play from 04:16
    I was gonna say, we see that today between the uh Fortnite and Roblox ecosystems, right? Like, not to shout out those 2 companies specifically, but if a game pops anywhere, there's a million clones on each platform. Well, is someone gonna make the AI-decompiled Newgrounds, if you will, where anytime something pops, they try to emulate it? We should shout out a couple things, right? Um First, we have to shout out um the Sega Seaman decompile. Did you see this thing?
    Play from 11:31
  3. Last Week in AI

    #258 - Opus 5.5, Sol and Luna, Muse, DeepSeek-V4.1-Flash, Xi

    It appears that the launch has been good. If these 2 investors, we'll see, you know, what the longer-term feel on this product are, but a lot of their users from, you know, the rest of the ecosystem of Facebook and so on are giving Muse a try.
    Play from 41:58
  4. Latent Space

    Academia is for Ambition — Alex Zhang, MIT

    What's an example? So is this different than, like, you know, pick your favorite company, Harvey, for example, using LLMs to do legal work or what's the—
    Open episode and seek to 1:23:20

Get the weekly AI conversation

The week’s key ideas and podcast moments, delivered to your inbox.

No account needed. One report a week. Unsubscribe anytime. Privacy

Confirm publication