Welcome to Exponential View, the show where I explore how exponential technologies such as AI are reshaping our future. I’ve been studying AI and exponential technologies at the frontier for over ten years.

Each week, I share some of my analysis or speak with an expert guest to make light of a particular topic.

To keep up with the Exponential transition, subscribe to this channel or to my newsletter: 
https://www.exponentialview.co/

----

Published in early March 2026, Andrej Karpathy's autoresearch AI tool makes autonomous scientific experimentation cheap and easy — but it was designed to solve machine learning problems. I wanted to see if I could apply its loop architecture to my own work: refining my worldview, testing arguments, solving business problems. In this video, I share how I adapted Karpathy’s autoresearch loops for problems that aren't easy to quantify, how to avoid the local minima trap, and the broader impact of these kinds of methods. I covered:

(02:11) The Karpathy Loop: what is it and how does it work

----

Where to find me:

Exponential View newsletter: https://www.exponentialview.co/

Production by EPIIPLUS1.

Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

Welcome to Exponential View, the show where I explore how exponential technologies such as AI are reshaping our future. I’ve been studying AI and exponential technologies at the frontier for over ten years.

Each week, I share some of my analysis or speak with an expert guest to make light of a particular topic.

To keep up with the Exponential transition, subscribe to this channel or to my newsletter: 

https://www.exponentialview.co/

----

Last week Jensen Huang shared the numbers from NVIDIA’s order book: AI compute demand has grown a millionfold in two years. Much GTC coverage focused on chips, robots, data centers in space, but I think Jensen revealed something far more important in his keynote: “the inference inflection has arrived,” and this is about to transform how all companies should manage their budgets. The inference era is already the operating assumption of the world’s most valuable company.

In this week’s podcast, I cover:

(1:20) NVIDIA's $1 trillion order book

(1:56) OpenClaw: our era's web browser

(7:54) Training vs Inference: how AI is changing

(12:50) Pre-fill vs. decode: the technical split

(18:06) The Harness: why OpenClaw changes everything

(18:59) The engine is useless without a car

(22:21) From 100M to 870M tokens per day

(24:29) Meet my agent R Mini Arnold's team

(26:16) AI focus group simulations at $10–50 a run

(29:36) Jensen's self-interest (and why he's still right)

(33:07) AI governance: token budgets don't belong with IT

(35:07) From training economy to inference economy

----

Where to find me:

Exponential View newsletter: https://www.exponentialview.co/

Website: https://www.azeemazhar.com/

LinkedIn: https://www.linkedin.com/in/azeem/

Twitter/X: https://x.com/azeem

Production by EPIIPLUS1.

Production and research: Baba Films, Chantal Smith, Marija Gavrilov.

Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

Welcome to Exponential View, the show where I explore how exponential technologies such as AI are reshaping our future. I've been studying AI and exponential technologies at the frontier for over ten years.

Each week, I share some of my analysis or speak with an expert guest to make light of a particular topic.

To keep up with the Exponential transition, subscribe to this channel or to my newsletter: https://www.exponentialview.co

----

In this episode I work through what I and many other commentators missed – and what road lies ahead for Apple. I cover:

(01:16) Why I was wrong about Apple

(02:40) What's behind the Mac Mini shortage

(04:07) China goes OpenClaw crazy

(06:28) Perplexity builds on a Mac Mini

(07:12) The edge case for Apple

(09:05) Apple Moat 1: hardware

(11:31) Apple Moat 2: privacy

(15:47) The K problem: when good enough beats genius

(18:08) Privacy, sovereignty & the diary problem

----

Where to find me:

Exponential View newsletter: https://www.exponentialview.co/

Website: https://www.azeemazhar.com/

LinkedIn: https://www.linkedin.com/in/azeem/

Twitter/X: https://x.com/azeem

Production by EPIIPLUS1. 

Production and research: Baba Films, Chantal Smith, Marija Gavrilov.

Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

Welcome to Exponential View, the show where I explore how exponential technologies such as AI are reshaping our future. I've been studying AI and exponential technologies at the frontier for over ten years.

Each week, I share some of my analysis or speak with an expert guest to make light of a particular topic.

To keep up with the Exponential transition, subscribe to this channel or to my newsletter: https://www.exponentialview.co/

AI has become so embedded in how I work that I can no longer cleanly separate it from my thinking. That raises a question I find genuinely unsettling: is intensive AI use making me a sharper thinker, or quietly doing the opposite? In this episode I pull back the curtain on my full research and writing process — the custom tools, the friction points, and the places where I'm still not sure I've got it right. For Ezra Klein, having AI summarize material is a disaster for original thought. But my AI systems are designed to protect the cognitive work that has to stay human, while they handle everything else. Knowing where to draw that line turns out to be the hardest and most important question.

I covered:

00:00 - Is AI worsening our thinking?

02:35 - Ezra Klein on AI and the death of original thought

04:02 - Cognitive offloading vs cognitive surrender

09:20 - Signal detection at scale

11:06 - Why I use several AI personas to scan for different insights

13:37 - AI tells me what NOT to think about

16:25 - The value of quietness

19:07 - Small notebooks, small ideas

20:01 - Writing reveals what you don't yet know

23:24 - The golden thread

25:20 - Speaking drafts aloud

28:05 - How I stress-test my arguments before publishing

29:35 - Using AI to stress-test my own house views

31:44 - Stylometer: my AI style and grammar tool

33:10 - Did AI make the thinking better?

For more on this week’s topics, subscribe to my newsletter https://www.exponentialview.co/

-----

Where to find me:

Exponential View newsletter: https://www.exponentialview.co/

Website: https://www.azeemazhar.com/

LinkedIn: https://www.linkedin.com/in/azhar/

Twitter/X: https://x.com/azeem

Production by EPIIPLUS1

Production and research: Baba Films, Chantal Smith, Marija Gavrilov.

Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

Welcome to Exponential View, the show where I explore how exponential technologies such as AI are reshaping our future. I've been studying AI and exponential technologies at the frontier for over ten years.

Each week, I share some of my analysis or speak with an expert guest to make light of a particular topic.

To keep up with the Exponential transition, subscribe to this channel or to my newsletter: https://www.exponentialview.co/

Meet R Mini Arnold - my OpenClaw chief of staff, which manages the equivalent of a ten-person team from a Mac mini in my garden studio. While I slept, that AI team debugged its own code at 3am, researched a trending Substack essay using five parallel investigators, and wrote a 4,600-word script for this very episode in 40 minutes. The gap between people who've started building this way and those who haven't is widening every week. 

I covered:

00:51 Introducing my OpenClaw agent “R Mini Arnold”

03:59 What my AI chief of staff actually does

07:58 The hardware and software stack

10:38 A morning brief before you wake up

12:05 Overnight agents: research and code

15:00 How I communicate with my agent

18:56 Example 1: the sovereign wealth fund

22:41 Example 2: how this video was written

26:34 What it costs

29:22 The soul.md personality spec

32:39 Am I losing the judgment muscle?

35:46 Individuals vs. Fortune 500s

38:25 What to try this week

-----

Where to find me:

Exponential View newsletter: https://www.exponentialview.co/

Website: https://www.azeemazhar.com/

LinkedIn: https://www.linkedin.com/in/azhar/

Twitter/X: https://x.com/azeem

Production by EPIIPLUS1

Production and research: Baba Films, Chantal Smith, Marija Gavrilov.

Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

Welcome to Exponential View, the show where I explore how exponential technologies such as AI are reshaping our future. I've been studying AI and exponential technologies at the frontier for over ten years.

Each week, I share some of my analysis or speak with an expert guest to make light of a particular topic.

To keep up with the Exponential transition, subscribe to this channel or to my newsletter: https://www.exponentialview.co/

This is the first episode of AI Vistas, a new series where I bring together people I trust and respect to tackle a major question collectively. 

Today’s question: are we in charge of our AI tools, or are they in charge of us? 

Joining me are Nita Farahany, distinguished professor of law and philosophy at Duke University and a leading thinker on cognitive liberty and mental privacy; Eric Topol, founder of the Scripps Research Translational Institute and one of the world's most cited medical researchers; and Rohit Krishnan, engineer, former hedge fund manager, and AI builder. Moderating the conversation is Nick Thompson, CEO of The Atlantic.

We covered:

(01:33) Introducing AI Vistas

(03:51) The AI agent that made a financial decision mid-drive

(05:48) What does it mean to act autonomously anymore?

(08:42) Why AI harms are rarer than you'd expect

(10:24) When AI outperforms doctors – and why that's complicated

(15:20) Constituent competence: the skill you must never offload

(18:50) De-skilling is already happening 

(31:20) What can schools do better?

(42:50) AI slop and "hollow-ware"

(46:40) What is lost when AI does the creating?

(49:18) When a tool gets good enough, we hand it off

(50:11) Deliberate intent: keeping AI as a tool

-----

Where to find me:

Where to find Nick, Nita, Eric and Rohit:

Production by EPIIPLUS1

Production and research: Baba Films, Chantal Smith, Marija Gavrilov.

Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

Welcome to Exponential View, the show where I explore how exponential technologies such as AI are reshaping our future. I've been studying AI and exponential technologies at the frontier for over ten years.

Each week, I share some of my analysis or speak with an expert guest to make light of a particular topic.

To keep up with the Exponential transition, subscribe to this channel or to my newsletter: https://www.exponentialview.co/

In this episode, I sit down with my friend Rohit Krishnan - writer of the Substack newsletter Strange Loop Canon - for a hands-on conversation about what it actually looks like to build with AI agents today. Between us we're burning through tens of billions of tokens a month - I hit nearly 100 million in a single day this week - and we share what we're each running on our own machines.

We dig into the quirks and surprising power of tools like OpenClaw, Claude Code, and Cowork, debate why AI remains stubbornly bad at good writing, and zoom out to ask what a world of trillions of agents might actually look like — and what economic infrastructure it will need.

We covered:

(03:15) What's on your screen right now?

(04:30) OpenClaw

(06:27) Rohit’s agent, Morpheus

(11:06) Azeem's agent, R. Mini Arnold

(19:25) The analyst is now a machine

(22:36) 100 million tokens in a day: the new normal

(24:44) Building tools to improve AI writing: Horace and Broca

(32:19) Why writing is the hardest eval for LLMs

(39:18) Towards a trillion agents

(42:09) The agentic economy: coordination, identity, and exchange

(46:33) How to get started with OpenClaw

(51:18) The hardest leap for new users

-----

Where to find me:

Exponential View newsletter: https://www.exponentialview.co/

Website: https://www.azeemazhar.com/

LinkedIn: https://www.linkedin.com/in/azhar/

Twitter/X: https://x.com/azeem

Production by EPIIPLUS1

Production and research: Baba Films, Chantal Smith, Marija Gavrilov.

Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

Welcome to Exponential View, the show where I explore how exponential technologies such as AI are reshaping our future. I've been studying AI and exponential technologies at the frontier for over ten years. Each week, I share some of my analysis or speak with an expert guest to make light of a particular topic. 

To keep up with the Exponential transition, subscribe to this channel or to my newsletter: https://www.exponentialview.co/

 ----

In this episode, I'm joined by Jaime Sevilla, founder of Epoch AI; Hannah Petrovic from my team at Exponential View; and financial journalist Matt Robinson from AI Street. Together we investigate a fundamental question: do the economics of AI companies actually work? 

We analysed OpenAI's financials from public data to examine whether their revenues can sustain the staggering R&D costs of frontier models. 

The findings reveal a picture far more precarious than many assume; we also explore where the real infrastructure bottlenecks lie, why compute demand will dwarf energy constraints, and what the rise of long-running agentic workloads means for the entire industry. 

Read the study here: https://www.exponentialview.co/p/inside-openais-unit-economics-epoch-exponentialview

We covered: 

(00:00) Do the economics of frontier AI actually work? 

(02:48) Piecing together OpenAI's finances from public data 

(05:24) GPT-5's "rapidly depreciating asset" problem 

(13:25) Why OpenAI is flirting with ads 

(17:31) If you were Sam Altman, what would you do differently? 

(22:54) Energy vs. GPUs; where the real infrastructure bottleneck lies 

(29:15) What surging compute demand actually looks like 

(33:12) The most surprising finding from the research 

(38:02) The race to avoid commoditization 

(43:35) Agents that outlive their models 

Where to find me: 

Exponential View newsletter: https://www.exponentialview.co/ 

Website: https://www.azeemazhar.com/ 

LinkedIn: https://www.linkedin.com/in/azhar/ 

Twitter/X: https://x.com/azeem 

Where to find Jamie: https://epoch.ai or https://epochai.substack.com 

Where to find Matt: https://www.ai-street.co 

Production by supermix.io and EPIIPLUS1 Production and research: Chantal Smith and Marija Gavrilov.

Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

Welcome to Exponential View, the show where I explore how exponential technologies such as AI are reshaping our future. I've been studying AI and exponential technologies at the frontier for over ten years.

Each week, I share some of my analysis or speak with an expert guest to make light of a particular topic.

To keep up with the Exponential transition, subscribe to this channel or to my newsletter: https://www.exponentialview.co/

-----

A week before OpenClaw exploded, I recorded a prescient conversation with Mustafa Suleyman, CEO of Microsoft AI and co-founder of DeepMind. We talked about what happens when AI starts to seem conscious – even if it isn’t. Today, you get to hear our conversation.

Mustafa has been sounding the alarm about what he calls “seemingly conscious AI” and the risk of collective AI psychosis for a long time. We discussed this idea of the “fourth class of being” – neither human, tool, nor nature – that AI is becoming and all it brings with it.

Skip to the best bits:

(03:38) Why consciousness means the ability to suffer

(06:52) "Your empathy circuits are being hacked"

(07:23) Consciousness as the basis of rights

(10:47) A fourth class of being

(13:41) Why market forces push toward seemingly conscious AI

(20:56) What AI should never be allowed to say

(25:06) The proliferation problem with open-source chatbots

(29:09) Why we need well-paid civil servants

(30:17) Where should we draw the line with AI?

(37:48) The counterintuitive case for going faster

(42:00) The vibe coding dopamine hit

(47:09) Social intelligence as the next AI frontier

(48:50) The case for humanist super intelligence

-----

Where to find Mustafa:

- X (Twitter): https://x.com/mustafasuleyman

- LinkedIn: https://www.linkedin.com/in/mustafa-suleyman/

- Personal Website: https://mustafa-suleyman.ai/

Where to find me:

- Substack: https://www.exponentialview.co/

- Website: https://www.azeemazhar.com/

- LinkedIn: https://www.linkedin.com/in/azhar

- Twitter/X: https://x.com/azeem

Produced by supermix.io and EPIIPLUS1 Ltd. Production and research: Chantal Smith and Marija Gavrilov.

Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

Welcome to Exponential View, the show where I explore how exponential technologies such as AI are reshaping our future. I've been studying AI and exponential technologies at the frontier for over ten years.

Each week, I share some of my analysis or speak with an expert guest to make light of a particular topic.

To keep up with the Exponential transition, subscribe to this channel or to my newsletter: https://www.exponentialview.co/

-----

At Davos 2026, the mood was unlike any previous World Economic Forum gathering. With Donald Trump arriving amid escalating geopolitical tensions and European leaders sounding alarms about sovereignty, I recorded live dispatches from the ground. In this special episode, I bring together observations from four days at the annual meeting, tracking the seismic shifts in global order alongside the practical realities of AI adoption in the enterprise.

Skip to the best bits:

(00:38) Day one at Davos

(02:10) Three recurring themes through the week

(03:55) Day three at Davos

(05:12) Mark Carney's stirring speech

(05:52) Why European leaders are sounding the alarm

(06:51) Why technological sovereignty just became urgent

(09:31) Day four at Davos

(12:59) What leaders really have to say on AI adoption

(14:07) The case for only using open source models

Where to find me:

Exponential View newsletter: https://www.exponentialview.co/

Website: https://www.azeemazhar.com/

LinkedIn: https://www.linkedin.com/in/azhar/

Twitter/X: https://x.com/azeem

Production by supermix.io and EPIIPLUS1. Production and research: Chantal Smith and Marija Gavrilov.

Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.