The world is freaking out about AI safety. AI researchers are resigning in their droves. Mathematicians are crying in public. But is the end really nigh?

In this interview I speak to Cybersecurity expert and Founder of cybersecurity firm Embroidery - Zack Korman - about whether agents are really going rogue or if it's just people?

Spoiler alert - the agent stuff is much less scary than you think.

Subscribe now to never miss an episode!

Key topics

The OpenAI Hugging Face hack explained

The beef between safety and security people

The twisted logic of effective altruism that dominates AI safety

The corporate negligence behind recent incidents

What would be the warning signs that that AI might actually go rogue?

Where you can watch/listen

Youtube⁠⁠⁠⁠⁠ - ⁠https://www.youtube.com/@HumansintheLoopPod

Apple podcasts⁠⁠⁠⁠⁠ - ⁠https://podcasts.apple.com/us/podcast/humans-in-the-loop/id1831285114⁠

Spotify⁠⁠⁠ - ⁠https://open.spotify.com/show/4BbslJR0KA5bfqrcYRhJER?si=b5a46211098b4f3e

⁠Or wherever you get your podcasts!

Get in touch

Instagram⁠⁠⁠ - ⁠https://www.instagram.com/humansinthelooppod/

⁠Email: ⁠seb@theuncommonfounder.com⁠

Website: ⁠https://www.theuncommonfounder.com

Where to find Zack

https://zkorman.com/

https://www.youtube.com/@ZackKorman

So much of the debate around AI is on the level of vibes or mystic predictions.

I wanted to come back to a grounded view of what AI is actually good at today. No doom, no hype.

I've long admired Rohit's work and his excellent Substack - Strange Loop Canon - much of which details his extensive experimentation and assessment of AI, so he was the perfect guest for this topic

Key topics

• AI's superpowers in verifiable domains like coding and problem-solving
• Limitations in reasoning and judgment
• The role of benchmarks in evaluating AI performance
• The concept of AGI and its misconceptions
• Societal responses and accountability for AI risks
• Legal and regulatory considerations for AI safety
• The importance of realistic expectations and responsible innovation

Where you can watch/listen

Youtube⁠⁠⁠⁠⁠ - ⁠https://www.youtube.com/@HumansintheLoopPod

Apple podcasts⁠⁠⁠⁠⁠ - ⁠https://podcasts.apple.com/us/podcast/humans-in-the-loop/id1831285114⁠

Spotify⁠⁠⁠ - ⁠https://open.spotify.com/show/4BbslJR0KA5bfqrcYRhJER?si=b5a46211098b4f3e⁠

Or wherever you get your podcasts!Get in touch

Instagram⁠⁠⁠ - ⁠https://www.instagram.com/humansinthelooppod/⁠

Email: ⁠seb@theuncommonfounder.com⁠

Website: ⁠https://www.theuncommonfounder.comWhere to find Rohit

https://www.strangeloopcanon.com/

Scott Aaronson is a Computer Scientist, Complexity Theorist and leading authority on Quantum Computing.

As he says himself, he's spent twenty years trying to dispel some of the commonly held myths about it.

As a non-scientist, I've spent some time trying to wrap my head around quantum mechanics and it still blows my mind everytime I try to think about it. This conversation was as close as I've ever been to properly understanding it.

Key topics

Quantum mechanics and probability

Quantum superposition and interference

Building reliable quantum hardware

Quantum algorithms like Shor's and Grover's

Cryptography and post-quantum security

Quantum error correction and scalability

Implications for AI and societal risks

Where you can watch/listen

Youtube⁠⁠⁠⁠⁠ - ⁠https://www.youtube.com/@HumansintheLoopPod

Apple podcasts⁠⁠⁠⁠⁠ - ⁠https://podcasts.apple.com/us/podcast/humans-in-the-loop/id1831285114⁠

Spotify⁠⁠⁠ - ⁠https://open.spotify.com/show/4BbslJR0KA5bfqrcYRhJER?si=b5a46211098b4f3e⁠

Or wherever you get your podcasts!

Get in touch

Instagram⁠⁠⁠ - ⁠https://www.instagram.com/humansinthelooppod/⁠

Email: ⁠seb@theuncommonfounder.com⁠

Website: ⁠https://www.theuncommonfounder.com

Sir Dieter Helm is the UK's leading energy economist and Professor of Economic Policy at the University of Oxford.

He is also the first Knight of the Realm on Humans in the Loop.

As a proud Brit and European, I care deeply about British and European prosperity. Technology is not the only driver of that, but it is an important one. And given the geopolitical state of the world, there is going to be more and more importance placed on sovereignty i.e. our ability to stand on our own two feet.

I often run into this argument that one of the major blockers to sovereign British technology is the high cost of British energy, so I brought Dieter on to talk about British energy and what a more sovereign technological future might look like.

Dieter's book - Climate Realism: Facing the Facts about Energy is out on October 8th: https://dieterhelm.co.uk/books/climate-realism/

Key topics

UK deindustrialization and its impact on emissions

Intermittent renewable energy and system costs

Cost structure of industrial energy in the UK

Global competitiveness and energy prices

The role of nuclear and firm power

Sovereignty and energy independence post-Brexit

Global energy resources and geopolitical risks

Technological innovation in energy and AI

Where you can watch/listen

Youtube⁠⁠⁠⁠⁠ - ⁠https://www.youtube.com/@HumansintheLoopPod

Apple podcasts⁠⁠⁠⁠⁠ - ⁠https://podcasts.apple.com/us/podcast/humans-in-the-loop/id1831285114⁠

Spotify⁠⁠⁠ - ⁠https://open.spotify.com/show/4BbslJR0KA5bfqrcYRhJER?si=b5a46211098b4f3e⁠Or wherever you get your podcasts!

Get in touch

Instagram⁠⁠⁠ - ⁠https://www.instagram.com/humansinthelooppod/⁠

Email: ⁠seb@theuncommonfounder.com⁠

Website: ⁠https://www.theuncommonfounder.com

Where to find Sir Dieter

LinkedIn: https://www.linkedin.com/in/dieter-helm-3077b214b/

Website: https://dieterhelm.co.uk/

Claire Stapleton spent 12 years at Google from 2007 - 2019. She joined as a fresh grad at a time of peak optimism and idealism in Silicon Valley, and soon became somewhat of a 'Bard' - responsible for shaping the story and sharing the folklore around Google with employees and later the outside world.

In 2018, feeling increasingly disillusioned with what Google had become, Claire was one of the organizers of a 20k employee walkout in protest against the company's handling of sexual harassment allegations against certain senior executives - one of whom, Andy Rubin, had allegations against him deemed credible by an internal investigation - yet walked away with a warm handshake and $90m exit package.

Claire and the other women who organised faced a backlash from Google with their roles and responsibilities being diminished in the aftermath.

Claire eventually left in 2019, and her memoir, 'Don't Be Evil' has recently been released and is available at any good bookshop near you: https://www.nytimes.com/2026/08/04/books/review/dont-be-evil-claire-stapleton.html

Claire's own journey maps to a time in which internal and public sentiment towards big tech in general has shifted so we sat down to talk not just about her own experience at Google, but also the wider vibe shift in the industry as a whole.

Key topics

• The early idealistic days of Google
• Claire's role as 'The Bard' - telling the internal folklore of Google to its employees and the outside world
• Google's Me Too moment - her role as organizer of 20k employee walkout
• The backlash that she and other women at Google faced as a consequence
• The vibe shift from idealism to corporate ruthlessness
• When she stopped believing the story she was telling
• Her experiences of parenting and grief and how that shaped how she felt about her work

Where you can watch/listen

Youtube⁠⁠⁠⁠⁠ - ⁠https://www.youtube.com/@HumansintheLoopPod

Apple podcasts⁠⁠⁠⁠⁠ - ⁠https://podcasts.apple.com/us/podcast/humans-in-the-loop/id1831285114⁠

Spotify⁠⁠⁠ - ⁠https://open.spotify.com/show/4BbslJR0KA5bfqrcYRhJER?si=b5a46211098b4f3e⁠Or wherever you get your podcasts!

Get in touch

Instagram⁠⁠⁠ - ⁠https://www.instagram.com/humansinthelooppod/⁠

Email: ⁠seb@theuncommonfounder.com⁠

Website: ⁠https://www.theuncommonfounder.com

Where to find Claire

LinkedIn: https://www.linkedin.com/in/claire-stapleton-868a935/

Substack: ⁠https://techsupport.substack.com/

X: https://x.com/clairewaves?s=20

Instagram: https://www.instagram.com/clairewaves

Benedict Evans is a tech analyst and former partner at A16z.

His newsletter with rational takes on tech goes out to over 175k people.

In amongst the noise around AI and the tech industry more broadly, Benedict has built a reputation for high quality, level headed analysis.

I sat down with Benedict to discuss topics such as:

• The current state of technology and ongoing platform shifts
• Whether this time is truly different or not
• Historical comparisons of technological revolutions
• Challenges in predicting AI's future trajectory
• Silicon Valley’s naivety about the world
• The cultural and philosophical aspects of AI and art
• The importance of understanding complex systems and markets

Where else you can listen

Youtube⁠⁠⁠⁠ - https://www.youtube.com/@HumansintheLoopPod

⁠⁠Apple podcasts⁠⁠⁠⁠ - https://podcasts.apple.com/us/podcast/humans-in-the-loop/id1831285114

Spotify⁠⁠ - https://open.spotify.com/show/4BbslJR0KA5bfqrcYRhJER?si=b5a46211098b4f3e

Or wherever you get your podcasts!

Chapters

00:00 Introduction and Benedict Evans' perspective on tech noise

03:33 E-commerce growth and retail media expansion

07:05 Historical scale of technological change and exponential growth

10:33 The unpredictability of technological evolution

14:08 The difference between automating tasks and jobs

17:24 The limitations of AI in creating authentic art

18:13 The organizational and deployment challenges of AI

20:39 The emergence of new pathways and the unpredictability of AI evolution

22:20 Historical perspectives on internet and mobile adoption

23:37 The importance of questions and uncertainty in analysis

25:01 The impact of the internet on industries and productivity

26:29 The exponential nature of technological progress

28:23 Organizational inertia and the difficulty of change

30:04 The complexity of automating jobs versus tasks

32:38 The role of models in understanding human work and creativity

36:26 The potential of AI as a pattern matcher and logic system

39:23 The philosophical limits of AI and human judgment

41:24 Cultural perceptions of art, authenticity, and AI

44:52 The enduring human element and cultural value in art

45:46 The future of AI and human creativity

48:47 The broader implications of AI as a platform for a new economy

50:15 Historical context of technological revolutions and their long timelines

52:24 Philosophical debates on existence, value, and AI

54:07 Benedict Evans' perspective on Silicon Valley and tech optimism

55:35 The influence of politics and regulation on AI development

56:49 The mindset needed for startup innovation and large-scale change

57:20 Benedict Evans' optimism about the future and long-term progress

58:47 Reflections on historical hardships and the progress of humanity

01:00:51 The long arc of human history and technological progress

Get in touch

Instagram⁠⁠⁠ - https://www.instagram.com/humansinthelooppod/

Email: ⁠seb@theuncommonfounder.com⁠

Website: ⁠https://www.theuncommonfounder.com

Where to find Benedict

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

Newsletter: https://www.ben-evans.com/newsletter

Website: https://www.ben-evans.com

Few people understand language quite like today's guest. Emily M. Bender is a Professor of Computational Linguistics at the University of Washington.

She's also the person who coined the term 'Stochastic Parrots' in the now infamous paper that saw Google fire its Ethical AI lead; and the co-author of The AI Con, a book which is unflinching in its criticism of both AI boosterism and doomerism (in fact she objects to the term AI).

I sat down with Emily to discuss topics such as:

• How language really works
• What LLMs are actually doing
• What Stochastic Parrots really means
• Why doomers and boosters are two sides of the same, misleading coin

Where else you can listen

⁠⁠⁠Youtube⁠⁠⁠ - https://www.youtube.com/@HumansintheLoopPod

⁠Apple podcasts⁠⁠⁠ - https://podcasts.apple.com/us/podcast/humans-in-the-loop/id1831285114

⁠Spotify⁠ - https://open.spotify.com/show/4BbslJR0KA5bfqrcYRhJER?si=b5a46211098b4f3e

Or wherever you get your podcasts!

Chapters

00:00 Introduction to AI and Linguistics

02:54 Understanding Computational Linguistics

05:47 The Stochastic Parrots Paper

08:32 The Implications of AI Terminology

10:57 Skepticism Towards AI Claims

12:58 The Illusion of Intelligence in AI

15:55 The Dangers of AI in Professional Settings

18:33 Democratization and Distrust in AI Development

28:02 Doomers vs. Boosters: The AI Narrative

32:57 Skepticism in AI: A Personal Journey

36:30 The Rapid Evolution of AI: Are We Outdated?

40:13 The Role of Software Engineering in AI

48:14 Automation vs. Human Expertise: A Balancing Act

53:02 Geopolitical Narratives: The AI Arms Race

56:33 The Future of AI: Regulation and Ethical Use

57:44 Pride in Expertise: Understanding Our Value

Get in touch

⁠⁠Instagram⁠⁠ - https://www.instagram.com/humansinthelooppod/

Email: ⁠seb@theuncommonfounder.com⁠

Website: ⁠https://www.theuncommonfounder.com

I sat down with Paul Kedrosky - investor, researcher and one of the smartest voices out there on the economics of AI and the risks of a bubble.

Here are a few of the questions we talked about:

• What is a bubble?
• Is AI a bubble?
• How does it compare to past bubbles?
• What would cause the bubble to burst?
• Why should the average person care?
• Will we see a bull market in laser tag in all the abandoned data centres?
• What needs to be true for AI to NOT be a bubble?

More info about Paul

Paul Kedrosky is an investor, writer, and researcher. He has been a regular on-air contributor at CNBC and Bloomberg, and has written for the New York Times, the Wall Street Journal, and publications around the world.

Paul started as an engineer, became a quota-carrying salesperson, then a technology equity analyst on Wall Street, and eventually a venture capitalist. He has started companies, held academic positions in the U.S. and Canada, has a popular newsletter, and is a research fellow at MIT's Institute for the Digital Economy. His Ph.D. research focused on risk, adoption, and path dependence in finance and technology.

Where else you can listen

⁠⁠Youtube⁠⁠ - https://www.youtube.com/@HumansintheLoopPod

Apple podcasts⁠⁠ - https://podcasts.apple.com/us/podcast/humans-in-the-loop/id1831285114

Spotify - https://open.spotify.com/show/4BbslJR0KA5bfqrcYRhJER?si=b5a46211098b4f3e

Or wherever you get your podcasts!

Chapters

00:00 The AI Bubble Debate

06:44 Understanding Bubbles and Their Characteristics

12:04 Historical Context: Railroads and AI

17:51 The Call Option on AI's Future

19:29 The Necessity of Bubbles for Progress

25:31 Consequences of Stranded Assets

26:51 The Dilemma of Data Centers

28:46 The AI Bubble Dynamics

31:20 The Impact of Economic Bubbles on Society

34:06 The Interconnectedness of Global Financial Systems

36:17 Public Sentiment Towards AI and Job Security

38:18 The Fallacy of Technological Inevitability

40:20 Regulatory Challenges in a Globalized Economy

42:50 Predicting the Burst of the AI Bubble

45:37 Market Sentiment and AI Valuations

49:21 The Token Economy and Its Implications

52:28 Possibilities of Change in the Current Landscape

54:57 The Future of Work and Economic Stability

Links

https://paulkedrosky.com/

Get in touch

⁠Instagram⁠ - https://www.instagram.com/humansinthelooppod/

Email: ⁠seb@theuncommonfounder.com⁠

Website: ⁠https://www.theuncommonfounder.com