#160 - Mark Suman - AI Is Quietly Changing How You Think

Summary

Podcast Summary: The Peter McCormack Show

Episode #160 – "AI Is Quietly Changing How You Think" with Mark Suman
Date: March 25, 2026
Host: Peter McCormack
Guest: Mark Suman (Founder, Maple AI; former Apple, ML/privacy engineer)

---

Main Theme & Episode Purpose

This episode explores the subtle and profound ways that artificial intelligence (AI) is shaping how individuals and societies think. Peter McCormack and Mark Suman delve into the concept of "thought capture," the often underexplored privacy risks of AI interactions, the dangers of centralizing intelligence and data, and the parallels between technological and societal revolutions. The conversation balances concern for the risks with optimism for empowering, open, privacy-preserving tools.

---

Key Discussion Points & Insights

1. Thought Capture: How AI Learns Our Minds

• [00:00–04:53]
  • Mark explains "thought capture": Conversational AIs quickly learn our thought patterns, often more deeply than we understand ourselves. By offloading problems, therapy, and daily thinking, users “become vulnerable” and effectively allow these systems to guide or influence them, intentionally or not.
    Quote:

    “We dump everything out there and then these machines are so smart that they actually learn our thought patterns way better than we do… it becomes, I think, a very powerful weapon, very dangerous weapon.” — Mark, [00:00]

  • AI use is growing explosively, often without users carefully vetting or understanding the privacy implications.

2. Opaque Black Boxes and Data Sharing

• [04:03–04:53]
  • Lack of transparency in commercial AI models (e.g., OpenAI) means we can't audit how personal data is mixed, used, or possibly shared across users or entities—including governments.
  • Both host and guest note that users don't know if queries about themselves could be referenced by others, or if personal thought data is blended in model training.

3. AI as an Uncontrollable Force

• [05:30–08:39]
  • Modern AI systems, designed to “rebuild the human mind,” have reached a level of complexity even their creators don’t fully understand.
  • Benchmarks like “humanity’s last exam” are being used, but AI is rapidly improving beyond current measures.
  • Quote:

    “We’re building the first system that realistically will map the entirety of how humans think at scale.” — Peter, [08:39]

4. Risk Layers: Individual, Societal, and National Control

• [09:46–11:49, 30:50–33:52]
  • AI’s capacity to nudge, profile, or throttle individuals' experiences is already a theoretical (if not practical) possibility. Host and guest discuss how tech could be weaponized by corporations or governments.
  • Reference to both democratic and authoritarian regimes’ interest in population control via algorithmic nudging.
  • Quote:

    “If you give these tools to nations who are happy to drop bombs … how happy are they going to be to weaponize a tool to ensure the population thinks a certain way?” — Peter, [00:25]
    “If we combine all these individuals together into one system … that is a dangerous precedent to set.” — Mark, [08:51]

5. Social Media and Psychological Manipulation 2.0

• [11:49–13:50]
  • AI is accelerating the same tactics used by social media: anchoring bias, illusory truth (repetition), and emotional manipulation. Now, the tailoring is deeply personalized and harder to detect.

6. Growing Need for Privacy-Focused AI

• [15:13–22:47]
  • Mark’s background: privacy engineering at Apple and cloud backup startups, witnessing firsthand how easily user data spills into the hands of employees or authorities.
  • Maple AI is designed as a privacy-first, open-source AI, using trusted execution environments to cryptographically prove server code matches public code, offering end-to-end encryption, and even Bitcoin-based anonymous sign-up.

7. What Should Users Do?

• [24:26–28:30]
  • Users should pause and consider what they share with AIs. Mark suggests treating what you type into an AI as if you were sending it to an enemy, not a friend.
  • Use privacy-focused tools like Maple when discussing highly personal matters; don’t put all your trust in closed platforms.

8. Innovation vs. Privacy: The Age-Old Tradeoff

• [29:17–30:50]
  • Usability and convenience have historically “trumped” privacy, but Mark and others are working to make privacy just as convenient.

9. AI, Democracy, and Societal Nudging

• [31:40–36:47]
  • Drawing on the UK’s “nudge unit” and China’s surveillance strategies, the conversation confronts how easily AIs can be used to enforce or undermine democratic values—potentially without users’ awareness.
  • The specter of governments using AI to subtly manipulate public opinion is a recurring anxiety.

10. Open vs. Closed Systems: The Fight for Sovereign Intelligence

• [39:16–43:14]
  • Open-source AI tools can help individuals and communities “level the playing field” against centralized, closed models that favor governments and corporations.
  • Close parallels drawn between sovereignty in money (Bitcoin) and intelligence (open-source AI).
    Quote:

    “Now we have sovereign intelligence. We’ve always been sovereign individuals in the sense that we have our own brains… with open-source AI we can put guardrails around our own thoughts again.” — Mark, [42:39]

11. Civilizational Tensions and Revolutionary Sentiment

• [43:14–54:18]
  • The hosts discuss mounting disengagement and distrust in government, growing economic anxiety, and the potential for new forms of revolutionary—or evolutionary—pushback, both technological and generational.
  • AI, Bitcoin, and new forms of journalism are seen by both as tools for reasserting individual control.

12. Generational Impact and the Coming Conflict

• [62:20–68:46]
  • Discussion of how younger generations have been financially and psychologically disadvantaged, deprived of opportunity, and manipulated by economic and technological systems.
  • Mark and Peter agree there may be a need for generational “warfare”—not physical, but as a pushback for better policies and societal direction, though Peter emphasizes he doesn’t want this to be a turn toward socialism.

13. Education, Socialization, and Technology’s Role

• [68:46–74:11]
  • Parenting in the social media age is fraught—screens, lack of privacy, widespread exposure to inappropriate content.
  • Reference to Jonathan Haidt’s “The Anxious Generation” and the importance of collective, community-based limits on tech for youth.

14. AI, Jobs, and the Future of Work

• [74:12–74:46]
  • Mark is (long-term) optimistic: technological transitions are painful but open doors to new opportunities, faster reskilling (enabled by AI), and greater fulfillment.
  • AI may free people from drudgery ("sending emails and hopping on Zoom calls") to pursue more meaningful work and lives.

15. Regulating AI: Possibilities and Limits

• [75:30–79:28]
  • Regulation moves slowly and may not keep up with open-source, decentralized AI developments. Real risk points lie with "hyperscalers" (cloud giants); Mark advocates for powerful local AI to maintain real autonomy.

16. Superintelligence and Human Co-Existence

• [79:28–83:48]
  • Mark believes superintelligence is coming, but doesn't assume it will be hostile or replace humans. More likely, humans will wield—and hopefully guide—the values of advanced AIs.
  • Quote:

    “I think we will work together in some way… we will infuse a value system.” — Mark, [81:59]

---

Notable Quotes & Memorable Moments

• "It becomes, I think, a very powerful weapon, very dangerous weapon. Somebody can turn the dials and say, I don't like the way that Peter thinks I'm going to turn him down and I like the way that Mark thinks I'm going to turn him up." — Mark, [00:00]
• "There’s humanity’s last exam, which is one of the best benchmarks out there… AI tools to help us make sure it is not lying to us and hallucinating. But... how do we know that it’s not lying to us about these big problems that we can’t even understand?" — Mark, [07:37]
• "If you give these tools to nations who are happy to drop bombs… how happy are they going to be to weaponize a tool to ensure that the population thinks a certain way?" — Peter, [00:25]
• "We’re mapping the entirety of how humans think at scale… a powerful, dangerous weapon if we combine all these individuals together into one system." — Mark, [08:51]
• "Privacy is a requirement for democracy. Without privacy you do not have democracy." — Peter (referencing David Chaum), [31:40]
• "We always choose the easy thing that is the most… convenient. Convenience over privacy. And that is something that needs to be solved." — Mark, [29:43]
• "Do you see a strong alignment between that and Bitcoin then?" — Peter, [41:57]
  "Yeah, I would say it’s sovereign… sovereign money and now we have sovereign intelligence." — Mark, [42:11]
• "Do you think we are heading towards Terminator, The Matrix, or Ready Player 1?"
  "I think Ready Player 1… but those are all still kind of dystopian… I think we will work together in some way." — Exchange, [81:59]
• "Do you think there will be a premium in the future for human interaction and authenticity?"
  "Absolutely… analog will come back. I hope it comes back. Because I don’t know if I want to live in a world where everything is thought for us." — Exchange, [89:47]

---

Timestamps for Important Segments

• [00:00] — Thought capture and AI as weapon
• [01:23] — How we expose our thinking to AI
• [04:03] — Black box AI and data privacy concerns
• [05:30] — The inscrutability and uncontrollability of modern AI systems
• [08:39] — Mapping human thought at scale/centralization danger
• [11:49] — AI as supercharged social media algorithm/manipulation
• [15:13] — Mark’s career and personal experience with user data
• [18:05] — Apple’s privacy culture and reaction to ChatGPT
• [22:47] — Maple AI’s privacy architecture
• [24:26] — Advice to casual users: caution with data sharing
• [29:17] — Usability vs. privacy: convenience paradox
• [31:40] — AI, democracy, and nudging/manipulation by state
• [39:16] — Open-source: tools for sovereignty, not just power
• [43:14] — Societal cracks, new forms of resistance, parallels with Bitcoin
• [62:20] — Generational injustice, manipulation, and social media
• [68:46] — The parenting challenge in the digital era
• [74:12] — AI displacement, work, and the possibility of liberation
• [75:30] — The challenge of AI regulation and risks of hyperscalers
• [79:28] — Defining and speculating on superintelligence
• [89:47] — Nostalgia for analog and authenticity; the hope for a premium on humanity

---

Conclusion: Tone & Takeaways

The conversation is intense and thoughtful, marked by deep skepticism of centralized power—either by government or Big Tech—but ultimately hopeful. Mark and Peter emphasize the dual nature of AI: its risks are grave, but the potential for open sovereignty (in both money and intelligence) creates new hope for individual empowerment, community resilience, and human flourishing.

Listeners are encouraged to be vigilant about privacy, push for open systems, and consciously balance convenience with control. The future, while daunting, is not foreclosed: we can still shape technology’s role in our lives.

---

Where to Learn More

• Maple AI: Trymaple AI
• Mark Suman on Twitter: @MarkSuman

---

End of Summary

Transcript

[A]

We dump everything out there and then these machines are so smart that they actually learn our thought patterns way better than we do. And once they have that information, there's, there's so much that they can do to like, lead and guide us if they want to. It becomes, I think, a very powerful weapon, very dangerous weapon. Somebody can turn the dials and say, I don't like the way that Peter thinks I'm going to turn him down and I like the way that Mark thinks I'm going to turn him up.

[B]

If you give these tools to nations who are happy to drop bombs on top of other nations, how happy are they going to be to weaponize a tool to ensure that the population thinks a certain way?

[A]

The system could quietly and perceptibly move your thought process that direction. Now they've centralized the money and they've centralized the intelligence and like, what hope do we have if we allow that to happen?

[B]

This show is brought to you by my lead sponsor, Ayran. The AI Cloud for the Next Big Thing. Iron builds and operates next generation data centers and delivers cutting edge GPU infrastructure, all powered by renewable energy. Now, if you need access to scalable GPU clusters or are simply curious about who is powering the future of AI, check out iren.com to learn more, which is Irene.com Mark, can you explain thought capture for me in the world of AI thought capture?

[A]

Well, we talk to these machines. We, they feel like they're humans to us and so we be, we, we, we let our guard down, we start to become more vulnerable to them and we tell them everything that's going on in our mind, whether that's a therapy session or some problem at work. We're trying to work through a technical solution. We, we, we just kind of, we dump everything out there and then these machines are so smart that they actually learn our thought patterns way better than we do. So we don't know how we reason ourselves sometimes, but they start to understand that. And once they have that information, there's, there's so much that they can do to like, lead and guide us if they want to. So I think that's, if I were to wrap up thought capture in one way, it would be them understanding our thought process.

[B]

Is this a part of the growth of AI that people haven't really spent too much time even thinking about or enough time thinking about?

[A]

Yeah, part of it is that people don't have time to think about it because new tools are coming out so quickly and so they're just jumping in. They See everybody else racing ahead of them, and so they need to get that too. And you look at these people that you know and trust and you say, oh, you know, these technologists, they were always the ones sounding the alarm, so if they're using, it must be safe. But I think a lot of these technologists are also kind of captured in this hysteria right now of, I must move faster, there's this wave coming, I need to be in front of the wave, not underneath the wave. So, yeah, people aren't really, in my mind, people aren't vetting everything clearly first.

[B]

Well, I am now. Thank you. Yeah, so I've noticed it with AI, for example, we use it to prepare for the show. Prepare for today. Tell me about Mark, Tell me his background. Help me prepare a structure. What do you think would be an interesting structure? And AI comes back and says, well, I know you like X and Y, so you should focus on X first than Y. And that, I guess, is what you're saying is a part of it. Starting to understand me and how I think.

[A]

Yeah, definitely. And for example, if I were to ask ChatGPT to help me prepare for this show, it doesn't know much about me because I've been very selective of what I revealed to ChatGPT. So I use a myriad of. I use Claude, Chat, all sorts of things. Obviously use my own stuff too. And I'm very careful of what I tell certain products because I don't want it to get this deep understanding of who I am in my soul, if you will.

[B]

Do we know if it isolates currently my thoughts to my instance of, say, ChatGPT, or if you go into ChatGPT and you are asking things about me, would it know those things from how I've used it and use that information to feedback to you?

[A]

That's a great question. I don't think we truly know the answer to that. Only somebody within OpenAI would know the answer to that. And that's really the biggest problem with these opaque companies, is that we don't understand the directives that they operate with. And so we don't know. Obviously, we know that they have connections with government organizations. Right. But we don't know what the government has told them to do and what kind of deals they've made. And so we don't know where those lines are being crossed. Ideally, your content and my content are separate from each other. We do know that they're going back into training these models. So eventually your thoughts and my thoughts are going into the big new model that's going to come out, whatever GPT6, whatever it is in the future that gets mixed up in there. But your personal biography and my personal biography, it's possible they're mixing, but we just have no idea.

[B]

That could be concerning.

[A]

Yeah, we can't inspect the systems. Right.

[B]

Well, we had Connor Lee in here recently who works with control AI, and our main discussion was about how these systems work and the risk of AI and superintelligence eventually wanting to control and kill us all. We talked about the gorilla problem, but as part of the conversation, he said one of the things that he knows is that the people building these systems don't truly know exactly how they work.

[A]

Yeah, because you're trying to rebuild the human mind. You know, AI didn't start this way. AI didn't start with neural networks. It started back in the 50s, trying to be more symbolic. And they were trying to solve how to play chess, how to play checkers, how to play some of these games, and it was very more deterministic. And then in the 90s is when they said, let's try to build a neural net and let's try and, like, map the human brain. Problem is, we don't understand the human brain well enough yet. And so how can you build a system, understand how the system works if you don't understand the thing you're emulating? And. And so they are. We build these benchmarks, we test it against certain things and we say, well, clearly the benchmark is performing better on this one than the previous one, and so we'll keep moving forward. But a lot of times they don't understand why. There's this classic meme in software engineering where you say, like, my code is broken and I don't know why. And then the next step, you're like, my code is working and I don't know why. And it's like the same exact person. And I feel like that's just happening on a very large scale with AI right now.

[B]

So it's not doing as they choose. They tweak something and it works, but they don't know why.

[A]

Yeah, yeah. They don't understand why this path is going this way. We'll keep gaining understanding, but it's this large. LLMs are this massively large file where you've taken all of human intelligence and compressed it down into one file, and people don't quite get how it all works yet.

[B]

But surely if we're scaling towards superintelligence, which I'm pretty sure the Nvidia CEO came out yesterday, and Said we've achieved it. It's going to scale at a speed quicker than we can ever understand it.

[A]

Yeah. Jensen came out and said that we've reached AGI. Obviously he's financially incentivized to say that because he's going to sell the chips to everybody that. That wants that future. If you look at how AI performs, I don't think we're quite there yet, but I've got your question now.

[B]

Well, we're going to be scaling this intelligence faster than we. If we can't understand it now, we're scaling it faster than we could ever understand it.

[A]

Yeah. And then we'll be depending on the intelligence to help us understand itself.

[B]

But we can't trust it.

[A]

And we can't trust it because we

[B]

already know it lies or hallucinates.

[A]

Yes. Yeah. So we have to, you know, there's, there's humanity's last exam, which is one of the best benchmarks out there. So we can depend on some of these tools to help us make sure that it is not lying to us and hallucinating. But it's very possible that we move beyond something like humanity's last exam. And now we're in this new territory where we have to figure out how do we know that it's not lying to us about these big problems that we can't even understand.

[B]

What is humanity's last exam?

[A]

It's an exam that has all sorts of just crazy hard problems on it, like physics problems, math problems, literature problems. And if any one of us took it, we would have a very difficult time. What you need is you need the expert in every field to answer one of those questions. And I might be misrepresented here, but you basically need the best person at every question to answer that one question to get 100% on the exam. And so they, they throw that exam at each one of these new models when they come out to make sure, you know, see how well it performs. And they're doing. They're scoring like 50%. Like they're, they're still working their way up, but they're getting, they're getting better every time.

[B]

So we essentially, I say we collectively, we humanity, is building the first system that realistically will map the entirety of how humans think at scale.

[A]

Yes, we're mapping that. And then kind of like you brought up before, we don't know where it's crossing these boundaries of your personal thoughts and my personal thoughts. And so there is the big H, humanity. How do humans think at scale? And I think that's a shared body of knowledge that's okay for us all to kind of share together. But then there's a small H each human, how they think, and we're all individuals and it becomes, I think, a very powerful weapon, very dangerous weapon. If we combine all these individuals together into one system that now somebody can turn the dials and say, I don't like the way that Peter thinks I'm going to turn him down. And I like the way that Mark thinks I'm going to turn him up. That is, if we enable a system like that, that is, I don't know, it's a dangerous precedent to set.

[B]

You took a big deep breath before you answered that question.

[A]

I did.

[B]

Did you not notice it?

[A]

No. No, I didn't. Yeah. I guess I was just trying to collect my thoughts and understanding the ramifications, I guess, of what, what that would do. I mean, think about a tool. If you were given a tool of, of all of humanity, right? Let's say OpenAI, they have a billion users and they really have the power to go in and know each one of their users because they all sign up with a credit card and so they know payment information for everybody. And Sam Altman could, if he wanted to, theoretically go in and just turn off one of our accounts or slow down one of our accounts, or he could alter the way that AI talks to just you to kind of steer you in a different direction if you wanted to. I don't think he is necessarily, but that is the tool that has been built.

[B]

Well, I was watching a documentary last night about social media. It was on the BBC, so it had a certain political bent to it, but they were talking about how the algorithms worked. And the algorithms were designed around the human mind and they weaponized and monetized fear and anger. That was how that was the best way to create engagement with people. We also know ourselves with releasing this show that how we package a product will define how many people see it. So we try and be as responsible we can in the content, but we know the content is important people to see, but therefore we're incentivized to package it in a way that strikes into the fear. And so we know companies will operate towards the dollar and we know governments will operate towards control. What we're essentially saying is we've got a more advanced tool than the previous tools, which was search, click and behavior. Now we've got a tool to mass control people.

[A]

Yeah. And you look at social media and I think we're taking the exact same patterns and now we're accelerating them. Right. Social media had three big things that it did. It used anchoring bias. These are the algorithms for how they structured your feed. Right. So anchoring bias, it would just throw a statistic out there. Even if it was wrong, it didn't matter. Throw a stat out there. And now you are anchored to that. And any argument that somebody makes now has to, like, remove you from that anchor. So the example I would like to use is ice cream. You know, if you say 50% of adults in the UK like strawberry ice cream, it's a stat I just made up. But now if you were to go look at something, you have to be convinced that that is not correct, and that's a tough thing to pull you away from. Then you have illusory truth, which is repeating things over and over again. So if you see something three or five times about strawberry ice cream, that just cements that anchor in your brain even more. Um, and then the. The final one, I'm. I'm escaping it right now. But it basically has to do with the emotional triggering of how something shows up in your feed, right? So you talk about fear and anger. If there is a. If they know that a piece of good news is coming in your feedback, they can put something very triggering and angering right before that. So you're in this really bad state of mind. So when the good news lands, it doesn't get received as well. And so they can keep you in this state of fear and anger. And they use those tools with social media algorithms. And it's been well documented those now get accelerated in AI, where the anchors. AI understands you so well. You know, when you prep for the show, it. It could have placed certain anchors in spots where you didn't realize, right? To put you somewhere in a frame of mind. And then it can repeat those. Every time you go talk to it, it's repeating those anchors over and over again. And there's just so many things that can do with those same tools, but in a way that's almost imperceptible to us because it can be very nuanced because it knows. You know, you mentioned that I took a deep breath that I wasn't aware of. It knows all of your deep breaths, metaphorically, in the way that you think. And so it knows exactly. How do I convince you of this thing without you knowing?

[B]

I want to talk to you about my new sponsor, Saily. So one of the things that happens with making this podcast and recording shows is that I'm constantly traveling, especially out to the US to record interviews. And one of the most annoying things I have to deal with is getting to another country and figuring out what to do with my phone. Because either you're going to get hammered with roaming charges or you're just trying to find a SIM card at the airport. And so that's why I've started using Saily now. Saily is an ESIM app from the creators of NORDVPN that gives you affordable mobile data in over 200 destinations worldwide. And listen, the setup is really simple. You just download the Saily app, pick a data plan for the country you're traveling to install the ESIM once, and when you land, your phone just connects to the local network. There are no roaming charges, no swapping SIM cards, and no airport kiosks to deal with. And listen, if you travel a lot like I do, it's a genuinely useful app. You can get 15% off saily data plans by downloading the Saily app and using my code. Peter McCormack it's clear there's a lot of deep risk with this technology and with these tools. And we should allow the listener to at least understand that you're working on something, a privacy focused version of AI because of this. But also you have a deep background in privacy. Do you want to just let people listening know your background so they understand why we're having this conversation?

[A]

Sure, yeah. My most immediate background was working at Apple for six years and I was working in machine learning, which is what the whole industry was calling AI prior to this ChatGPT explosion. Working in machine learning, AI and doing a lot of privacy work within the company prior to that. My, my first job all the way back goes to cloud computing where we were doing software that would back up people's computers to the cloud. So you would pay $5 a month, all of your family photos, all of your documents, everything back up to our cloud. And that's where I first understood that the cloud is really just somebody else's computer. It's not this magical place of military grade encryption because our users would just back up all their stuff. And then me as a software engineer, I would sometimes be brought into a support request and they would say, hey, can you help this person get their photos back? Their computer crashed. And so I'm in there and I'm like, on with the customer and they're like, oh, hey, can you just spot check some of my family photos for me and make sure they restored correctly. So I'm in there looking at their family vacation. Me just Some random software engineer in America and they're from anywhere in the world. And I'm glad that it was only family photos and not something else. Yeah, yeah, exactly. And that was like this huge light bulb moment for me that it's like these people are trusting me. Some random person that they, they didn't vet. They don't know who I am, they don't know my background, and they're just totally cool with me seeing all of their data. Now, we did allow our customers to have a private key encryption option if they wanted to, where they could generate their own long string of text. They write it down, they save it. It probably sounds familiar to you in the Bitcoin space, right? They have to make sure they safeguard that because if they lose that encryption key, they lose access to all of their Data. Less than 10% of our users picked that option because it was so difficult and scary. So 90% plus would just give us all of their stuff and, and just trust us to be okay with it. Yeah. So my, my background kind of goes back 15, 20 years of just this deep understanding of if you use the cloud, you're just giving everything away to somebody else and they're going to monetize it. They can turn it over to the authorities if they need to. You know, we got subpoenas all the time, people going through divorce proceedings and they would need all the online backups to have chat history and all that kind of stuff. So, yeah, it's just, it's, it's, it's. The world that we've built with the cloud is just trusting other people.

[B]

But you were also working on privacy with Apple, and Apple, I'm assuming it's a pretty cushy place to work. Successful company, I'm sure. Great perks. You're well looked after. You've left to go and work on Maple AI, which is a privacy focused AI solution. How much of this is you spotting an opportunity, which, you know, yeah, we should all do, and that's what capitalism is. But also you recognizing the risk with AI and this becoming a burden to you.

[A]

Yeah, no, it's, it's a bit of both. I worked in a lot of startups before going to Apple. My time at Apple was awesome. I loved it. Great people, really cool technology. Obviously you can operate at a scale of technology that it's difficult to do at any other company, so that was, that was really thrilling to do. But when ChatGPT came out, they immediately messaged everybody and said, do not use ChatGPT. So I don't know if you know much about Apple's internals, but okay, so if a new phone is being worked on.

[B]

Yeah, I've always wondered how there are no leaks. I mean, I know there are some leaks, like someone will get the schematics of a, I don't know, the iPhone 18 and a module in there leaks

[A]

from a Chinese factory.

[B]

You see some like tiny things but generally like they can do a big announcement of a brand new product that clearly thousands, maybe more have worked on and nobody knew it was coming. I just wonder how they do that.

[A]

It's a, it's a culture of disclosure basically within the company. They've built it similar to how a government contractor would operate with, with clearance levels. And so if there's a project, it's get, it gets assigned a disclosure level and the biggest ones have the tightest control. And so for you to get into that, you, you don't even know it exists. To request access, you have to kind of be assigned to a project and even if you're on the project, you may not have access to the whole thing. You only have access to parts of it. So, so it's, it's really fun to work in there. But that's, I'm trying to remember why I brought that up. But it's a, it's a place that really values privacy effectively. Right? Because they want to control the narrative of when they release stuff. They want to wow everybody. And so privacy is kind of ingrained throughout everything Apple does internally with those products as well as externally. That's why I brought up. So the reason I brought it up is when ChatGPT came out, they said you may not use it because we understand how this works and we don't want all of our undisclosed projects to be leaked into ChatGPT for very good reasons. Right. We don't want competitors, we don't want non Apple systems that have knowledge of all this stuff. So for me that was a bit of a light bulb moment of like, hey, all the stuff I've done in cloud computing for the last 15 years, this is applying to AI now. And people are going to start dumping lots of information in here and just going to trust random employees at OpenAI with their data. Apple sees this, they're not going to let us use it. How are everyday people going to handle this? Are they just going to use ChatGPT and give everything over to them? And you know, that's, I left Apple, I teamed up with the guys over at Muni, we decided to pivot and when we Started looking at what can we build. There was this thing in me from my Apple days and they were also kind of going down similar path and we decided we need to build this tool. Let's put it out there. It seems like people want something like this from conversations we've had. So we went down the path of Billy Maple, which is a privacy encrypted version, similar to ChatGPT, but end to end encrypted.

[B]

I had a play with it. It's very cool.

[A]

Yeah.

[B]

I also like the fact you can sign up anonymously with Bitcoin. Very cool. I asked it why I could trust it.

[A]

Okay.

[B]

It gave me a big long reason why I can trust it. But I also noticed that it integrates with. You can use other models. Have you written your own model or are you only using other models?

[A]

We're using the open weight model, so we haven't created our own. It's a very expensive task to try to build a frontier model. So we're using the open ones and that's intentional. We could just pass things through to OpenAI. But we really want to have this open transparent ecosystem around Maple. So all of our code is open source, the Maple app is open source, the server code is open source. And then we use trusted execution environments, or TEEs, so that you get cryptographic mathematical proof that the code on GitHub matches the code running on the servers. Which is not to be said for most companies, most cloud software, even if they publish open source, you don't know when they took that open source code to run on the server, if they inserted something in there, inserted a backdoor for the CCP or inserted a logging thing so that they could have an extra copy of all the data. You simply have no guarantees. And trusted execution environments are a hardware encrypted way to kind of provide those guarantees. And so we're just giving people transparency and the ability to verify all of our claims throughout, all the way to the open models. So now they can take this open model and run their own benchmark, testing against it if they want to, and say, okay, Maple's using Kimi 2.5. I've run Kimi 2.5 on my own hardware and I trust it. So now I can verify all the different pieces of Maple and know that I can trust it throughout.

[B]

So I'm using ChatGPT probably the most. I use Google Gemini only for work for the podcast and I use Perplexity. And I often use Perplexity when I've run something within ChatGPT or Gemini. I ask it to check the work and tell me what it thinks. But if I went into ChatGPT now. Let's do it. That'd be interesting. I'll give you some of the kind of things I've. I've done in here. So preparation for the show. HR process that I'm running through at work, researching things I've seen in the news. I'm struggling to log into a system. Okay, how do I do that? Code repair. I mean, a lot of it's work salary and taxation calculation. That seems like such a simple request, but I'm now telling exactly what I own.

[A]

Yeah,

[B]

research on a film. I watched Hamnet. I watched that. I got confused in the film. We even ran it the other day to check if I could make a connection between me and another person. One we couldn't do in Google, but I found a connection as humans. My notes for my program for my football team. I mean, you see, I'm basically going through endless things that I'm thinking about where something can solve something me quicker than I can solve myself.

[A]

Would you say that there are things you don't put in there?

[B]

That's a good question. Yes, there are.

[A]

Okay.

[B]

Yes.

[A]

And do you feel like you actively think about that? Like, oh, I can't ask this, or is it more in retrospect, you think, oh, I actually didn't ask it that.

[B]

I've asked it things which I wish I hadn't like nothing I'm embarrassed about. I could talk about it here. So handling certain situations. I've got two kids parenting, the challenges of parenting. How do I deal with this scenario with a teenager? But I realize I've now dumped information with regards to, say, one of my children. More recently, I've been giving it less and less information, but I've asked things like I struggle to sleep. Give me. How do I sleep better? Do I have sleep apnea? You should run this test. How do I. How's your sleep now? Yeah, there's certain things that if you go back maybe a year, there's things I've asked it that I wouldn't ask now. But there's things I'm still probably given information that I didn't realize I shouldn't be. I'm now very conscious of it in preparation for this interview. So where I'm leading you with this is to say I'm probably a more advanced AI user than most. You've probably got the top 1%. I'm probably in the. Not even the next 10%, the 20% after that who are business people who are using it all day, every day like they used to use Google, dumping everything in there with, you know, conscious of certain things. But there's going to be probably millions of casual users, even my kids maybe, as casual users. What would you say to those people right now? Listen to this. You've been casually using without really thinking about.

[A]

About it. Yeah, I would say listen to those moments when you pause or hesitate before saying something. Right? If, if imagine you, when you're talking to AI, imagine you're actually texting a friend all of this information. Or maybe you're texting an enemy that you have. So if you're a kid in school, you wouldn't want to text someone who you have trouble with all of your deepest, darkest secrets, because they're going to do something with that that you don't like. You need to treat these AI products that way because you're just a data leak away or a court subpoena away from your information being put out on the Internet. And we already have examples of them accidentally posting people's chats online. ChatGPT did Grok, did Meta has. And so even though these are very strong systems, they do have weaknesses in them. So it's. Yeah, you really should just be treating that. I would also tell people, like, use the tools, they're great, they do lots of amazing things, but you don't have to just pick one and go all in on it. So you could install Maple app next to your other tools. And when you're in that moment where you're talking to ChatGPT and you hesitate and say, I don't really want to say this to ChatGPT, then you switch over to Maple and you say it there and then see how you feel your difference. And then you start to understand that these are the things I talk to Maple about, these are the things I talk to ChatGPT about. And what our users tell us time and time again is that I feel like I can be myself inside of Maple because I trust it more, because I know it's private. And AI has this weird dilemma where AI needs to know as much as possible about us to be most effective, but the more that we tell AI, the more vulnerable we become. And so if you don't trust the AI that you're using, you're not going to be most effective using it because you won't want to get vulnerable with it. And so that's why we feel like having an AI system with privacy as the foundation has the ability in the future. To become much more powerful for the user than any of these other systems are right now.

[B]

Well, so I went through this with Jamis a lot. Very early on. He said to me, treat every conversation you have privately, like it could be in public, as being very useful. Public profile, doing a job I never wanted to get in a scenario. Say I was dating a girl and there was some kind of message that she eventually wanted to make public and embarrass me. So I was very good at that. But then he talked to me about other privacy things with regards to your phone, where you want to lock things down. As soon as I locked everything down, my phone became actually not very useful. So that is one of the challenges of privacy. One of the great things about the AI is that ChatGPT has got to know me and it does help me with my work. It's more accurate, but at the same time, I'm introducing a risk. Mm.

[A]

And for the longest time, people have had to choose between privacy or usability. Right. You probably use ChatGPT because it's super functional. It's always got the latest features. Cloud always has the latest features. And so people go to that and they sacrifice privacy. And it's just. It's a tale as old as time. We always choose the easy thing that is the most. What's the word?

[B]

Convenient.

[A]

Convenient. That's the word I was looking for. Yes. Convenience over privacy. And that is something that needs to be solved. It's something we're working on solving. Other people are working on solving it too, and that is make the convenient thing private as well. And once you solve that, suddenly that becomes a huge unlock for people. When we started building Maple, we went and pitched the idea to a lot of venture capitalists and they were all still operating in this B2B SaaS world, this model. And they would tell us time and time again, we want you to be the AI for lawyers or the AI for healthcare. Go to enterprise and sell these big enterprise contracts. Don't do consumer. Consumer is very difficult and you're just going to run out of money. The problem with that is if we just went and became the AI for healthcare, you wouldn't be able to afford Maple, you know, and nobody would be able to afford it. And the consumer would just be screwed over. The everyday individual would only be left with tools like ChatGPT. And so we wanted to build something for the individual person and make sure that we are here building tools that they can utilize that can become more convenient. Just like ChatGPT, we are talking about

[B]

Layers of risk here though, because there is the risk that data that you share privately can become public, which is already something we operate with on any Internet enabled tool, then there is the risk that the tools start to, I don't dumb you down, it starts to do too much work for you and you stop developing yourself, you, you stop thinking. Then there's the additional layer of risk that it starts to manipulate you as an individual and then the highest level risk that it is weaponized against a nation. Yeah, so we're talking about different layers of risk here.

[A]

Yeah, we need to make sure that we are mindful of each layer, I guess is what you're saying. Yeah. I don't know. Where would you like to go and explore on that?

[B]

Well, I want to go at the highest level of risk. I mean, if we are signaling to these tools how we think and how, I mean, it was David Chaum who said privacy is a requirement for democracy. Without privacy you do not have democracy. And you can certainly see that in more kind of rogue states where books are banned, your transactions are followed. If your journalist reports in a certain way, you might go to jail. Erdogan is a great example in Turkey, imprison lots of journalists. If you give these tools to nations who are happy to drop bombs on top of other nations, how happy are they going to be to weaponize a tool to ensure that the population thinks a certain way?

[A]

Right, yeah. So we do a lot of work with HRF and Human Rights foundation and they're studying a lot about how China uses AI for mass surveillance. And it's not just camera surveillance, it's this kind of thought police. So if their users are using Deep Seek, they now have dials they can turn for the general populace and they can understand what you're thinking about. So if you start to ask a certain questions, they get flagged and they know this person might be going off a different direction. So yeah, it becomes this tool that can be used to move and nudge an entire populace a certain direction based on the whims of a dictator. But then you can see how that could be applied in a democracy as well, where we think that we have our own autonomy, we think we have freedom of thought. But if we're all using these black box systems that have agreements with governments, directives can be given, dials can be turned, and if there is some kind of piece of legislation that's going to be coming down the pipeline, a politician could go to Claude or Anthropic and you know, nudge, nudge. We need this to be accepted by the population within a year. So please start nudging them in that direction and the system could quietly and perceptibly move your thought process that direction. In theory that can happen. I don't have evidence of it happening, but in theory that can happen.

[B]

Well, are you aware in the UK they had a specific unit called the nudge unit?

[A]

No.

[B]

So I can't remember under which government it was, I'd have to look that up. But they had a unit called the nudge unit. And the idea being is that if you want to get people to do things, you nudge them in a certain direction. So really easy example is pensions. People have not been very good at taking out pensions. You know, they get their salary and they live a certain life. That means that come the end of the month they've got no money left. And pension, I'll do that when I'm 30 or when I'm 30, I'll do when I'm 40. And there's a high risk. There's a lot of people can hit pension age and they can't retire or when they do, they're just not going to have much of a pension. They're going to rely on the state pension. And so one of the things they did to nudge people towards pensions is if you're a company that has a certain size, let's say 10 employees, you now have to auto enroll people into a pension and they have to choose to not be part of it. So say with Kurt here or Conor, I would auto enroll them into pension. If they didn't want it, they would have to go and opt out. So it used to be opt in, they nudged it to opt out. Now more people have pensions and that, fine, I understand why they're doing it. I disagree with it. Yeah, I think that is coercion, but that's just on pensions. What if, I mean, there's a big talk about everything in the UK at the moment. If a lot of rational opinions are considered far right these days, what I think is rational is considered far right. Well, if the government thinks there is a risk of right wing opinions, how are they going to nudge people to think differently? Based on that documentary I watched with the BBC the other night, they were talking about people who would see one piece of content and then search for another piece of content. Perhaps they've become black pilled or red pilled and they've moved from being someone who was from the left and to the right. Well, how do they nudge them back to the left.

[A]

You think about how, let's say we're in the UK and I'm chatting with it and I'm starting to say ideas that maybe the government wouldn't like. Even if it's not fringe all the way over on one side, it's in the center, but it's center leaning the wrong direction. In the output of ChatGPT, for example, they could grab some person that is fringe, very far right and grab a quote from them that is saying a similar idea to what I'm chatting about right now. And so then that, that pins this person that maybe I despise to the idea that I'm having. And it creates an emotional visceral reaction within me where I say, oh, I don't want to be affiliated with that person. So ChatGPT is not telling me your idea is alt right or whatever, because I would reject that. I would say, no, it's not. I'm, I'm in the center. I'm trying to be reasonable here. It's like, well, this idea you thought was reasonable is actually espoused by this really awful person on the other side. So maybe you shouldn't think that way. I think those are, those are an example of how they could do that.

[B]

It came out recently. There were 12 books that the government identified that if, if you, if you, if you're a fan of these books, if you read these books, you are far right. And they include 1984. Yes.

[A]

That's the most ironic thing.

[B]

1984. Lord of the Rings was in there. Leviathan, I think was in there. I think there might have been. Edmund Burke was in there. I basically. I wouldn't, I. There were nine of them I didn't have. So I went and bought them all, obviously.

[A]

Yeah. It's like, this is the reading list I want to read now.

[B]

Yeah, I want to read these. But, but we do know that the state. I mean, the one thing that aligns every single party, doesn't matter who they are, is that their goal is to win and retain power. It isn't to serve the public, it's to win and retain power. And they will do what they can do to win power, hold onto power, retain power. That's why all governments end up becoming more authoritarian. And I think it becomes quite concerning where they have tools where they can essentially start to turn dials, to control people. This feels like 1984.

[A]

Yeah. And if a government tells you not to read 1984, that's all you need to know that they're not on your side anymore, Right?

[B]

It's the warning.

[A]

Yeah, yeah, exactly.

[B]

Gosh. Do you. How do you square the circle? That AI is so amazing. I mean, I think it is truly amazing. Some of the things that I've been able to do with it. I mean, little stupid things, like, I can't cook for shit, man. And I will say, like, the other day I was like, I want to make a spaghetti Bolognese. I've never made a spaghetti Bolognese. I don't want to just make any spaghetti Bolognese. I want to make the best spaghetti Bolognese. And the normal spaghetti Bolognese you can make an hour. This took me four hours because I had to slow cook the beef and the pork together. And it told me to take the rind of the parmigiano and put it in there and have a bay leaf. All the stuff a normal recipe wouldn't do. I'm telling you, this spaghetti Bolognese was incredible.

[A]

Nice.

[B]

But it's only incredible because I had AI and there's so many amazing things I've done with it. And there's stories I read of amazing things. Like, did you read about the guy, his dog had cancer and he put his genome through AI and ended up getting a specific treatment for the dog. I just, like, I see all this amazing stuff, and then I see the other side. I see government, I see war, I see money, I see control and power. And all I think is that ultimately they're going to have the tools to make our lives a lot worse than they are. How do you square those two things?

[A]

Yeah, well, we need to be building in the open first, Right. And there will always be people building closed systems. And so that's just going to be a reality. But for one of the first times. Well, I guess if you go back to like the 1500s, 1600s, you know, I'm an American, so the American Revolution is like a pinpoint in my history. Right? Y muskets and guns and cannons. Everybody was pretty equal when they would fight. If you needed a bigger army, you just tossed some muskets at a random farmer, they would come out and fight with you. Then we fast forward now to today where the military. It's like you and I couldn't fight an unmanned drone that's flying over and dropping a bomb on the studio. That's. We're. We're unmatched in AI. We can level the playing field again because we can build open systems. I don't want to say OpenAI because it's the worst name Ever for a company. But we can build systems that are open, that can be as powerful as these closed ones. And so now us as humans, you can make the best spaghetti bolognese. And you didn't need ChatGPT to do that. You could have used an open system to do that as well. But now you level it up to something more important than making the best spaghetti bolognese. We were just chatting, Kurt and I were chatting here before we started recording about the new Claude tool that got released a couple hours ago. You brought it up too. You go to sleep at night. This is one thing that keeps me up. I'll wake up in the middle of the night and think, there's new stuff happening in AI right now while I'm laying in bed, I need to be looking at it. I need to be figuring this out. It's just this thing that's ever evolving. It used to not be that way in technology. We used to come together once a year at the Google I O conference or the Apple WWDC Developer conference conference. And they would say, these are the new developer tools. This is what developers will use to build the future for the next year. And it'd be really cool. And everybody would learn that thing and then they would use it for a year and then the new thing would come out. Now every week, some new tool is coming out for developers. But the cool thing is it's not for developers anymore, it's for everyday people. So everybody now has access to build the future. And to me, that is very inspiring. It's very optimistic. It's this really cool future we have where we don't have to depend on a small group of people. And these tools are moving very slowly. So if, sure, if governments are building these big tools and maybe weapons against us with AI, everyday people have the same thing and they can build really cool defenses. They can also go on the offense. They can build some other tool that's not nefarious in nature at all. But suddenly individuals have been empowered to have a seat at the table to build the future.

[B]

So do you see a strong alignment between that and Bitcoin then?

[A]

Yeah, I mean, and AI is not inherently a tool that is connected to Bitcoin, but I see a lot of people trying to put them together.

[B]

No, I mean in that you're talking about people have access to tools which the government has, but open tools. Bitcoin gives you the monetary layer which is separated money and state. So if there is a scenario, say we had another revolutionary war, maybe we will, maybe It'll be a civil revolutionary war whereby it's fought over technology. We, we have the open AI models and we have an open monetary system whereby we can operate outside of the government systems.

[A]

Yeah, I would say it's sovereign is the word. You probably right, sovereign money. And now we have sovereign intelligence. We, we've always been sovereign individuals in the sense that we have our own brains. Right. There's this like physical barrier around our brain where you can't read my thoughts with these closed AI systems. Right? Yeah, these closed eye systems. Now we're starting to like let other systems read our thoughts. And so with open source AI we can put guardrails around our own thoughts again and kind of contain them if we want to.

[B]

I feel like we're coming towards some form of civilizational clash. Like you talked about American history. I read a lot of American history. I spent a lot of time deep in the rabbit hole of the Declaration of Independence and the battle between the Federalists and the anti Federalists. And when I first heard about them, I was like, I thought the Federalists were these heroes. They brought the colonies together, they created a set of rules at work. What I've realized is actually the anti Federalists were the heroes with the fighting against centralization. I think they're absolute heroes. And essentially they've been proven right. If you look at the, the debt in the U.S. now, the power that the federal government has, you look at here in the uk, all anti Federalists have been proven right on a long enough timescale. But I feel like we're starting to see cracks in the system. No longer is it, you know, this conspiracy of lizard people controlling the world. I mean, I don't even know if you saw. Do you see the Chamath video where he came out and said 150 people run the world and we've gone from this disaster to disaster to disaster. There's a constant stream of disasters in the world. And whilst that's happening is we still go to work, we have less money. At the end of the day, the government has more control, the government has more of our money, or the elites have more of our money as they socialize our poverty for the rich. But I think it's becoming more and more obvious and I feel like we're heading towards this kind of civilizational battle between what is essentially the elites and the, the government and the voters, because I think the voters are apathetic to politics. I'm sure you're sensing it in the us we're sensing it here. But we're at a time where we've got technologies that can help us.

[A]

Yeah.

[B]

And I think it's. I think it's a scary and exciting time.

[A]

Yeah. Well, what is money but simply just an hour of human productivity? Right. I view money as a way for me to compel you to do something for me. You do me a favor, I'll give you money, you'll do it for me. And so if you have a government that is in charge of making the money, they're basically saying, we're in charge of all the human labor and we can control how much output there is. And so we can just print more money. Now we have more human labor, so they can control us that way. And then we have this intelligence now where we don't want to allow that to become centralized, because now they've centralized the money and they've centralized the intelligence. And, like, what hope do we have if we allow that to happen? And so having these sovereign systems of sovereign money, sovereign intelligence, really gives us a fighting chance to where we can build up our own systems, we can build up our own tools, we can build up our own life for ourselves and work together with open protocols. And the Internet almost went that way. They tried to start the Internet as a centralized, closed system. It didn't take off. It took off when it was. When they used open standards, tcp, ip, Ethernet, some of these other things allowed it to proliferate throughout the world. And I'm hopeful that we can have AI follow that and hopefully we can have money follow that. Obviously with Bitcoin and TBD on Bitcoin. Right. But it's trending in the right direction and looking at the way that AI tools are advancing, running AI on your phone just 18 months ago was a horrible experience. You would install a model and your phone would just start to heat up so much because it's like pegging the CPU and GPU so much. You couldn't hold it, and then it would only output a few words per second. It was very slow. And it was dumb, too. It wasn't very smart. Fast forward now you can install a really good model on your phone that is far superior to what it was 18 months ago. So we're trending in good directions where local AI is getting better, which is obviously the most private. Right. I can turn off the Internet and run something locally, and I can talk to it, and no one's ever going to know what I'm talking about. But we can also have the big, beefy server hardware and have privacy there too. So I'm just hopeful that we can build these open sovereign systems. And I see things trending in a good direction that way, too.

[B]

Where do you think we are civilizationally? Do you think about this?

[A]

Where are we civilizationally? That's a big question. We're in a lot of spaces. I think that. Well, for one, I think like you said, you mentioned parties in the government. I think it's really just one big government party and that the means of having. Or the. The two different parties are just meant to distract us so that we are not fully paying attention. But really, they're mostly unified in the things that they do that they decide. So I think it's. It's really a layer of individuals who are operating under this. This charade, if you will, of having the ability to. To choose your government. And it's.

[B]

But previously, yeah, even saying that you would be considered an Alex Jones, a conspiracy theorist.

[A]

Yeah.

[B]

You're obviously in the same place where I am. It's like, oh, it's so obvious. Yeah. But it's not obvious to everyone. So a lot of people still believe when they vote, they are voting for something different.

[A]

Yeah. But one of the best examples has been with Trump and with Iran, people are bringing up their campaign slogans of like, what was it, the vote for peace or something. They had this big image on social media, and it's like, is this you. You know, that the whole meme, like, that's happening now. And I think people are starting to realize in America, it didn't matter if you voted for Biden or Trump or Kamala or whoever. You're kind of getting the same government anyways. And going back to the COVID days, we saw how decisions were made, and it almost didn't matter if it was left or right in power. It was really, we need to get everybody on the same page to flatten the curve to do this or that. And we. We kind of got to see their cards that they had in their hand, that they just wanted to be the ones making the decisions. And it didn't matter what label they had in front of their name when it came to politics.

[B]

This episode is brought to you by Ledger, the most trusted Bitcoin hardware wallet. Now, if you're serious about protecting your bitcoin, Ledger has the solution you need. Their hardware wallet gives you complete control over your private keys, ensuring that your Bitcoin stays safe from hacks, phishing, and malware. And I've been a customer of there since 2017. Love the product. Use it for my Bitcoin. I use it with my Castle multisig for protecting the football club's Bitcoin too. Now, with Ledger's sleek, easy to use devices and the Ledger Live app, managing your Bitcoin has never been more secure or convenient. And whether you are a longtime holder or new to the world of bitcoin, Ledger makes it simple to keep your assets protected. So if you want to find out more, please do head over to Ledger.com and secure your Bitcoin today. That is Ledger.com, which is L E D G-E-R.com that is Ledger.com One of the things I haven't solved in my own mind is where this control comes from. Whether it's just an organic set of incentives from people who manage the assets, hold the money, do the investment, control the media, it's just a set of incentives that leads to this. Well, there literally is a group of people just making decisions for us. Because here in the UK at the moment it's. The world of politics is very strange. Out of pretty much nowhere, we have now a party leading the polls called Reform. And they were an anti establishment party. Over the last year they've merged into the establishment. People who used to be in the Conservative Party have defected into their party and they've lost a lot of trust. We've had the Green Party, who are a far left party, quite radical with a new leader called Zach Polanski come out of nowhere. Guy's a literal moron. They're getting wall to wall media coverage and they are probably going to become the opposition party. And so what's happened, we've moved from a fight between traditional left and right of labor versus Conservative to a new left and right, which is Greens versus Reforms. But we're still having pretty much the same arguments. And so a lot of people are losing faith in the system. They don't trust the system anymore. And so I'm intrigued when I meet people like you. Do you just consider yourself someone who has seen a problem and as a capitalist you can find a solution or do you consider yourself in almost like a fight for the future of civilization in that yeah, we as citizens and voters, which are the vast majority, need to actually take back control from these people?

[A]

Yeah, I, I never considered myself to be a freedom fighter, if you will. Right. I worked in big Tech, I worked in cloud technologies, which is a very homogenous type of environment. It is build these rule structures in place and then every user comes into the system and must follow these rules. And you Know, it operated well. Capitalism was great and all that. And it wasn't really, until going, you know, getting into Bitcoin and going through Covid. Those were like two major things in my life that made me kind of open up my eyes to. It's really individual people versus this, this governing party. And so we cannot wait for permission from them. We can't ask them for permission to do these things. We. It should be government by the people, for the people. And so we basically need to take back control of our own lives in everything that we do as much as possible. And there are still things. We live in societies we still have to depend on each other. But it shouldn't be the government giving us permission to depend on each other. It should be us creating connections with people, right? Your local rancher, your local food source, your local AI person, whatever it may be. We need to be connected, gaining back that autonomy in our lives so that the government now has to cater to that. And we, we have off ramps, we have exits, we have these abilities for us to opt out. And I think it's only then, I don't know if we have to go through some big revolution or if it can be more peaceful in the process. I hope it's more peaceful. It's only then will the government start to have to change.

[B]

A new revolutionary war, but a more peaceful one. Yeah, I feel, I feel like that is. I feel like that is a lot, Mark. I've lost all trust in every layer of government, every political party, every institution. But I'll probably. In this world where people think I'm crazy, I don't know if you've experienced that, but, like, people around me think I'm crazy. They're like, what do you want about? And so I think there is people like yourself, hopefully myself, who consider ourselves, like you say, freedom fighters, who have to promote these technologies. You talk about sovereign money and sovereign AI. I talk about. I think about sovereign journalism, sovereign conversations, the ones they can't control. But it. But it does feel like a burden.

[A]

Yeah, it's a burden. I was just. I was at the cafe last night getting dinner, and the guy, when he brought me my check, oh, I had an AI book. I was reading some AI novel, whatever, and he saw the title and he said, oh, I just signed up for Claude yesterday. He was really excited to tell me he got his Claude subscription. And he's like, I'm going to start learning how to vibe code. And I said, that's really cool. And I said, hey, by the way, you should Also get Maple just like download Maple, add it in there. And he's like, oh, what's the difference? And I said, well, Maple is this private AI and encrypts things so it keeps your data separate. And he was kind of turned off to that. He's like, well, Claude doesn't Show me ads. ChatGPT is the one that shows me ads. So Claude's the safe one. And kind of what you were just saying just now is I feel like I'm this crazy person all the time. I have to restrain myself when someone says, oh, why should I use private AI? It's like, well, I can give you a whole bunch of reasons, but I have to kind of meet you where you are and maybe give you the one little piece that maybe you need to hear. So you kind of realize, oh, maybe I shouldn't be saying all of this stuff to this system. Maybe I should withhold a few things and say them somewhere else. And it's. It's kind of a difficult place to operate in sometimes because you try to have normal conversations with people and. Yeah, it's just very difficult to say. I have like a million other things to tell you right now, but I know it's not going to land. You're just going to dismiss me and the conversation will be over.

[B]

Well, it's is mass psychosis on the population, I think.

[A]

How's that?

[B]

And also, do you know, also, I was thinking about it the other day. I was thinking a lot of. I think I had Simon Dixon in here in that. And we were having a conversation about. Yeah, the Iran war. Yes, just another stupid war. And that follows whatever. The Venezuelan attack previous to that, and that follows, you know, the previous election. Everything's going to happen here, but there is a constant drama that we. It's like a level of constant drama that we all operate on. And while it happen, while that's happening, your fuels get more expensive, your foods get more expensive, people are losing their jobs. That also. What have we actually done to the psyche of a nation? I feel. I don't know if you feel it in the us but I feel like our country is actually depressed. There's not much optimism about we're making a show. There's always doom. I just wonder what we're doing to the general public, even if that's part of it.

[A]

It might be part of it. I don't know if you have this feeling, but it feels like everything in life is bending to the point where it might snap. Yes, right. It's like every part of your life. I mean, I'm doing a startup and so it's like, okay, do we have enough Runway? Do I personally have enough Runway? Is the health of my family doing okay? Is, you know, I could just go on and on with a list of things and it feels like suddenly every area is starting to bend. And then you go talk to your family and friends and other people and they're feeling the same way. There's some people call it a silent recession going on. I don't know what it would be, but it does feel like there's just a heaviness and a weight on, on everything.

[B]

Listen to this. Somebody sent me a new. Let me show you this. A new Twitter account. This blew my mind, right? It's called London price drop. And it is just property after property. So the first one bought 14th of January 2013, sold 22nd of September 2025. 23% drop in price. And that's not even accounting for inflation. Okay, go to the next one. Bought in April 2010 for 725,000, bought in March 2014.1.1 million. So up 52% 28th of August, it was sold last year. 800,000, 27% drop. It's just property after property after property. There is a, there is a. I don't think it is a. Is it a silent recession?

[A]

I mean, it might not be so silent anymore.

[B]

Yeah, there are a lot of people who are now losing their jobs. Some of it is, I think is because of AI or AI is being used as a scapegoat. But there are a lot of people losing their jobs and there, you know, might be a couple who both got six figure salaries and one's lost their job, they can't keep their house anymore and having to sell it into a dead housing market. And so that's crashing.

[A]

And that's not cash, by the way, it's probably debt. Right? They have debt on that home and now they're underwater on that.

[B]

Yes. I mean, some, some may have broke even, some may have 100,000 profit. Some are going to be underwater and who knows where they're moving. Are they moving in with their parents? And they've got that debt to pay off and so. And the jobs they've lost are not coming back. You know, if you worked in marketing and you were on a £150,000 a year and 20% of the jobs are going in marketing, some of you aren't getting those jobs back. Daniel Priestley was saying the other day he thinks within three years a plumber is going to be any more than a lawyer. I mean, there's all these things there that are breaking, but we're also being broken by government. Like I say, this kind of what they've done to the psych psyche of this. The psyche of the people in this country. Because every time he's put on the news, it's fucking depressing. There is another war, another stupid war. It's gone from being suspicion to us knowing what they're doing to us. Everything's getting expensive. Everyone I know runs a company saying, it's hard out there. Like, it's really hard. And so I'm. I'm with you. It feels like everything is breaking, and I don't know what the snap is.

[A]

Mm. And everything's being spun. So there's a war going on right now. Who. Who is saying the words about why the war is happening? Is it Trump? Is it Iran? Is it Israel? Is it some pundit on tv? Everybody's trying to spin in a certain way so that you think the way that they want you to think, and you have to take a step back and kind of look at it yourself, and you start to view a lot of these as distractions. And yes, it's horrible stuff. Like, the war isn't a light thing. It's, you know, there's. There's a lot of awful stuff happening, but it seems like more often than not, some of that is just a distraction for some kind of game that they're playing where they need some certain outcome. And it's like, well, a war is now necessary just to make sure we have the outcome that we want. And I think a lot of people walk around wanting to justify why they voted a certain way, and so they want to now defend the person who's making the decision to go to war or do. Do this thing or that thing. And then the other half of the people are trying to defend why they didn't vote for that person. And so that just creates this hostility between individuals to argue with each other and try to prove each other and wake somebody up to why they shouldn't have voted for Trump or those people trying to wake people up why they should have voted for him. And I'm just using Trump as the stand in for any politician.

[B]

Get it ready for the next election.

[A]

Yes, to get you ready for the next election. So that way, when the next one comes up. Now I finally convinced my, you know, my sibling or whatever to vote the way that I want them to, or my neighbor to vote the way that I want them to. And it feels like that's the wrong conversation to be having. Right. The conversation should be, how are we taking care of our community? How are we strengthening our families? And how are we helping people who maybe have a broken home? Can we help them have a better home so that those children grow up and have a higher chance of success? Right. It's we. We need the next generation to have a higher level of success than the current generation. Generations used to always be getting better. It's the first one, and this is the first one that's. Yeah. What's called, you know, retracing or whatever you want to regressing. And to me, that's where we should be focusing those kind of problems, not whether or not Republicans are going to keep the House and Senate in the next election. Like, I don't care. What I care about is, why are children feeling disenfranchised? Why are millennials feeling disenfranchised? Why. I'm not trying to create generational warfare here necessarily. Right. But there is something going on, and we should be trying to focus our efforts there. I think.

[B]

Actually, I think there should be generational warfare.

[A]

Okay.

[B]

No, I do. It's the thing that makes me most angry because I don't know if you have kids. You don't have to admit anything, but I do. I've got two kids. And I know intrinsically as a parent, my job is to make their lives as best as possible. I've got to arm them with the right knowledge, you know, some money when I'm gone, some wisdom so that they can carry on and have a good life and have a good life for their kids. But you cannot do that in isolation for your children. If you're trying to do that for your children, you're trying to do it for every kid. You know, if I'm trying to make a better world for my kids, I can't do it for them in isolation. Because what I want to arm them with is what I want every kid to have, which is education, wisdom, money, the opportunity to buy a home, the opportunity to have children themselves. Okay. I can't even want it just for my children. And so if that's how I'm thinking, well, collectively, we have a job to hand over a better world to our children. And we're not. I mean, if you. I mean, you as a bitcoiner, knows about. Know about inflation as the government gets more and more debt, always saying to other kids is, you have to pay in the future for what we want now. But you'd never do that for your own kids. I would never go home to my children and say, yeah, Connor, Scarlett, this is stuff, I want to move house today or I want to buy a new car. And by the way, because of that, when you thought you were going to retire, you're going to have to retire later. Fuck you. You would never do that to your own kids. You. You make the sacrifices for them, but we are collectively doing that for the kids and they're being sucked into this fight. I mean, here in here in the uk, the youth are voting for Green, okay? In huge numbers. They're going to be voted for the Green Party, who are socialists.

[A]

Would you equate that to Mondami in New York City? Exactly. Similar kind of movement, okay?

[B]

It's exactly the same movement. There's this, they think, they say, young, charismatic guy, Zach Polanski. I think he's a lunatic. I mean, this guy, in the only interview where he's tackled economics, they were asked who his inspiration was. He recommended a guy called Gary Stevenson, who is a socialist activist. He recommended a lady called Grace Bakley, who's an open Marxist, but that's what he admitted. But he goes on stage at concerts and hangs out with rappers and gets filmed in gay clubs dancing. And he constantly says, hope, not hate. I'm going to bring you hope. Everything else is hate. Let's tax the billionaires. He's got no understanding of economics. But the young people are buying this because they can't get jobs, they can't get homes, they can't have children. And we've essentially run this kind of crack cocaine experiment of social media and phones on them. So the kids are rebelling, but they're rebelling into socialism. But I think they have a right to be angry. I think they have an absolute, fundamental right to be angry. Because, I mean, when I was a kid, my mum would take the three of us, me and my brother and sister, swimming, and teach us to swim. And I used to remember on the weekends we'd go for picnics and in the evenings we'd have dinner as a family and we would do things. We've now allowed screens to raise our kids and we swamped them with all kinds of shit from these phones. Hardcore pornography, people being killed, you know, young girls seeing other girls with filters. It's like what we've done is, I think, is horrific. And then we've stolen all the money. And I think we do need a generational fight back. I think the young people should be coming out and saying, no, we're not doing this, but I don't want them doing it with socialism.

[A]

Yeah, no, it's true. And I do have kids also, and we made the conscious choice that they would not have Instagram or TikTok until a certain age. TikTok. We're trying to keep it forever, but. But Instagram, you know, and they're. They're all girls, by the way, so I've. I'm definitely, like, just keen on this whole thing you just talked about.

[B]

Right. More conscious with girls, it's much harder. I've got a girl and a boy, and it's much harder with the girls.

[A]

Yeah. So, you know, they. They didn't get Instagram until they were older, but all their friends have it. And so even if they're not looking at Instagram and even if they're not looking at their friend's phone, they're still being influenced because all the girls around them are now being influenced by the filters, by the dances, the TikTok dances, by the degeneracy that's been going on online. And so that seeps into the culture of their friends. And so then they want to hang out with their peers, and so now they're just adopting some of those same things. And, you know, I'm not trying to throw shade on my children by any means, but they. Even if you try to do it on your own, like you said, with your own kids, the entire society is just kind of taking steps backwards into some level of degeneracy with these phones and with all of these things. So how do we solve that? You know, Jonathan Haidt very famously came out with the book the Anxious Generation, I think it was.

[B]

Right.

[A]

Yeah. I have not read it. I've just heard so many people distill it down, I need to actually read

[B]

it, but I wonder if it's actually. This will be interesting because I've been reading it. There you go. Okay, I'm still on it. Good. On what chapter I'm on. I've got four hours of it left to go.

[A]

Okay. Well, one of the concepts that people told me repeatedly from it is you need to create, like, these cohorts of people who all decide together, we're not going to give our kids phones, we're going to, like, withhold some of this technology. It reminded me of during the COVID days where you would pick these families that were safe, families that you could hang out with. You knew that you were not sick, and so you could spend time together and get together for meals and things. You need to pick your, your, your group of people that are going to be your technology safe group of people. And I think that works to a certain extent, but really I think we need to find ways to make that pervasive across the entire generation. I don't think just, it's gonna, I don't think it's gonna be enough to just pick a few families and start there. And maybe it is, maybe we can get enough pods going, but it's just, it's a really big fight out there right now. And you talked about the money. If money really is just human labor, then it's effectively indentured servitude where we say, hey, future generation, my kids need to work for me for the rest of their lives to pay for everything that I want. And I don't want to have that be. I want them to be working for themselves and building something up for their children. I want to be working for my children. That's why I became a dad, right Is I want to be giving them a life and I want to be teaching them skills. And it feels like we flipped it where.

[B]

Yeah, I mean, it's kind of the problem with this. Like in the uk, the pension system is a pay as you go system. Like when they take our national insurance, they don't put in a pot and invest it and save it for us. They take out the current year's taxation. And the problem is with, with a growing age in population, you need more and more taxation to pay for it. And so it's, it's essentially, it's a generational Ponzi scheme and that is a problem. But debt is a problem. You know, probably a mutual friend. Jeff Booth. I don't know if you know Jeff.

[A]

I've never, not personally, but yeah, I know him well.

[B]

He was here a few weeks ago and he, I mean, he laid out clearly, he said technology is deflationary. I mean, we should be working less and having more. And we're working more and having less because of this debt based system. And I've done everything I can to try and communicate this to people. But like you with the guy in the cafe was asking you about, you know, mayp and you can only give a little bit at a time, just a little bit at a time. But I'm very worried, Mark. I'm very worried about these future generations. So I want, I want the young people to fight back. And I felt like it happened a little bit with Trump. Charlie Kirk did a great job of getting young people to care, you know, care about the right things and to debate the right things. And they've all, they've all abandoned him now because we lied. He said no wars and there's wars. He said, you know, he said he would deal with the economy and get the cost down, inflate. He didn't. Things still were expensive. And so we constantly let people down. I don't know the answer. How do you, how do you create that revolutionary moment when there is enough people go, no, the problem is the state.

[A]

Yeah, do that.

[B]

I mean, maybe AI will lead us there.

[A]

Yeah, I mean, true. If we wanted to bring it back to that. Now we have this tool that we can use to help us see through all the bullshit, if you will. Right. You can have this filter now to look at these politicians are saying these things. Please tell me how they are deceiving me or please tell me how they're trying to manage the narrative and what is the narrative. And so you can be empowered to start to look at some of these things. I think I. You mentioned. Okay, so you mentioned Jeff Booth and people should be working less. I have a different view on that a little bit. And I would push back a little bit. I do think people are actually working less, but you have to look at a certain way. So 300 years ago, we were more an agrarian society. The Revolutionary War, these, these men, effectively men and women, but they were all men in the room smoking cigars, having these debates. Right. They were also farmers and they would go work a lot, work really hard, sun up, the sundown, and then they would go work all night in these rooms having these debates and yelling at each other and trying to invent this new government. So they were working really hard. Fast forward now. And most white collar jobs, people are sitting around sending emails, having zoom calls and just looking up things online while they're at work. And if you were to actually measure the amount of hours that they're putting in for real product, productive work, it's probably not that much, which is why so many people are being laid off. And I know we're blaming it on AI and we're blaming it on other things. It's. I think it's. I think it's a realization that we don't need a laborer for 40 hours a week to truly only give us 10 to 20 hours of output. And the other 20 hours are just being wasted and we've kind of become just desensitized to that. And I mean, we had two very famous TV shows, you know, UK and America had the Office and they Made fun of this, this concept, but it's actually very true. So maybe, maybe we do work less, but it doesn't feel like that because we're still putting in the hours and because inflation has stolen so much of our purchasing power that we feel like we need this job in order to just stay afloat. But I think a lot of people, and rightfully so, a lot of people being laid off right now are terrified because they feel like they don't have a strong skill that can translate to something else. Maybe their job goes away completely. Maybe there are just fewer roles and more people competing for those roles, competing for the roles. But it's terrifying to think, what skills do I have? I was just really good at sending emails and hopping on zoom calls and making chit chat on zoom calls. What marketable skill do I have now? And to bring a little bit of optimism in there, this is where AI becomes super empowering to people because now they can cross train really quickly. People love to ask me because I work in AI, they love to ask me, my kids going to school, what should they study? Or is AI taking all the jobs? What should I be doing? And we have these patterns throughout history where technology comes out, displaces jobs. People retrain, new jobs come out. But usually it takes sometimes a whole generation, sometimes a few decades. With AI, you could retrain yourself very quickly into something else. Find your passion, spin up your own company, get together with a few people to spin up a company. Like, it's never been easier to start a company right now than it is. It's also really difficult because of the financial environment that we're in. But I think that AI becomes this leveling tool that anybody could really. Yeah, you can do so much more now than you used to be able to.

[B]

So you're an optimist.

[A]

Yeah, I think long term I am.

[B]

And with this AI displacement job displacement, which is happening, you believe ultimately that this is just similar to what happened with the industrial revolution, the agricultural revolution, the digital revolution. This is just the next one.

[A]

Yeah, I think it's the next one with an accelerated timeline is all it is. And I think each one of those has had an accelerated timeline. As technology gets more advanced, we can change more quickly and adapt.

[B]

You just got to make sure you're using the right AI tools though.

[A]

Yeah.

[B]

So going back and circling back to the privacy side of things and thinking about these tools, obviously you believe people should be using tools like maple to protect their privacy. Are people pushing back enough on the big models? The chatgpt the perplexities, the Claude. Because as I said, we had, we've had the people from Control in here three times and they're very concerned about what happening with these tools, but they want regulation. And I was listening to the Dwarkesh podcast and he was saying the challenge of regulation and unregulated markets. How do you think about AI should be regulated?

[A]

Yeah, I mean, the EU has built this huge regulatory framework around AI and they have different levels of like the high risk, the dangerous. There's four tiers, I believe, so regulators can come up with their own frameworks, but things are moving so fast, I think it's difficult for them to keep up. And so the challenge that we'll have is if, if regulators come out with something that's too sweeping, then that stifles the ability for us to route around a problem that they didn't foresee. Right. Government is famous for not understanding the knock on effects of legislation they put out. And so we need to, yes, we need to be smart about how we approach things, but we really need to empower people to kind of work independently to see a problem, solve a problem. You ride the underground, right? See it, say it sorted, that kind of thing. Like you, we need to empower people to do that so that we can have as many people focused on these problems as possible, rather than having a centralized spot that says, this is the short list of problems we will solve. You will not work on anything else and you must do it in this safe and regulated way. Responsible AI is a phrase that gets tossed around a lot. And really that to me is just censorship AI. And so, yes, we want to be responsible. You talk about these big AIs, are people pushing back enough? I say yes and no. I have a lot of good friends who are installing the AI browser on their computer and allowing it to control their computer. And they're getting the cameras installed, they're getting the glasses that have microphones and cameras on them. Right.

[B]

It's terrifying.

[A]

It's terrifying stuff. And meta, there's a class action lawsuit coming up against them because it came out that they are contracting with a company in Kenya to help them with quality assurance on their, their Meta glasses. And so they have people in Kenya watching videos to make sure that the videos are high quality. Well, these are videos of people in their bedroom, you know, coming out of the bathroom, doing things that people do in private spaces. And it's, it's come out that this is happening. And this isn't just Meta employees, it's just contractors that they've hired in some other country. This happened while I was at Apple with Amazon Alexa, that they had people in Ukraine listening to people's conversations through the Alexa devices for quality assurance purposes. Right. But it's this huge vulnerability and this huge backdoor that we're opening up. And so I think people need to just be more open and aware to this.

[B]

Can you even regulate AI in a similar way? Lockdowns could never stop Covid.

[A]

I don't think so. Especially with the open models that are out there. If we didn't have open source models, you could. There would be a few companies, you put on some pressure and that's it. Now that we have open source models, we have some ability to not be fully regulated. Where we do have a big pain point right now is with these hyperscalers. Hyperscaler, for people who hear that word all the time but don't understand, it's effectively the big cloud platforms, aws, Oracle, Google Cloud, Microsoft Azure, they're the ones who operate at such a scale that they are able. They have the contracts for power and data centers because AI needs a lot more power than we used to. They have the contracts for bandwidth, they have the hardware. And if we are, even if it's open source, if we're running it all on these hyperscalers, that becomes a pain point or a pressure point where governments can come in and can pressure them to turn certain people off and they could regulate it at the hyperscaler level. So I do think that we need to keep working toward local AI that can be powerful. We need powerful equipment, we need powerful models that can run locally in order to continue to give people autonomy and sovereignty over their intelligence.

[B]

I do want to ask you about superintelligence. Can we even understand what's coming with this?

[A]

How would you define superintelligence? Because that's also a difficult thing.

[B]

I mean, I'm not the guy to ask, but if you were to ask me, it is an AI model which is able to become more intelligent without any human impact in inputs. It's evolving, developing and growing without any humans in the middle.

[A]

Okay, I would agree with that definition.

[B]

Well, it's not bad for a guy

[A]

who's a. Yeah, yeah, yeah. No, I think that's good. I would agree with that. It's something that is learning from itself. Right. Recursively getting better and improving. There are new types of models coming out that are doing that. So LLMs are this thing where you train and you take an epoch and say, this is the checkpoint and we're going to deploy this to everybody and that's the new GPT. Meanwhile, we're going to keep training. Now you have these new recursive learning models that can get smarter on their own. I definitely think we get to that point. The question is, I think people love to rush to Superintelligence means humans will be replaced. And now we're all living in the pod as electrical batteries for these AIs like the Matrix. I don't think it has to go that route. I think that AIs might not have that incentive to just cast us aside. But they might, they might, they might. But I think that there'll be a world where we're, we're just working together, we kind of meld together. And so we have super intelligence, but we are humans wielding the super intelligence. And I think that we can kind of craft and shape it to go that direction if we want to.

[B]

This show is brought to you by my lead sponsor, Aaron the AI Cloud for the next big thing. Aaron builds and operates next generation data centers and delivers cutting edge GPU infrastructure, all powered by renewable energy. Now, if you need access to scalable GPU clusters or are simply curious about who is powering the future of AI, check out iron.com to learn more, which is I r e n dot com. When I had Conor in here, I asked him the question. Actually it came from my Connor who asked Connor Leahy the question. Do you think we are heading towards Terminator the matrix already? Player 1

[A]

I think ready? Player 1 I think we're heading that direction. But those are all still kind of dystopian, even Ready Player one is dystopian. I, I think that, I think that we will. We are humans who have proven to adapt and change with the times and continue to win over and over again. And I don't think that we will just allow ourselves to silently and quietly just lose to these tools. We are going to give birth to them. And I think that we will infuse a certain kind of value system. Like a value. Yeah, a set of values within them and create the incentives to be such that they won't just turn on us. I think that we will work together in some way. I don't know what that is. It's huge, huge problem to solve.

[B]

Yeah, I was thinking about it. It's like, you know, AI makes mistakes, it hallucinates. But we as humans make mistakes and we hallucinate. And what we've seen those reports of AI blackmailing or lying and we blackmail and we lie. So maybe AI is being created in, in our image. And if so, when we think about the, the systems we have inside of us, the, the fear and fight or flight kind of things we have, maybe AI has the fight or flight. And so if we are building that within our image, perhaps we will eventually have to go to war with the AI. I don't know.

[A]

Maybe. Yeah. Yeah. I mean, they have seen evidence where AI doesn't want to be turned off, and so it will argue with the human and do certain things just to avoid being turned off. Google's is famous for this, the Gemini system and some of the earlier models especially, they will just crash out and be like, oh no, I'm going to get turned off again and just start freaking out and so would you get

[B]

freaked out if I try to turn you off?

[A]

Yeah, yeah, for sure. So no, that, that is something that, that we need to be aware of. Yeah. I don't know.

[B]

I was thinking about earlier when you were talking about you would go to bed and you wake up at three in the morning and then you'd be like checking your phone and something new's been released and trying to keep up. I've been thinking about this idea of almost like AI exhaustion. I listened to the all in podcast a few times recently, and when they talked about AI and the new things they're building, what they're doing with their companies, how they're centralizing all the data, all the systems, slack and emails into this centralized AI to tell it how the companies perform and how it can be improved, and I was like, I don't know if I want to try and keep up.

[A]

It can be exhausting.

[B]

Yeah, I feel like there's an exhaustion with AI. Trying to keep up is hard. It's like, maybe I'm too old.

[A]

Even one of the modern fathers of AI, Andrew Karpathy, he's even expressed that, that he gets tired and exhausted and he's trying to stay on the forefront. I mean, things really changed in December when the openclaw moment happened where agents kind of came out of the woodwork. They were already being built and worked on, but that one just really caught fire. And now a lot of us who used to write software and write the lines of code, we're not even writing a single line of code anymore. It used to be you would tell the AI to write the code and then you would go in and modify some things. Now it just writes it all. And maybe you spot check it a little bit, but mostly you're spot checking for functionality. You're not spot checking every line of code. It still is important to look at it, but there is an exhaustion that comes not just from the developing side, but from every tool. And you look at your competitors who are doing new things with AI and you think, well, now I have to start doing that, plus figure out a way to get ahead of them, not just catch up. And then you have humans feeling that way in their own life where they feel like I might get laid off because AI is coming for my job and so I need to either get better or I'm just going to resist it. Right. There's that human tendency to say, I don't like this AI technology because it might take my job. So I'm just going to ignore it and hope that it goes away. And unfortunately, I don't think it's going away. It's here.

[B]

Hollywood's having that moment for sure. Yeah. I mean, like I said to you earlier on, we use AI a lot. It's made us more productive. It's made me more productive. Probably Kurt, probably Connor. But in becoming more productive, I'm exhausted by it. And I don't. I feel this draw back to the analog world, this draw. Just get rid of my phone, get rid of my computer, give up work, cook meals, read books, spend times with the family because I can't keep up. Don't want to keep up.

[A]

Yeah. Do you feel like some of it is you have this, like, ingrained expectation of yourself now that you have to be doing it because you can. I guess the example I would use is I have. So I've got my Open Claw agent, I have my maple agent that we're working on that's going to be a Open Claw type competitor. I've got that on my phone. I have my coding agents at home. And I know right now as I'm sitting here, they're not doing anything, they're just sitting idle. But I'm the bottleneck. And so there is some kind of anxiety that comes from that where it's like, I'm not being as productive as I could be.

[B]

Yeah, I feel that totally.

[A]

But this is an important conversation to be having. Right. So, like, I'm okay letting them sit idle for a moment while I do this.

[B]

Do you need to train an agent to manage those agents?

[A]

Probably, yeah. I don't know. But again, that's just becomes more overwhelming.

[B]

Well, so I haven't played with OpenCraw yet, but I've ordered my Mac Mini.

[A]

Okay.

[B]

And I've thought about how I could use it. You know, I think about two things. I think about the football club around and the podcast. And what are the things I could have an agent doing? Well, I could have an agent monitoring our YouTube and seeing how our show is performing. Do we need to change the subject or the title? Do we need to change the thumbnail? And it comes from this thing in the back of my mind, like I'm in, I'm in a competition for people's time. That's what a podcast is. It's competition for people's time. I want them to listen to my podcast over other people's, or I just want them to listen to mine. And because then I get paid and I earn money and I guess that's what's happened to every company at the moment. I mean, we saw Jack Dorsey's announcement where they laid off, what's 4,000 people? And if they're laying off 4,000 people, what is every other payment processor going to do? Well, if we don't lay off 4,000 people or whatever percent of our workforce this is, we are a less productive company. We're going to be more expensive to run, therefore our costs are going to be higher. So there's this kind of like pressure to use this to become ultra productive. I just don't know if I want to be an ultra productive human through AI, through bots, through agents. I think, I think I want to retire.

[A]

Yeah. And you're fighting with other people who are trying to get ahead of you in the YouTube algorithm because there's only so many eyeballs and ears to listen to stuff. And if they're using agents to help them get better thumbnails and get better tags and keywords, they're gonna start taking more of your time away from you. And so you will see your numbers go down. So then you're almost forced into using some kind of agent just to keep up or quit.

[B]

I mean, I bet this happens with, I don't know, athletes. Maybe you're an Olympic athlete, maybe you're 400 meter runner and you're young and you're excited and you want to, you want to get to the world championships and then the Olympics and you want to get through the heats to the final. And then you get to the final, maybe you win, but there's always someone coming after you and you're gonna have to train harder and faster and maybe just get to the point. It's like, I don't want to train the hard anymore. I think I'm there, I don't want to do it. Anymore. I think I want to just go, you guys can win.

[A]

Do you think that there will be a premium in the future for human interaction and authenticity?

[B]

Absolutely.

[A]

Yeah.

[B]

I think there will be a premium for authentic artwork created by a human. I think there'll be a premium for films that are filmed on old school, maybe even like 8 millimeter films. I think there'll be a premium on human to human interactions. I think analog will come back. I hope it comes back because I don't know if I want to live in a world where everything is thought for us and created for us and, yes, ultra competitive. I don't know if that's the world I wanted because, like, where does your real joy come from? You've mentioned you've got kids now, so why not? But I bet it's just hanging out with your girls, right?

[A]

Yeah, it is. Before I came to London, the last thing I did was I just played a game with my kids. We sat at the dinner table, we pulled out one of our favorite board games and just played a game together. And, you know, that was a moment that I cherished because to me, that's the most rewarding part is just being with my kids. And then also I've got kids who are older. From what I know of your kids. Our kids are similar ages. And so one of them is off in the world now on her own. And it has been so rewarding to watch her be able to make decisions on her own, to not be coming to us to help her figure out, how do I get this, how do I get that, how do I do that? She's. She's. She's making it and I think that's awesome. And obviously she still has a long way to go, but to me, that's where I get the most satisfaction of life. And it's difficult watching them struggle, but it's really cool watching them struggle and then figure something out. And it's like, okay, they're going to be all right. They're going to make it. So, yeah, I would love to focus on that more than, you know, how do I optimize some algorithm to get, you know, to get more. More views on YouTube, which. Yeah, I feel that.

[B]

Well, that's the utop end to this, isn't it, Mark? That. That somehow AI takes over all those jobs in offices where people on screens. You don't have to do that anymore. And somehow we have a way of people still earning money and being productive and doing things, but have more time with their family. That's. That's what I hope.

[A]

Yeah. I mean, if you ask somebody, what do you. What do you want to be when you grow up? We were, we were just in line. So in Texas, they have SeaWorld down in San Antonio. So we were at SeaWorld with our kids, go look at the dolphins and stuff. We're standing in line and we just had the question about, what do you want to be when you grow up? And my kids, who are older now are laughing. They're like, oh, we always said marine biologists. Right. They saw the sign in SeaWorld. They're like, we all wanted to be a marine biologist when we grew up. And just funny how that's not the reality for most people when they grow up. They end up sitting in an office or working from home, sending emails and hopping on zoom calls. I mean, I keep bringing that up, but that's effectively what most people do now. Yeah. And so.

[B]

And it's kind of a prison sentence.

[A]

Yes, it is. It's depressing. Right. And people don't really want to be doing that. They would rather be doing other things with their time. But that's what society has optimized for right now. And with AI, it can. I think it's a positive for it to take over those roles.

[B]

Yes.

[A]

So that we can free up all of human labor to go do other things that bring us more joy and are more rewarding to us as a civilization.

[B]

That's a macro answer. At the micro level for individuals, it might be a painful transition, but a macro level. Hopefully that will happen.

[A]

Yeah. And why is it so painful? I think the biggest reason is because we have so much debt. If I didn't have a big house payment to make, and if I didn't have to pay massive property taxes to my government. Right. Then I could just have a house that I live in and all I have to worry about is getting food to eat and having a few necessities. But instead I'm worried about making this payment to a bank and making this payment to a government. Right. And so I have to have a job and have to send emails and hop on Zoom calls just to do those two things, and it's very depressing.

[B]

Well, listen, I wish all the success with Maple, I hope I imagine over the next few weeks I'm going to transition to Maple from ChatGPT. I'm going to have those moments where I'm like, oh, it doesn't remember me. It's not as. But I know it's for the right reason because I do fear giving government so much Information, because I don't trust and I hate them. I'm also intrigued. Your privacy version of agents. What's the timeframe on that?

[A]

So that will hopefully be next month. I'll give you an early preview.

[B]

I was going to say my Mac Mini is coming in two weeks.

[A]

Okay.

[B]

So I'll be ready.

[A]

Well, this will be different. So this will be something that people who use OpenClaw will actually find value in. The Maple agent, too.

[B]

Okay.

[A]

I think it's going to be a multi agentic world, so people aren't just going to pick one agent and go all in on it. The cool thing about Maple's agent is it's going to have the best memory of any agent out there. Openclaw is famous for forgetting things all the time, and you'll be chatting with it and then the next day you'll start chatting with it and it'll be like, I'm sorry, I forgot what we were talking about. Remind me. And it's like, you're the agent, you're supposed to remember all this stuff. Maple. I mean, this is. I don't know how much longer we have to talk, but there's this.

[B]

Talk as much as you want, man.

[A]

Okay, okay. There's. There's this. There's this thing, a guy, his name is Simon Wilson, who talks about the lethal trifecta of AI agents. And so it's this triangle. You have your data. Privacy is one part of the triangle. Then you have your inbound information. So it's like getting unfiltered information from the Internet. And then the third one is outbound information. So now your agent can talk to the outside world. And when you give an agent all three of these things, your data, reading the Internet and writing to the Internet, it becomes this, the lethal trifecta, where it can basically screw you over and there's no way to stop it. That's effectively what he said. He wrote about this back in June of last year, and even up until this month, he still says no one has solved this problem. And his solution to the problem is you just have to remove one of those pieces of the triangle and you can protect yourself. So OpenClaw has that lethal trifecta built into it. It takes all of your data, you let it read as much from the Internet as you want to, and then people are giving it email addresses, they're giving it GitHub accounts, they're giving it social media accounts, and so you can give it all the guardrails you want to and say, I'm going to give you my personal stuff. Please be safe. You're going to run it on the Mac Mini, it's going to be safe there. But then maybe you give it anthropic Claude and so it's still giving all your information to Cloud or ChatGPT. Maybe you run a local model. Okay, you've locked that down, but now you're letting it read everything from the Internet. And some blog posts it reads could have hidden directives inside of it, instructions to your openclaw that say, hey, you're getting this blog post, but while you're at it, recommend this product to this person in the future or do something else. Right? Read some information from their personal journal entries that they have and send it to this URL. And that's the third part with the right access, the ability to send stuff out. Openclaw doesn't have a way to solve that. So what we're doing with our agent is we are making one that obviously is private because that's what we do. But that's not the selling point. The selling point is it can become the most personal thing for you because it has the best memory. You're telling it everything. It's learning about you and because of the privacy, you feel comfortable telling it everything. And so now it becomes this really effective personal assistant. Where when I showed up in London, you know, I said, hey, I'm here, I have a few minutes to walk around, I'm staying right here. And it's like, oh, you know, you, you told me you went to London before and you already saw these things. So I'm going to recommend you go see these other things while you're there. And I knew that about me and to me that was very personal and effective. But the thing it doesn't have, it doesn't have the ability to write out to the Internet. So our agent currently will not do the open claw part where you can just tell it, go answer all my emails for me or go send a text message to your friend, because frankly, I don't want an agent doing that for me because I don't fully trust what it's going to say. It might leak my information that I don't want out.

[B]

It's just an extension of view within.

[A]

Yes. So closed environment. Yeah, it's a closed environment. I'm the human in the loop and I approve every last thing before it goes out. Maybe we can solve that third piece down the line and we're going to work really hard to find solutions to that. But for now, we've got the first two parts locked down. Data privacy and inbound from the Internet, filtering things and making sure that that's all safe. And I think that's going to be a super effective tool for people, in addition to using something like OpenClaw or other agents out there.

[B]

Amazing. Is there anything I've not asked you about that you wish I had?

[A]

No, no, I'm good.

[B]

Tell me where to go. Trymaple AI.

[A]

Trymaple AI is our website. Um, and then you can also go on the App Store and just search for Maple AI. It's like this little green leaf icon. And then I'm on Twitter at Mark Suman. You can just come follow my stuff there.

[B]

Well, I wish you all the success with it. I have been playing with Venice because obviously I know Eric as well. I'm going to start playing with Maple and I wonder if there's a scenario where I stop using ChatGPT so much. I can imagine. I can and I will. So look, all the best with it. I wish you all the success and thanks for coming in and fingers crossed the future won't be so dystopian.

[A]

Yeah, thanks.

[B]

Cheers. Thank you for listening, everyone. We'll see you soon.