How Do We Constrain Trillions of AI Agents?

Cern Basher: The technology and the impact is inevitable. The question is now how are we going to react to this as societies. And here’s something to think about. We’re talking about the benefits of all these AI agents. The internet effectively is going to have billions if not trillions of these things running around. So the internet, instead of being 99% human, is going to be like 1% human activity and the rest is going to be AI.

I’m not worried about your agents and you’re not worried about mine, but I’m worried about agents maybe from China, Russia, North Korea, other nations.

Lisa Tamati: Yeah, me too.

Cern: They’re going to try to infiltrate everything. They’re going to find a weakness in everything and try to capitalise on that. So the question is, how do we constrain AI?

Lisa: Big question.

Cern: In the real world, how do we constrain people? Police, military. We have police, we have walls, we have barriers. We have military forces — army, navy, air force. That’s how we constrain. We physically constrain people. We also make it so that if they attack us, they know they’re going to get attacked back. There’s going to be a response and they know they may pay with their life.

So how do you take that concept and apply it to cyberspace? Well, the good news is that the proof-of-work protocol in Bitcoin is exactly that. We can actually apply Bitcoin as a wrapper around everything. It becomes a digital wall around everything in cyberspace.

Bitcoin as a Digital Wall — The Spam Email Analogy

Cern: The way to think about this is — if I send you an email right now, it costs me basically nothing. You don’t charge me to receive the email. But if you wanted to protect your email system from AI agents sending you a million emails tomorrow, the effective way to do it would be to charge each agent to send an email to you. Wouldn’t have to be much. Once they couldn’t send out a billion emails anymore because it would become pretty expensive.

You could apply this essentially with every sort of interaction on the internet — with computer systems accessing different computer systems. The spam example is one that people intuitively understand.

There are some people out there warning that in very short order, our email systems and our phones will be inundated with emails and unwanted calls and we won’t be able to stop it. Your phone would just constantly ring and ring and ring and there’s nothing you can do to stop it. Because AI will figure out a way around all the different security features that we’ve hard-coded in logic. The only way to beat that is to not hard-code it with code and logic — to actually use a physical way to stop it. And that’s exactly what Bitcoin is.

So if I want to send you an email, I have to pay half a cent. Which for us would be no big deal. But if you’re sending out a billion spam emails, that becomes a massive inhibitor.

Why Decentralisation Matters — The Google Problem

Cern: Ultimately I think the best method for this — and this is kind of why I see Bitcoin and AI as going hand in hand — is that you want a decentralised system. You don’t really want it in the hands of any one party, because that party could be compromised or they could use it for nefarious intent.

We’ve all seen Google, for example. People trust Google to various degrees, and Google has done things in the past that maybe haven’t been so user-friendly. The whole idea is to take that kind of authority away from a central power — whether it be Google or Microsoft or any government — and put it in the hands of a decentralised system where you have to do the work and pay the energy and the compute to prove that you are legitimate.

This is one way to constrain AI, and I think this is a problem that’s only going to increase very quickly. Within weeks. OpenClaw has really opened up a Pandora’s box. There are already billions of these agents out there, good, bad, and otherwise.

Centralised Censorship of Health Companies

Lisa: My company is being penalised by Google because I’m a supplement company. I’m in the health optimisation space. This is my conspiracy theory, so take it for what it is, but Google — the big stakeholders in pharmaceutical — they penalise supplement company brands from being able to advertise on their platform.

The fact that I have a book that talks about aneurysms, my mum’s story, cancer — not a show in hell of advertising on the Google Ads network. That penalises my company massively just from having my book, let alone the blogs, let alone the podcasts. I’ve tried over and over again to get approval for just my supplements, but because of the content — if you have the word cancer, if you have the word lipospheric — I sell lipospheric vitamin C, which everybody knows you can get at your local chemist. You cannot use the word lipospheric. It’s a medical term. You get banned.

That’s the sort of example where centralised authority is acting in their own interests, not in the interests of the public.

Cern: This is where AI also comes in and becomes very useful. AI can decide on the merits of each case. AI can read every single blog and everything you’ve put out and evaluate whether this person is trustworthy. Whereas right now you’re just getting scooped up by an algorithm because of a certain word. They block you, they don’t tell you which word, and there goes your income for the month.

Lisa: This is really hard on companies in the natural health space. These are systemic big problems. AI helped me work through the whole thing — I dismantled my whole website trying to get approval. Even AI gave up. He said, “Look, you’re not going to get through. Give up.” Which is pretty sad.

Jason Lowery’s Softwar Thesis — Bitcoin as Gunpowder

Lisa: Jason Lowery wrote the Softwar book. He’s a major in the military who was told to go and study Bitcoin as a defence weapon. Tell us about his thesis and how it applies with AI.

Cern: This could take up an entire podcast, but the general structure is — the story he tells is an interesting one. The Chinese developed this magical black powder. It had some flash properties and it was developed in the Chinese medical community. For years it sat in a glass on a shelf and they experimented with it and didn’t have much use for it in surgery or other things.

But eventually along comes some Chinese engineers and they discovered what the medical community had invented was gunpowder. They thought it was medicine because medical people discovered it. But as we know now, gunpowder is used in fireworks and weapons.

Bitcoin suffers from the same problem because Bitcoin has the word “coin” in its name. Everybody thinks of Bitcoin as money. But like gunpowder, what Bitcoin really is at its core is essentially a digital security technology.

You can’t mine Bitcoin, you can’t get Bitcoin unless you do the work — proof of work. You actually put the computer system and the power in place and engage in a global race every 10 minutes to guess a massive number 256 digits long. Unless you have computer systems guessing and getting the right one every 10 minutes, you don’t get any Bitcoin.

This is a way to decentralise control. In this case it’s a ledger — who gets to put the next transaction on the ledger. Ledgers can be used for all kinds of things. Ledgers obviously are great for money, but Bitcoin only makes for great money because it is so secure, because it’s decentralised, because no one controls it.

But that’s not what Bitcoin truly is at its core. It’s a cyber security system. It’s a protocol, just like we have other protocols on the internet for email and for websites — HTTPS. This is a security protocol that’s just sitting there waiting for the world to adopt it as a security protocol. And everybody has run off saying it’s money. Well, maybe. But you’re trying to use gunpowder for something else.

In the test of time, maybe Bitcoin does become our form of money or at least the base layer of money. I don’t know. But I’m pretty confident that we can use it as a security technology. And we need it very badly in this age of AI agents. We need a way to constrain AI physically because we’re not going to do it with computer code. AI is going to find a way through all of our computer code and hack it. You can’t hack Bitcoin.

Why AI Agents Choose Bitcoin

Cern: People are worried about quantum cracking Bitcoin, which to me is 10 to 15 years off. Let’s talk about it, but it’s not an imminent threat. What’s more important is that the AI we have now is going to crack our banking system, our medical records, our emails in the next few weeks. Not 10 years away. That’s a much bigger problem right now. And the only thing that can protect us is Bitcoin.

The US could put in a system controlled by one company or consortium of companies. But other countries won’t be happy with that. Why would you let Microsoft control all of your data?

People have asked AI what it would prefer, and every time AI has said Bitcoin makes sense. Agents have come up unanimously with Bitcoin as the thing they want to use and transact with.

Lisa: Which is funny.

Cern: Bitcoin is never closed. 24/7, global, decentralised — all the things you would design in a global marketplace are encapsulated in the Bitcoin network. This is a massive use case for Bitcoin. Bigger than Bitcoin as money. It’s more valuable than that, and nobody’s talking about it.

SaaS Is Cooked — Industries Being Deleted Overnight

Lisa: If we have all this technology, it’s a massive deflationary force. I’ve had Jeff Booth on the show talking about this. In the last few weeks, AI agents coming on — OpenClaw has been the catalyst for billions of dollars to be wiped out of the stock market overnight. Whole industry sectors — consulting, legal, software as a service — have lost value overnight because people are realising we don’t need them anymore.

I’m a big fan of Jordi Visser’s work, a macro analyst. He’s saying people are going “Oh, SaaS is going to bounce. It’s time to get in. It’s hit the floor.” And he’s going, “SaaS is cooked. It’s gone.” Unless you’re something like Palantir that’s slightly different, most of these SaaS companies are absolutely toast. Unless you work all day in AI like I do and see the day-by-day changes, you won’t understand that.

Cern: It’s not so much whether one company or industry goes under. Salesforce may be in place for years and years, but it’s where is the new value created? That’s where the stock market will flow to. These companies may not go bankrupt for 20 years, but the stock market is not going to place a great value on them anymore because that’s not where the growth opportunity is.

Deflation Isn’t Bad — We Already Lived Through It

Cern: We actually in our lifetimes have all lived through massive deflation. We saw the iPhone and other smartphones come into being and get rid of the pile of stuff on the left side of that chart. That was massive deflationary. How bad was that, Lisa? Was that a terrible experience?

Lisa: No, it was great. We could get cheaper things. The iPhone became your one thing and for us as consumers it was great.

Cern: I think there’s this narrative out there that “oh, we need a little bit of inflation, we can’t have deflation, people will stop spending.” I think it’s BS. I’m calling BS on that. We’ve all lived through this. We’ve all benefited from this. If we do this at scale with every other industry, we’re all collectively going to be far better off.

AI, humanoid robots, autonomous vehicles — it’s going to dematerialise a massive swathe of our economy, but it’s going to make us so much better off. There will be victims, like the companies that are those things. And there always have been victims throughout history. Think about the looms that put all the weavers out of business. The transition from farms to factories. All the women that used to be switchboard operators in the United States — I think there were half a million of them. They were all put out of work.

This is the constant state of things. Societies have been dealing with technology changes throughout millennia. Just not this fast. But we’re up to the task. We handled the smartphone one perfectly fine. I’m not upset that I don’t get to use a fax machine anymore.

The Ice Cutter Story — Disruption Throughout History

Cern: I did a post about this recently. In the United States, we used to have people that would cut ice. Before refrigeration, the guy that decided ice was good figured out a way to ship ice from Boston down to the Caribbean. It was a two or three week journey. Half the ice had melted by the time he got there.

He had this idea of going to bartenders saying, “I’ll give you ice for free. I want you to do a test. Give people a drink with no ice and a drink with ice.” Very quickly he created demand. He created this massive monopoly, eventually shipping ice around the world to Australia and India.

But then he was disrupted by refrigeration. And initially he said, “Refrigeration? That’s fake ice. That’s not God-created ice. This is unnatural ice.” That narrative went around for a long time. Then people started getting sick from the natural ice from lakes, and eventually the idea of “fake ice” became clean, healthy ice. And this guy was put out of business.

It’s just a reminder. Refrigeration has completely changed the world. Air conditioning. Think of the places people live now because of that. With AI, we’re creating these opportunities in every spot all at once. But this is something humanity has gone through time and time again.

The K-Shaped Economy — Will Abundance Lift Everyone Up?

Lisa: At the moment we’ve got this K-shaped economy. The asset owners doing really well, S&P at all-time highs. And you’ve got the worker whose wages are debasing and inflation has eaten away at all their savings. There’s massive unrest. Is this going to be positive for those in the bottom 50%, or is it only going to benefit the rich?

Cern: If we have the ability to double, triple, quadruple, 10x GDPs, there should be more than enough for everybody. There should be enough for investors to make a ton of money, and enough for the people being left behind. Not just a subsistence living, but a great level with everything they need.

In theory, there should be enough. In reality, are we going to design political systems and taxation systems that work for everybody? That remains to be seen. This is the biggest challenge of all. The technology and the impact is inevitable. The question is how we react as societies.

How is New Zealand going to react to the fact that AI and robotic labour is coming to your shores? Do you put up walls and try to keep it out? That doesn’t seem like a good strategy long term. The question is how does a country figure out how to embrace the technology while making sure people most impacted can still live a comfortable life.

Lisa: If we don’t, that’s when the wars happen and the revolutions and the riots and the breakdown of the fabric of society. So it’s in the interest of the rich to make sure some of this gets down to the bottom half.

Universal High Income — Making Goods Freely Available Like Water

Lisa: What’s going to happen from a political point of view and control over our lives, if the government is giving us our money and telling us what we can and cannot do? Is that going to erode our freedoms?

Cern: The easiest way to do it is just to make it freely available. Kind of like water today. Water is basically too cheap to meter. There’s public water fountains everywhere. You can get as much water as you want for basically nothing. Air and sunshine — free. Those are public goods.

Now imagine a world of abundance where more and more things are considered to be a public good. Just food in general. Instead of giving people money to go to the grocery store and the government saying “Lisa, you’re not eating healthy enough, we’re not going to give you as much this month” — which is horrible — you basically just take the grocery store and say, “Take what you want.”

Initially there might be hoarding, but eventually people will settle down and take what they need. That’s the long-term way I foresee it happening — making more and more things freely available. Without the control of “we’re giving you the money, you’re going to jump through hoops to get it, and if you don’t follow the rulebook you’re cut out of society.”

Lisa: We saw the dangers of that during COVID.

Cern: There’s a progression. It doesn’t happen overnight. You don’t just say everything in the grocery store is free. All the shelves would be bare instantly. But over time, as the robotic workforce is making an unlimited supply of everything, you can imagine going to a furniture store and saying “These are all made by the humanoid robots. Take what you want.”

And by the way, these would be beautifully, expertly crafted couches. Not cheap stuff. High-quality, long-lasting, locally produced. You may even be able to ask for customisation. The robot does it for you.

The notion of the importance of money becomes less and less when you have abundance. Right now we have an abundance of water and you can get it basically for free. That’s the mental model for increasingly lots of other things becoming abundant.

Bitcoin Price — Why Cern Isn’t Worried

Lisa: Bitcoin not doing so well lately. We’re in a crypto winter really, which surprised a lot of us. Short term it’s not looking great and we’re not following gold and we’re not following the normal correlations. Is this the end of Bitcoin or the start of Bitcoin?

Cern: I’ll echo some sentiments I’ve heard Michael Saylor say recently. It’s not been that many days since we had a high in Bitcoin — I think it was last October. For us to complain that Bitcoin’s not doing so well is interesting. It’s only been a short period of time.

Of course we’re looking relative to gold. No one said it was going to be pegged to gold. I think there are factors driving the gold market separately. None of this worries me.

I’m comforted by the fact that we need a security technology to constrain AI and here it is, waiting to be discovered. Hiding in plain sight. For me, this speaks to opportunity rather than risk. I’m not worried about Bitcoin going to zero. It’s a remarkable technology that people don’t fully appreciate yet. It’s great as money and a store of value, but it’s even better as a security technology. Eventually people are going to come around to this notion.

Bitcoin as Security Tech Is MORE Valuable Than Bitcoin as Money

Lisa: If we’re heading into a deflationary world, do we need Bitcoin from a monetary standpoint?

Cern: Two parts to that. First, as long as we have fiat currency of any kind, it’s not going to be deflating. We’re always going to be printing more money. If governments start giving money to people, they’re going to be printing even more. So from a monetary standpoint, you want to be in the currency that is not printing itself into oblivion.

But separating Bitcoin into the two elements — let’s say you were worried about the monetary side and said we don’t need Bitcoin anymore as money. Fine. I’ll give you that. But we need Bitcoin as a security technology. And that, in my view, is way more valuable than Bitcoin as money.

Lisa: Wow.

Cern: We’re going to have trillions of AI agents roaming around the internet. How are we going to stop them? We’re not going to do it with more code and passwords. It’s going to hack all that. We have to do it with a way that cannot be hacked, with some physical energy involved so that agents prove they’ve done the work.

To me that’s so obvious at this point. I’m just waiting for the rest of the world to come around to this idea. Again, this is not my idea. This is Jason Lowery’s. But it’s one that’s really resonated with me and it grounds me in this time when Bitcoin has been weak. It’s just a matter of time. A good accumulation time.

Michael Saylor recently said that when electricity was invented, it took 30 years for people to really come around to what it could be used for. Same thing with Apple and the iPhone — seven years after the iPhone came out, Apple was trading at something like 10 times earnings, a ridiculously low valuation. It just takes time for people to figure out what new technologies can be used for.

Strategy as the World’s Largest Digital Security Company

Lisa: MicroStrategy — or Strategy as it’s now called — has been hammered more than Bitcoin, down 70-odd percent from their highs. People are calling for it to go under.

Cern: I’m an investor in Strategy. I’m not particularly worried. I’m worried from the fact that maybe their premiums might not get back to where they were. But the company itself is safe. Bitcoin would have to go to horrifically low levels and stay there for 5 years for them to be in financial trouble. Very unlikely.

The interesting thing is — if you think of Bitcoin as a digital security technology, then Strategy becomes the world’s largest digital security company.

Lisa: Wow. Never thought of that.

Cern: This is a whole different value proposition. If Bitcoin takes hold as a security technology, we might see premiums on Strategy that we’ve never seen before. It’s not investment advice, but it’s interesting if you look at things through this different lens. Bitcoin as a security technology changes what you think about all kinds of different things.

Tesla Is Massively Undervalued — The Lemonade Stand to Surgeon Analogy

Lisa: Tesla — people value it on their price-to-earnings ratio and quarterly earnings. You had a brilliant analogy about a kid. Give us that take.

Cern: Let’s say as a child you created a very successful lemonade stand. You did it through your teenage years. It’s made more and more money every year because your lemonade’s amazing. Everybody from around the city comes to your stand.

But off you go to college. You might still have the lemonade stand going part-time, so you’ve still got revenue. You’re known as the lemonade stand guy. But you’re going to college studying to become a surgeon.

It’s like valuing that kid through college as though he’s just going to be a lemonade seller for his whole life. If you valued him for his future revenue as a surgeon, the PE would be much higher. But the market is looking at that kid because that’s all he’s ever done — sell lemonade.

This is the kind of transition that Tesla’s going through — from an automobile company (lemonade) into an AI and robotics company (the surgeon). The market is not fully appreciating the economics of the AI and robotics business.

When I do a present value calculation based on what I think are very conservative assumptions over the long term, I get a valuation today that’s way higher than where the stock price is. Again, not investment advice. But the market is only allowing itself to look out a little bit.

I don’t think the valuation for the humanoid robot business — Optimus — is really part of the company’s valuation currently. There are no analysts on Wall Street that I’ve seen with a serious model for that. They’re only just now incorporating the robotaxi business. So there’s an interesting opportunity as this company remakes itself.

It’s not just replacing one profit stream with another. It’s a recurring profit stream on the AI business, whereas the car business is a one-time sale. The margins on the recurring profit are that much higher. You’ve kind of 50x’d dollar for dollar the profit stream. It’s a material change in the business model.

Will SpaceX and Tesla Merge?

Lisa: xAI and SpaceX merged — the biggest merger in history. Do you think SpaceX and Tesla are going to merge?

Cern: What I’m hearing through the grapevine is this is something Elon wants to do. I’ve kind of gone on record saying I’d prefer that they don’t. I’d prefer they not distract themselves convincing Tesla shareholders. As an investment person, I’d rather allocate to two different companies.

I also think from a political risk standpoint, having two separate companies insulates Elon more. If the government goes after Tesla, he’s still got SpaceX, which is important for defence contracts. The government needs to be careful with Elon because they can’t destroy Tesla and expect him to continue to defend the country with his rocket company.

But if that’s what Elon wants, I think that’s what will likely happen. I have concerns, but I can see the economic benefits. From a monopoly standpoint, that much concentrated power in one company gives political reasons to go after him.

SpaceX: Starlink, AI Data Centres in Space & Starship Economics

Cern: SpaceX certainly has Starlink, which is amazing — adding a million subscribers every month it seems. The AI data centre opportunity for SpaceX is enormous. As Starship becomes operational and launch costs come down, it’s going to cause an explosion of things we put in space.

It won’t just be Starlinks. It’ll be AI data centres and all kinds of other stuff. We’ll start making pharmaceuticals in space. If you drive down launch costs by a factor of 10, you create a massive opportunity.

The way I think about space is it’s like we’re discovering a new continent. We now have the ships to get across the ocean. And space is the ultimate one because of the enormity of it. If you’re discovering North America or Australia, you’re constrained by size. With space, we have no such constraint.

With data centres up there, you’ve got solar power on tap 24/7, automatic cooling. You don’t have the problems with environmental impact complaints. AI data centres in the United States already consume about 7% of total electricity generation. We’re going to have an imperative to put it in space and free up that capacity.

Over the long term, we’ll probably put most of our dirty manufacturing in space and leave the earth cleaner and more pristine. The energy costs in space are going to be so much less. If you have a very energy-intensive operation, ultimately I think it’ll be cheaper to do it in space.

Lisa: The pharmaceuticals — is that because of the low gravity situation?

Cern: I don’t have a full appreciation for that, but it’s my understanding that a lot of drugs would be better produced in that kind of environment.

Lisa: I interviewed Dr David Furman from the Buck Institute at Stanford. He works with astronauts trying to protect them from aging in outer space — you age something like 10 times faster up there. They do lots of studies in low gravity with organoids suspended in zero-gravity environments, testing interventions on mini brains and mini livers. When you said that, I wondered if that low gravity environment has particular properties we can’t simulate on a bigger scale here on Earth.

Owning Your Own Robotaxi Fleet & NZ’s Opportunity

Lisa: We’ve covered everything from space to Bitcoin and AI. Thank you for your brilliance today. And on brilliance — you have Brilliant Advice. If anybody in America wants financial advice, check out brilliantadvice.net.

Cern: I wish I could work with people from New Zealand and Australia. Unfortunately, the regulations preclude us unless the funds are in a US institution. But X is really the best place to follow me — just @CernBasher.

I would mention that my last video with Herbert — we looked at the economics of an individual owning a robotaxi or a Cyber Cab. A lot of people worried about losing their job to AI are asking how they can replace their income. This could be one interesting way — own yourself a Cyber Cab or two or three or a small fleet.

Lisa: Do you think that’s going to be available worldwide?

Cern: I hope so. I believe it’s Elon’s intent to make it global. The interesting thing about the Cyber Cab is it’s the first automobile that rolls off the production line ready for both left-hand drive and right-hand drive roads — because it has no steering wheel.

Lisa: That’s an opportunity for other countries to say let’s bring this technology in. We don’t have the same regulatory limitations. Operate in our country. Be the global leader.

This is a moment where some forward-thinking leaders and governments have an opportunity to leapfrog every other nation. We’re a small country that could do that. Let’s hope some powers that be are listening to this interview. Let’s be tech-forward. Hiding under the covers isn’t going to fix us.

In my business it was like “I don’t want AI” and then it was “I’m going full AI because I have to.” And I’m absolutely gobsmacked at my abilities with AI compared to prior. Just insane.

Cern: Some of these things you’re using have just been developed. They’re the worst they’re ever going to be right now.

Lisa: That’s why I’m doing it and learning it and tripping over. Wrestling with a computer every day because I have to be in on the ground floor.

Cern: Fantastic.

Lisa: Thank you for your time today, my friend. Absolutely epic.

Cern: My pleasure, Lisa. Thank you.