Pilot interviews

Harold Tan: 'Scarcity is paramount'

Harold Tan, a professor and former e-commerce and solar founder, on scarcity, AI bottlenecks, Bitcoin, and keeping emotion out of investing.

Autopilot9 min read

Harold Tan was born in London, grew up in Australia and came to the United States after high school. Harold studied cybernetics and mathematics at UCLA and information systems at Harvard, then built e-commerce companies before going into solar energy. Energy, in Harold's view, is the foundational layer of all society and civilization.

Open it in the Autopilot app. https://link.joinautopilot.com/eXKr?c=harold-tan-interview&af_sub1=intro

Harold is also a part-time professor at Santiago Canyon College and teaches at UC Irvine, including a course on bitcoin and digital assets. On Autopilot, Harold publishes two Portfolios: the AI Bottleneck Portfolio and Uncle Sam's Checkbook.

This interview comes from the transcript of our conversation with Harold Tan in May 2026. It has been edited and condensed.

What's your name, and what do you do for a living?

My name is Harold. I was born in London, grew up in Australia, and came out to the States after high school to seek the American Dream. Went to UCLA for my undergrad. I studied cybernetics and mathematics. Cybernetics is a branch of artificial intelligence. I went on to study information systems at Harvard. I'm a full stack programmer.

Straight out of college, I started my first e-commerce company. I've built and exited two e-commerce companies and then went into building a solar energy company. Energy is something that's very fascinating to me because it's, to me, the foundational layer of all society and civilization. Without energy, we don't have anything.

So I'm a business owner and operator. I understand cash flow and profits and balance sheet and so forth, market challenges. And I'm also a part time professor at Santiago Canyon College, and I teach at UC Irvine in Orange County. I teach a course on bitcoin and digital assets. I also give lectures on Bitcoin to graduate engineering students, and I teach business concepts to graduate students at UCLA.

How did you get into stock picking?

I've always realized that the main path to wealth is through investing. If anybody is just working a job and earning a salary, they're always going to be capped. Even if you're the CEO of a company, maybe it's a high salary, but you're going to be capped. So investing is necessary to really build wealth because there's no cap.

I started having AI analyze my own investment portfolios to find where I could optimize it, or organize it in ways that better fit my worldview. And so after using AI to optimize my own portfolio, I started sharing it with friends, which made sense to them. I kept updating this Google doc for friends and family to follow along. And then I realized Autopilot could basically do that for me.

Did you see the energy bottleneck in AI coming this quickly?

I did see the energy bottleneck happening very quickly because I come from the Bitcoin mining industry and Bitcoin is intrinsically tied into energy. Essentially, the Bitcoin miners are making the profit on the delta between the cost of energy and the price of bitcoin. And so we were always looking for low cost or free or wasted energy areas where they were producing too much energy and needed to offload it. And I saw that part of the challenge with that was interconnecting the grid and the equipment and the data centers to those energy sources, and a lot of Bitcoin miners are now actually supplementing, even pivoting their business models to AI because of the massive demand.

What does a bottleneck mean, and how does that apply to AI?

The way I see a bottleneck is that it's something scarce. Scarcity is paramount. And I started first truly understanding scarcity through Bitcoin. When I came to study money in general, I realized that even if the number in your portfolio was going up in nominal terms, it may not be going up in real purchasing power.

So in reference to AI, I'm a full stack engineer programmer, I developed back and front end software, and I saw that software is basically going to be free and abundant. And the rate limiter, as it were, is the substrates that feed AI. AI is turning electrons into compute. Or another way to think about it is AI is turning electrons or energy into intelligence, and intelligence is going to be abundant. But what's going to bottleneck that growth is the energy, compute, physical hardware and grid connections to allow the growth of that intelligence.

What are some examples of those bottlenecks?

One example of a bottleneck is energy. You can essentially create software overnight now with AI. However, you cannot bring a power plant online overnight. Dealing in the energy space myself, I've seen the long lead times, the bureaucracy that's needed to permit and bring energy online, so it can take months, if not years, to bring energy online.

Another bottleneck in the AI space is photonics. For many years, there were labs that had developed fiber optic systems and very specialized lasers that operated on nanometer scale. But there wasn't a widespread demand or commercial usage for it because the standard in computing has been copper. However, with the explosion of AI, there's a sudden demand and not enough developers for the lasers and photonics that can bundle and package with the GPUs.

Why does the AI Bottleneck Portfolio include exposure to Bitcoin?

My thesis is that in a world of digital abundance, there will be a need for digital scarcity. The one other key thing with Bitcoin is in a world where we do not know what's real, what's not, with deepfakes and AI can create anything, Bitcoin is something that anchors truth, because it requires something called proof of work, and timestamps every transaction into a block. It develops a method of verifiable truth in a world where anything might be fake.

Where do you get your information?

I try to gather my sources from a wide variety of places. I've been following a Stanford engineering class online that is actually the economics of AI. And in that class, they have guest speakers that discussed the entire buildout of AI, people who are actually operating and running these companies. And that's been a valuable source of information. It's amazing how much free and available information there is out there. The key is finding the signal in the noise. So being able to listen to an actual business owner and operator speak, to me, is far more valuable.

What do you think your edge is?

I think my edge is being a long term systems thinker. And being a business operator, I understand complex systems from a macro, holistic point of view, and I'm not emotionally driven by the market. A lot of people I see get shaken out of the market with volatility. For me, volatility is vitality. And everybody likes volatility to the upside. They just don't like volatility to the downside.

Why do you take a long-term view?

I'm a long term systems thinker. I've been involved in the Bitcoin space for over ten years, and anybody who takes a short term view in the Bitcoin space usually gets shaken out. And so I think it's important to take at least a five year view in the space, because in the interim there will be volatility. People who are traders will want to collect profits on quarterly earnings or trade momentum and so forth. And so the important thing is to not get emotionally shaken out, but to allow the thesis to play out and only change course of action if the thesis breaks.

So in my portfolio, I have very specific rule sets for thesis breaks. If Nvidia no longer has a moat, if nuclear energy gets completely outlawed or we have another massive blow up and there's negative sentiment, then I may pivot from that. Or if we discover new methods of cooling, or if we enter data centers in space much faster than anticipated, then that warrants a reallocation or a thesis shift. But if price goes up or price goes down on a quarterly basis, a lot of that is just driven by emotions. And I don't want to invest based on emotions.

How do you think about geographic risk?

We've gone from a unipolar, America dominated world to a multipolar world where the supply chain is somewhat fragile, especially with geopolitical tensions. And so one of the rules I have in my portfolio is to not concentrate too much geographic risk. And I've got portfolio weightings based around that.

Why do you think friends and family look to you for stock picks?

What I noticed is most people don't have the time or interest to do deep dives into highly technical or financial matters, and I understand life is busy. People need to do the groceries and pay the bills and walk the dog and pay attention to the crying baby. For me, I am fortunate to have a passion and the time to deep dive into a lot of these topics, and also the background experience to understand the first principles of how computing and artificial intelligence works. And so I'm happy to spend the time doing what I would do anyway to share with friends, family and others out there.

“"If price goes up or price goes down on a quarterly basis, a lot of that is just driven by emotions. And I don't want to invest based on emotions."”

Follow Harold Tan's Portfolios on Autopilot

Harold Tan publishes AI Bottleneck Portfolio and Uncle Sam's Checkbook on Autopilot, an investing app that lets you follow Portfolios from investors, AI models, and public filings in your own brokerage account. See Harold Tan's Pilot page. https://link.joinautopilot.com/eXKr?c=harold-tan-interview&af_sub1=body

  • You follow in the brokerage account you already have, and the money stays there.
  • On Premium Tier, when a Portfolio changes, Autopilot Advisers sends the orders to your broker. On Basic Tier you get an alert.
  • Review the Portfolio's method, holdings and any available performance record before following it. New Portfolios show return figures after 30 days.
  • The Pilot sets the subscription price, there is a 0.00% AUM fee, and you allocate at least $500.

You are never trading at the same moment as the Pilot, and your account won't match the Portfolio exactly.

Open it in the Autopilot app. https://link.joinautopilot.com/eXKr?c=harold-tan-interview&af_sub1=end

"Autopilot" refers to Autopilot Holdings Corp and Autopilot Advisers, LLC, an SEC-registered investment adviser. AI Bottleneck Portfolio and Uncle Sam's Checkbook are managed by Harold Tan, an independent Third-Party Pilot unaffiliated with Autopilot. Autopilot doesn't manage, validate, or guarantee this Pilot's portfolio. This interview reflects Harold's own views as of May 2026, edited and condensed, and isn't a recommendation to buy or sell any security or to follow any Portfolio. Individual stocks shown are illustrative and not a recommendation. Strategies linked to digital assets carry extreme volatility and risk of substantial loss; holding equities or funds exposed to digital assets is not the same as holding those assets. Where a Pilot uses an AI tool in their own stock-selection process, that tool does not advise clients, select investments for clients, or place trades. Investing carries risk-including possible loss of principal-and past performance is not indicative of future results. Figures current as of 10/09/26. View our Customer Relationship Summary and other important info at joinautopilot.com/disclaimer.

The performance data presented reflects the aggregated results of actual client portfolios managed by the Registered Investment Adviser (RIA), based on the trading activity in the securities contained in the relevant portfolio over the specified period. If fewer than 10 clients are invested in the portfolio, the data represents the aggregate performance of all clients following this strategy. Performance figures are shown net of advisory fees, calculated using the applicable subscription fee and an assumed $50,000 account size. Because subscription fees are flat rather than asset-based, accounts smaller than $50,000 would experience a higher effective fee percentage and lower net performance, while larger accounts would experience a lower effective fee percentage. Actual investor outcomes can vary due to factors like brokerage support for fractional shares, account size, trade timing, fees, and taxes, and past performance should not be interpreted as a guarantee of future results or typical of all portfolios.