Gold surge and AI boom raise new bubble warnings, strategist says

Gold’s explosive rally and the rapid rise of AI-linked stocks have raised warnings from global watchdogs about possible bubble conditions, sparking debate over whether markets can sustain their momentum. Analysts have pointed to unusual parallels between bullion and tech stocks, as both reach record levels amid heightened geopolitical tensions and aggressive corporate spending.

BNN Bloomberg spoke with Skyler Weinand, chief investment officer at Regan Capital, about the factors fuelling gold’s climb, the risks emerging in AI-driven markets and why leverage, borrowing and data quality will play decisive roles in determining whether a true bubble is forming.

Key Takeaways

• Weinand says gold’s rally is partly driven by foreign governments shifting away from U.S. Treasuries, though he argues the metal has limited intrinsic value.
• He warns valuations may be overextended, noting gold above US$4,000 an ounce could prompt investors to consider trimming holdings rather than betting against it.
• Weinand says an AI bubble has not yet formed, but rising corporate borrowing to fund massive AI capital spending could increase future risk.
• He argues AI models remain limited by their reliance on low-quality data sources and restricted access to primary material, raising concerns about reliability.
• Weinand says younger generations may become overly dependent on AI tools, making traditional skills — such as sustained reading — increasingly rare and valua

Read the full transcript below:

ANDREW: Let’s take a broad look at the markets, including gold and those AI stocks. We’re joined by Skyler Weinand, chief investment officer at Regan Capital. Skyler, thank you very much indeed for joining us. I’m not sure if you saw that headline — the Bank for International Settlements, the international banking watchdog, says gold may be in a bubble. It’s concerned that both tech stocks and gold have been rising this year, hitting records, and it says that’s very unusual. What do you think — could gold be overheated right now?

SKYLER: It’s definitely had a tremendous run, and part of that is this trade war, so to speak, where China and foreign governments are getting a little bit nervous about owning dollars and U.S. Treasuries, and are instead plowing that money into gold. And gold, for most purposes, has very little value. It yields nothing. There is some industrial use, but it really is just stockpiling. So if you look at gold, you look at crypto, you look at some of these AI stocks that are spending a tremendous amount of money — the big question looming in that bubble question is: Is this money going to productive use, or is this actually going to generate anything in the future, or are folks just getting nervous about where their money might be going?

ANDREW: Would you be inclined to bet against gold, though, after this huge run-up?

SKYLER: You know, gold’s always going to have that soft spot in an investment portfolio, where you should have one, two, three per cent of your portfolio in gold — maybe a little bit in crypto — just in case we hit World War Three and there’s a huge meltdown. But trading at $4,000-plus an ounce, it’s getting to that point where, well, maybe I should start selling my gold. So I’m not going to bet against it, but I’m going to think twice about, you know, if that jewellery gets inherited down, whether I wear it or I go sell it.

ANDREW: What about the tech stocks themselves? You think we’re not quite in an euphoric enough environment to actually call it a bubble?

SKYLER: Yeah, I don’t think we’re there yet. And the big reason I say that is leverage. These tech stocks just started borrowing within the last few months. Up until now, they were funneling a lot of cash into R&D and future investments. Only recently have Meta and Alphabet come out and started borrowing money to fund future AI expenditures. Now, the jury’s out on whether they’re going to run a profit on AI, right? And so you’re looking at hundreds of billions, if not trillions, of dollars in expenditures with the hope that somewhere out there they’re going to generate something out of that. So up until now, you take an Nvidia, for example, which has gone from a trillion- to a US$5-trillion company — but you had revenues behind that. You had earnings behind that. So we’re just hitting this point now where companies are starting to borrow, and it’s going to be a three-year time frame, I think, in order to figure out whether that borrowed money was actually worth it for them.

ANDREW: Because apparently evidence is emerging — one study, a widely watched study reported in The Atlantic in September, by a crowd called MITRE, found that programmers were using AI, but the AI was just not reliable. And one person said it was like shoulder-surfing — you’re watching an overenthusiastic rookie. And sure, it’s fast, but it makes loads of mistakes, so the programmers end up wasting time having to fix the mistakes. Now, I know AI is getting better all the time.

SKYLER: Yeah. The problem with AI is these models don’t actually have access to a ton of raw data and primary sources. If you look at the sources of data they’re using, the top two or three sources are Reddit and Wikipedia. And so they can’t actually access a lot of copyrighted data out there — music, primary sources, books, journals, etc. They don’t have access to that information. And you’re actually starting to see these machines create AI off of AI, instead of writing their own books or writing their own text, and that becomes a new primary source. So you’re kind of going down this rabbit hole of: What are we looking at? We’re starting to put this filter on — is this image real? Is this movie real? Is this text real or not? And mark my words, being able to read books and being able to read primary sources is going to be a superpower in 10 or 15 years from now, because the next generation won’t be able to exist without AI. But the data that AI is using really isn’t high-quality data.

ANDREW: You mean the next generation of people will be overdependent on AI?

SKYLER: Oh, for sure. I mean, my kids are 11 and nine, and being able to sit down and actually read a book for 20 minutes is getting harder as time goes on, because this generation is relying on YouTube, ChatGPT, the internet — you name it — for a lot of their ability to do homework and to do actual work. And like I said, being able to read a book a generation from now is going to be a superpower.

ANDREW: Whoa, that’s a scary prospect, isn’t it?

SKYLER: I mean, there was a study done that even reading actual text — printed-out material — you’re able to absorb about 25 per cent more information than reading it on your phone or on a computer screen, first of all. So then take: Hey, I don’t know how to do this problem. Back in the day — don’t use a calculator, write it out. Now it’s: Everything can be done by AI. Everything can be done just by writing in a prompt. And so the high-level AI folks now are: “Oh, I can set good prompts, I’m good at that.” So we’re getting to this precipice of: Tons of money is being spent — what is that actually going to result in? And we won’t see for three to four years, at least.

ANDREW: That’s interesting, yeah. I’m glad my kids got through school before this stuff arrived. My kids are in their early 20s — they missed the real white heat — but you’re right in the middle of it. Skyler, well, thank you very much, and all the best to you and your kids, and I hope you get them reading books. Thank you so much, Skyler. I really appreciate it. Skyler Weinand, chief investment officer at Regan Capital.

—

This BNN Bloomberg summary and transcript of the Dec. 8, 2025 interview with Skyler Weinand are published with the assistance of AI. Original research, interview questions and added context was created by BNN Bloomberg journalists. An editor also reviewed this material before it was published to ensure its accuracy and adherence with BNN Bloomberg editorial policies and standards.