As I’m writing this, it’s Christmas 2025, the end of the year, and a wonderful time to enjoy space with family, hopefully a chunk of time off, and a moment to breathe before the introspective/retrospective punch of New Years (!)

The topic of the issue?

The AI investment bubble.

First of all - what is a bubble?

Using AI while we have it, the quick definition is: “An economic bubble is a market phenomenon where asset prices (such as stocks, real estate, or commodities) rapidly increase far beyond their intrinsic, fundamental value. Driven by investor hype and speculation rather than value, this unsustainable boom continues until confidence crashes, causing a rapid, severe price drop known as a ‘burst’“

Here’s a pretty solid video breakdown of the current state of AI investment, with a handy infographic.

One extremely direct example is OpenAI, the company of ChatGPT, and the group that hit the gas on the whole situation. OpenAI needs to make $125 billion dollars annually, with a rough deadline of late 2028 to hit that. In 2025 they hit about $20 billion, but with staggering losses relative to income (Q1 they made $4.3 billion, and lost $13 billion, for roughly 3 dollars lost for every 1 earned).

The good news is, there’s a perfectly reasonable path to hitting that revenue target. Their growth has barely slowed, and that target is “only” around half of Microsoft’s annual revenue, a little more than a third of Alphabet (Google’s) annual revenue.

The bad news is that their current growth is fueled in large part by circular investment. As in, they have a massive investment from the chip maker, Nvidia… as a re-investment on a percentage of the money OpenAI is going to spend on Nvidia chips. There are some major questions about this deal, and it just looks shaky, at best.

Another good tidbit is that the massive investment in AI is, in part, funded by massive spikes in all of the involved companies’ stock prices, which they’re able to reinvest. That’s bad too though, because it means tight competition, short supply, short timelines, and every single piece of the pie that everyone needs keeps getting more expensive. Every company involved is in some way investing in the others, and they’re all driving up each others’ value while costing each other (and themselves) more at the same time.

On paper they’re also justifying this eye-popping company valuation by showing “growing user demand.” Which on the user side we see as all of those AI tools and features being crammed into every single software platform you’re used to using, and each company counts you for each feature you use, not how many of “you” there actually are, or whether or not you have any paid subscriptions.

BUT. There are legitimate ways AI companies can meet these revenue targets, and they’re not even that far out of reach:

So the overall picture is (surprise!) mixed. There’s cause for concern/pessimism, but plenty of room for optimism as well.

…to dip into the dark one more time though, it’s important to note again that this is a highly competitive space, there are multiple LLM AI players in the game, and not all of them are going to hit their targets. The worse part about that is because of the enormous amount of money in ALL of them, any of them going under will have significant repercussions, and potentially could bring down other players who otherwise have made it (by tanking Nvidia’s value, for example). That also means that for each company’s big win, they’ll get a smaller win than they could have in a more open market. It’s a staggering arms race right now.

Cross your fingers - but there are both opportunities, and hazards.

*Please note this is NOT formal investment advice.*

In my first issue, I covered two startup companies and their founders, both of whom represent great examples of where the big AI companies can buy in to profits. To augment internal features, and/or gain immediate profit opportunities from what these players are already doing.

Both Clarity Inbox and IntentPost stand out for being small, nimble, and having immediate paths to profitability in ways little guys can jump on where big players will have a harder time seeing.

Both companies also represent companies that can do what they do on their own, even if the bigger AI companies end up not making it. That’s another important “out” as we’re talking about startups - if the big players, or even some of them, go down, AI startups need to have roadmaps to survival too.

I also mentioned, above, tangential industries like humanoid robots and commercial manufacturing. There are several other somewhat obvious opportunities here too:

Now that’s all speculation, very specific advice, and a lot of information that may not apply to the companies reading this.

When it comes to actually surviving in the AI age, as a small startup, there’s one crucial thing to focus on: Positioning.

That’s always true, but because of how much of the room’s air large-scale AI takes, you have to choose between tying yourself to - or distancing yourself from - the biggest players. How do you stand out from them? How do you function by yourself? Are you NOT using AI and so a little bit insulated, or do you have a clear understanding of what your tools do if AI crashes?

We’ve all had some practice here too, with even just the outright backlash against the flood of AI tools. More specifically against hallucinations and “slop,” every AI startup I’ve seen has already practiced messaging how these problems don’t affect them, what they do personally to mitigate it, or use it as a reason for why they’re not using AI in the first place. All of those are lessons to re-purpose here!

This also is a fascinating tightrope moment to bring to angel investors and VCs. Coming to the table with a rough understanding of the AI market at the top level, and where you either fit in or are insulating yourself against.

Investors are in a tight space here. Having almost universally invested in AI at some level, needing that investment to succeed, but also needing to hedge their bets to prepare for worst-case scenarios. This is an excellent time to have blunt, straightforward, “What if?” conversations about where every company stands whether AI successfully expands, implodes, or some combination of both.

*Being a Bonus Issue to Week 1, these are the same companies and founders from last week - but they’re squarely in the focus of the topic today as well, so it bears repeating them!

I want to showcase and shout-out the folks doing really, really hard work. Who are very early stage, but with something special - and I want shout them out while they’re going through major growth:

Startup News Spotlight

In a shout-out both to The Portugal News and European AI progress, British startup Zango AI is expanding their presence in Portugal heavily. It’s one of the many, many cases of small, hyper-focused AI companies (especially one mixing directly in FinTech) with potential to both right a rising AI tide, or survive a general AI crash.

Rumbles around the World

Right now, the AI market in the US is red, red hot. It’s explosive, massive, and there are no guardrails. Is it where the top talent and biggest players are? Sure.

You could also argue, on many levels, that the US AI giants are misusing the technology. I recently compared the vibe-coding platform Lovable to v0 and Replit, and Lovable just buries the bigger, US-based competitors on every metric in the most hilarious way possible: It actually works.

The EU’s pragmatic approach and focus on solving smaller problems first has seen some of the cleanest, most successful implementations of AI tools that I’ve seen, anywhere.

Europe is far from insulated from the US financial market, we’ve seen that many times. But you have a lot of varied tech talent, actively learning and understanding AI, and while it’s not as red hot as the US market it’s also FAR less exposed right now, and a perfect place to explore smaller-scale, safer investments that will survive whether Big AI does… or not.

Want to support my work? Share this article of course, and/or come say hello on LinkedIn!

Ready for more? Read my other articles and subscribe here!

This newsletter is lovingly, joyfully, crafted on beehiiv.

Whew! You made it to the end!

See you next week ;)