Mind the Warnings… But Follow the Incentives

If you own AI stocks, the past few weeks haven’t been fun. I speak from experience…

Seemingly out of the blue, over the weekend, some of the most powerful people in artificial intelligence started sounding alarms.

The warning bells tolled, and a negative narrative took off like wildfire in the mainstream media – and just as rapidly across social media – hitting AI stocks heavily on Monday.

The sobering message was that AI could be advancing too quickly. The technology breeding so much excitement (and dollars) could become dangerous. Development might need to slow. We need stronger safeguards and perhaps stronger regulation.

Investors heard “slow down AI” and translated it into something much scarier:

The AI boom is slowing down.

Wall Street has this great habit of shooting first, then asking questions later.

But as an intelligent investor, before you make that same leap, consider something:

What if we’re reading the message backwards?

The Problem

Fear is incredibly powerful because it shuts down rational thought.

Wall Street has known this for years, but the news media has known it longer. A reader will click a doom-and-gloom story faster than a cute story about puppies. Every. Single. Time.

It’s in our wiring. Negative information moves faster and louder, which is what advertisers want. And right now, the fear is loud.

AI spending is exploding. Technology companies are pouring hundreds of billions of dollars into data centers, chips, memory, networking and power infrastructure.

Naturally, results-driven Wall Street has started asking when all that spending will bring returns.

At almost precisely the same time, the biggest names in AI began giving dire warnings about the technology they themselves are building.

That’s odd.

It made me wonder: Why are calls to slow AI becoming suddenly louder, after billions have been raised, valuations have exploded, and IPOs loom?

OpenAI and Anthropic are preparing to eventually go public. That means public-market scrutiny while the industry simultaneously faces questions about spending, profitability, and cheaper open models.

I think there is a logical business reason and benefit to sounding the alarm now.

That doesn’t mean AI dangers aren’t real. They very well may be. But as investors, our job isn’t merely to listen to what people say.

Our job is to understand the incentives behind their actions.

The Proof

Imagine you’re in a race to build the fastest car anyone has ever seen.

You spend billions. You hire the world’s best engineers. You build an enormous lead.

Then, at the crucial moment when you feel dominant, you announce racing could become dangerous.

Perhaps we need speed limits, safety inspections, expensive testing and licensing.

There may be excellent reasons for every one of those rules. But there’s something else worth noticing… You’re already leading by a mile.

The largest technology companies have already spent staggering sums building models, securing chips, constructing data centers and hiring talent.

Now imagine serious AI development requires extensive safety testing, outside audits, cybersecurity systems, compliance teams, lawyers and government approval.

Who can afford that?

The giants certainly can. The startup trying to catch them may not.

Regulation could therefore give the companies already leading AI another advantage. They may help influence the rules while expensive compliance creates another obstacle for smaller competitors.

Regulation doesn’t necessarily destroy a competitive advantage. I think, in this case, it might cement one.

There’s another incentive too…

Time.

Slower development gives AI companies more time to monetize what they’ve already built and demonstrate returns on enormous investments.

Think about the iPhone. Apple only completely reinvents every few years. Spacing them out gives Apple time to squeeze enormous value from each generation before moving consumers to the next.

AI companies have the same incentive. Winning the race to the frontier is great. But once you get there, you also need time to monetize your lead.

I’m not saying that’s why these executives are warning us. I’m saying investors should recognize that the incentives exist.

The Payoff

Everything for me always comes back to the stock market. It’s the purest expression of attitudes towards risk and opportunity.

Investors hear AI needs to slow down and immediately move several steps ahead. Shoot first…

Less AI means fewer GPUs. Fewer GPUs mean fewer data centers. That means less memory, networking, optical equipment, power, and cooling.

Therefore, sell AI stocks.

There’s just one problem… The first step of that chain is entirely wrong.

AI regulation does not mean AI disappears. AI safety doesn’t mean Microsoft stops building data centers, Google suddenly needs less compute, or Meta needs less networking bandwidth.

History shows clues.

Look at airlines. For decades, the government controlled which airlines could fly which routes and even what fares they could charge. The goal was a safe, reliable transportation system.

But there was a huge side effect… Regulation protected the airlines already inside the system. Competition was limited, and entry was difficult until deregulation began tearing down those barriers.

Banking offers another example.

Post-financial-crisis rules increased capital, reporting, compliance, and stress-testing requirements. Those rules addressed real risks, but complying with them costs a fortune.

JPMorgan can absorb those costs. A tiny bank cannot. That’s the part investors need to understand.

Regulation can protect consumers while simultaneously protecting the giants already in place. The two aren’t mutually exclusive.

Now, bringing this back to AI.

If AI really is becoming as powerful as its creators claim, we may need more testing, monitoring, cybersecurity, and infrastructure surrounding it.

Those things cost money. Lots of it.

If frontier AI requires massive compute clusters, specialized safety teams, audits, government reporting, and expensive compliance, who is best positioned to handle it?

Probably the companies already spending tens of billions of dollars doing it or about to raise hundreds of billions in the public markets.

The regulation intended to slow them down could actually make them harder to catch.

And remember, these monstrous companies have huge PR departments whose job is metering, collaring and spinning outgoing messages for one key beneficiary:

The company itself.

Those weren’t random tweets from a CEO’s couch on Sunday. I believe we witnessed a coordinated message intended to produce a desired effect.

Suddenly “Slow down AI” doesn’t have to mean, “The AI boom is ending.”

It could mean, “We want to keep the AI boom within our tight-knit group of leaders.”

Follow Incentives, Not Fear

Naturally, I don’t know for certain whether these warnings are a coordinated strategy or genuine fear.

But the people inside those rooms do. That’s precisely why investors should follow incentives.

• Take the warnings seriously, not blindly.
• Show me why AI safety means less compute.
• Show me why regulation means less data.
• Show me why increasingly powerful models require less memory, networking or electricity.
• And show me where the world’s largest technology companies are abandoning the hundreds of billions of dollars they’re spending to build AI infrastructure.

I don’t see it.

What I see is an AI boom that remains alive and well.

The spending continues. The models keep getting larger and more capable. The amount of data being created and moved keeps exploding. The need for compute, memory, networking, power, cooling, and optical connectivity isn’t disappearing.

The only thing that changed was the narrative. But narratives change faster than fundamentals.

Markets occasionally need to shake out the excess. Expectations get too high. Stocks get crowded. Then fear arrives, investors rush for the exits, and great companies get thrown out with everything else.

I’ve seen that movie many times.

Eventually, investors return to revenues, earnings, orders, backlogs, and the staggering amount of infrastructure still being built.

By then, prices may have already adjusted.

That’s the cruel thing about great investment opportunities.

They rarely feel comfortable when they’re actually available.

Six months from now, I suspect many investors will look back at this AI scare and see something almost impossible to recognize while living through it:

A missed opportunity.

So take the warnings seriously. Understand the risks. Follow the incentives.

But above all, watch what these companies do, not simply what people say about them. Because from where I’m sitting, the AI boom didn’t end… fear just put it on sale.

Regards,

Jason Bodner
Founder, Outlier Intel