AI is Overturning the Military World Order

Operation Spiderweb, Implications, and Impacts on Deep Tech Investing

DefenseTech and war are not my usual bailiwick. On the bright side (kind of), much of this content is adapted from a chapter in my upcoming book, which is how I had much of the information on hand. Now, of course, I need to update it where much of my commentary on “will happen” has now happened...

---

Lenin, supposedly, once said:

"There are decades where nothing happens, and there are weeks where decades happen."

This week was the latter.

Operation Spiderweb is the cause. Early this week, Ukraine attacked Russian airfields and claimed to have hit 41 Russian aircraft—including expensive, irreplaceable strategic bombers. Tu-95s are old—they first flew in 1952 during the Korean War—but they're still part of Russia's nuclear umbrella.

Due to various budgetary factors and degradation after the Soviet Union fell, they are, in fact, the majority of Russia's current intercontinental nuclear strike capabilities.

It isn't totally crazy—the US's own B-52 strategic bombers were also introduced in 1952. They are also still in service today. Often literally the exact same aircraft.

The US would also have a lot of trouble replacing all of their B-52s if they suddenly disappeared. Though at least the US would have other options.

This was paired with a cyberattack on Tupolev (the "Tu" part of Tu-95) that either added insult to injury (they put an owl gripping Russian aircraft on the website—the owl is Ukrainian military intelligence's informal mascot) or will have greater significance we'll find out later.

What does all of this have to do with AI? We don't know how autonomous the drones truly were—that'll likely come out in more reporting later. Supposedly, there was some aspect of autonomy.

But it does three things:

1. It shows the asymmetric potency of drones. This was already clear to most military planners and thinkers, but this makes it blindingly obvious.

2. It similarly shows the relative helplessness of expensive, conventional weapons platforms that are orders of magnitude more expensive. This may seem "unfair" to say in a sneak attack, but surprise strikes are certainly fair game in war—and, besides that, drones are doing just fine against expensive weapons systems like tanks in the Ukraine conflict already.

3. As Ben Thompson highlighted on Sharp Tech this week, it deals a massive blow to the Pax Americana world order. Why? The humble shipping container, which drives standardized global trade, was the vector used to infiltrate Russian borders. As he points out, the global trade system just doesn't work if shipping containers are assumed to be threats. Noah Smith points out that thousands of those containers arrive in the US every day—what happens if China suddenly pulls the same kind of op and destroys a huge portion of the US's multi-trillion-dollar air force and navy in minutes?

Turning Points

To explain why this kind of turning point is significant, it's useful to look back on history.

Bob Work, then deputy US Secretary of Defense, gave a speech in 2016. He described AI and autonomy as the Third Offset—the third major technology that would completely change the nature of war. What were the first two?

The First Offset was nuclear weapons. The Second Offset was basically the Information Age—microprocessors, sensors, and stealth.

AI's impact is still "to be determined" (well, not really, but I'll explain later). Nuclear weapons are obvious. Why does the Information Age, with our IT stuff and precision weapons, count? It seems a little odd next to the other two.

The answer, in part, was the 1991 Gulf War with the US coalition that invaded Iraq (the first time, not the second).

It is, after all, what helped China realize that its military was helpless against the US.

The US crushed Iraq in that war. Air power was ridiculously effective—BVR (beyond visual range) missiles with US aircraft, cruise missiles, etc. all dominated the skies. Precision weaponry and modern command and control made the entire "ground war" a rout—a significant portion of Iraq's military ended up gone before it could even start reacting.

That operation, Desert Storm, was closely studied by the PLA (Chinese military). Why? China had believed their military was fairly powerful. It had all of the conventional hallmarks in terms of numbers, powerful ground platforms, and training. It was something that could likely crush a lot of militaries in WWII. Iraq was also similarly "modern." Sure, China had more stuff than they did and likely expected they were more powerful, but Iraq's military was categorically the same as China's.

That's the US's—and other modern armies'—issue. They may be far more powerful and technologically sophisticated than Russia's. But they're categorically the same thing. They would likely fare similarly in such an attack. None of the fancy, special, multi-billion-dollar systems are designed to fight cheap, autonomous flying drones that, not long ago, were children's (and "big kid" tech nerd) toys.

AI and Autonomy

I live in the Bay Area. I hear a lot of, "We should not hand the decision to kill to robots or AI!"

Ah, so little do they know—we passed that point a long time ago.

Since 2023, there have already been Ukrainian drones with low-power chips that use AI to identify manned tanks—and autonomously attack them. They don't need to check in with a human operator.

Sure, this "AI" may be more classically called "machine learning"—it isn't exactly the latest ChatGPT model running on there—but that hardly matters. Categorically, that line got crossed. And practically, it's extraordinarily useful.

Congratulations! We've already reached AI drones killing people. And it's likely just beginning.

Again, just how much autonomy versus teleoperation played a role in Operation Spiderweb is still a question. It's almost certainly the case there was some. It seems unlikely that such a large, careful operation that was planned for "one year, six months, and nine days" was left to spotty wireless connections.

Regardless, Ukraine—which has never been flush with military hardware to the level it desired, even before the Trump Administration—has been showing just how effective adapted, cheap "tech stuff" is for combatting conventional militaries.

To some degree, the biggest loser (from a macro perspective) of this attack is the US. It's the one with the biggest, most expensive horde of this military hardware that's now... likely outdated.

I gave an analogy of the Iraq War. Noah Smith gave an analogy of the British Navy in 1940—proud of its own incredible power, especially of its battleships—being utterly humiliated and forced into a quick retreat by Japanese aircraft. A battleship and battlecruiser received the dubious honor of being the first capital ships actively defending themselves to be sunk by air power in the open sea.

Drones and AI

I've seen multiple startup pitches at this point for drone swarming using full AI control (dual use, of course...). The potential of that technology against conventional weapons platforms is terrifying.

But honestly, just watch some videos of drone shows in China. Those are pre-programmed with computer control, but it does show how effective the platform can be—especially with the right AI brains behind it.

A screenshot from that drone show video I linked. It is, after all, just entertainment, right? That’s why they comment on how they can’t be blocked and the nationalistic pride they have.

We've already crossed a Rubicon here. The US and China already were pushing their AI capabilities, in part for new weapons in the coming war.

Investing in a Changing World

This has... not been the best environment for investors. Generally, the preference of most investors is stability. Markets are hard enough—you don't need to add in the volatility of, say, tariffs that move from nothing to historical levels to historical records (and back again) from day to day.

Obviously, this creates opportunity as well. But those opportunities tend to be much more binary now.

On one hand, the upside is that technology is now clearly a way to gain geopolitical advantage—and military power. Just look at Anduril and Meta's new partnership to develop XR for the US military. Inevitable, that’s going to be a place that could be (as it was in history) a huge source of revenue for tech companies.

On the other hand, you do end up needing to make your bets much more carefully. I'm not as much of a "doomer" about the US's future as... well, many people I talk to, especially in Asia. I'm actually not much of a doomer at all—the US still has a huge number of technological, geographic, and cultural advantages. At least for now.

But even most of those "US doomers" expected a slow stagnation until the eventual sack of Rome by the Visigoths.

I wasn't so sure. A definitive outcome in the Taiwan Strait or some other confrontation might change the landscape quite quickly. Now? I think that's even more the case. With so much technology throwing so much of the "historical mass" of military hardware potentially out the window, the entire world order ranking is a bit of a free-for-all.

"There are decades where nothing happens, and there are weeks where decades happen."

We could very well wake up in a world where the US—who technically is still ahead in most metrics in AI—will suddenly reestablish a commanding, dominant position. I'm not one of those who thinks that AGI is just a few years away, but for the sake of argument, something like that (or, more likely, a far more diverse combination of systems and AI capabilities that are less flashy but extraordinarily effective) is first reached by the US. It'll be like 1945—the US, alone, with nuclear weapons, shocking the world.

Or, the US could rapidly be humbled. Maybe it's China, but maybe it's some other innumerable conflicts flaring up around the world as Pax Americana weakens. Is it that the Houthis sink a US aircraft carrier with a bunch of cheap, AI-powered drones from Iran? Do we see some US ally, armed with lots of modern weaponry, get crushed by Chinese-supplied AI drones?

We'll have to see.

I've already been somewhat put off by how much my job in terms of deep tech investing has taken on a tinge—or much more than that—of national security and power. I'm not directly in DefenseTech, but it's kind of hard to avoid any advanced technology, AI included, falling into that category. As this week showed, however... that likely isn't going away anytime soon. Maybe not even within my lifetime.

---

Thanks for reading!

I hope you enjoyed this article. If you’d like to learn more about AI’s past, present, and future in an easy to understand way, I’m working on a book titled What You Need to Know About AI that will be published later this year.

Sign up below to get updates on the book development and release here!