Machine-speed warfare has arrived — and most military and political systems are still built for human-speed interpretation.
In this episode, we explore how drone saturation, cheap precision strike, AI-assisted targeting, and electronic warfare are creating the conditions for escalation not through intent, but through misinterpretation.
From Gaza and southern Lebanon to Ukraine and the U.S.–Mexico border, small, expendable drones now shape the tempo of conflict. They drift, jam, spoof, maneuver and react faster than commanders can fully interpret — and that gap is becoming the defining risk of our century.
A recent analysis of the Ukraine war describes drones as cost-imposing systems evolving faster than traditional military structures can adapt. Erik Prince, in a March 2025 security lecture, warned that inexpensive FPV platforms and AI autopilots have democratized precision strike far beyond what state doctrine ever anticipated.
This episode asks the central question of the drone age: What happens when autonomous or semi-autonomous systems respond to noise, spoofing, or jamming faster than humans can de-escalate?
The future of conflict will not be determined only by who has the fastest systems, but by who can encode restraint into them.
Power is cheap now. Speed is cheap. Intelligence is cheap. Wisdom is scarce.
This is part 2 of a 2-part series.
Chapters: 00:00 Introduction 01:15 The Tom Clancy Analogy - "Sum of All Fears" 04:48 Robust vs Loose Systems 09:28 What Nations Need to Do 09:35 Medium Powers 10:39 Israel 12:22 Taiwan 14:02 United States 16:15 Discernment at Machine Speed
When people think about revolutions in warfare,
they imagine the first jet engines,
the first nuclear test, Sputnik in orbit,
the first Gulf War with laser-guided bombs on CNN.
Those moments came with fanfare,
a sense that history had just shifted.
But the most consequential shift in warfare
since the nuclear age arrived quietly
through little workshops, YouTube tutorials,
telegram channels, and cheap components sourced online.
It arrived via teenagers, militias, smugglers,
and small companies improvising with hardware
that costs less than a nice phone.
This revolution arrived as the drone,
not just the big, great predators and reapers,
though those were part of the early story,
and 30 years ago, they were very secret.
We're talking about the cheap, expendable,
modified, improvised flying machines
that anybody can buy, hack, and deploy.
Welcome to Global Risk Profile,
where we break down the headlines that matter
and the ones that should.
Now let's go to a fictional story
and bring it back to today's reality.
We go back to our favorite author
in this podcast, Tom Clancy.
He was not a control theory guy,
but he grasped something deeply important about war.
Escalation can come from misinterpretation
under time pressure.
In his book, The Sum of All Fears,
the plot hinges on a false flag nuclear attack
designed to make the United States
and the Soviet Union blame each other
and stumble into war.
A lost Israeli nuclear device from the 1973 war
is secretly rebuilt by terrorists
and detonated at the Super Bowl,
while coordinated attacks in Europe
are staged to look like Soviet aggression.
The deception only begins to collapse
when the hero Jack Ryan traces the bomb's
isotopic signature and realizes
it wasn't a new Soviet weapon,
but an old Israeli one.
The scheme was basically a manufactured crisis
meant to destroy peace by triggering escalation
through misinterpretation.
The core pattern looks like this.
Events create ambiguous signals.
Humans interpret those signals under stress.
Each side's interpretation feeds back
into the other's picture,
and the loop tightens until escalation feels forced
on both sides.
Now, let's bring drones and autonomy into that model.
Picture two opposing drone swarms
operating along a tense border.
One drone loses GPS lock due to spoofing
and drifts towards a sensitive site.
To the opposing swarm's algorithms,
that looks like a probe.
The defending swarm responds,
perhaps automatically,
with a short jamming burst to protect itself.
The first swarm registers hostile jamming
and switches to evasive or defensive behavior.
That could mean spreading out,
changing altitude,
or shifting closer to their own protected assets.
To the defending system,
those evasive maneuvers look like
more aggressive positioning.
It tightens its defenses,
maybe even preemptively targeting
what it now classifies as high-risk contact.
All of that can happen at machine speed.
By the time a human sees the trend,
a few seconds or a minute later,
dozens of autonomous decisions
have already been made on both sides.
Nobody sat in a war room and said,
let's escalate.
But the system escalated anyway.
Clancy wrote that story at the strategic level
with submarines and nuclear weapons.
Autonomous drone warfare creates miniature versions
of the same story at the tactical level,
potentially hundreds of times per day.
The danger isn't that someone wakes up
and decides to start a war.
The danger is that imperfect, stressed humans
are trying to supervise ultra-fast machines
in a noisy environment full of deception.
You can't fix that with better hardware.
You need doctrine, you need design, you need restraint.
Which brings us to the next question.
How do different states choose to design these systems?
Once you see that machine speed feedback loops
can spiral, the way nations design their drones
and their rules of engagement
becomes the heart of the story.
Not all drone systems are created equal.
Some states try to build robust systems.
That means conservative rules of engagement
embedded in software.
Multiple layers of verification
before lethal effects are triggered
and clear off-ramps and circuit breakers
when sensors disagree.
Also a bias towards de-escalation under uncertainty.
Other states may build looser systems
with faster, more aggressive automated responses,
a greater willingness to operate in gray zones
and ambiguous airspace,
and higher tolerance for collateral risk.
And a doctrine that treats ambiguity as an opportunity
rather than a red flag.
On a different axis,
some actors are technically competent
with good sensors, good comms, and well-trained operators.
And others are crude using hacked together systems
with poor discipline and bad situational awareness.
I guess some smugglers maybe.
If you sketch that out,
you get basically a two by two matrix.
Robust and competent.
For example, the US, the UK, France, Japan, Australia,
and in many domains, Israel.
Loose and competent.
China in some areas, Russia in some of its areas.
Robust and crude.
Perhaps newer NATO drone users
trying to behave responsibly with limited tech.
And loose and crude.
Like militias, cartels, insurgent groups,
and other non-state actors.
The fourth category is the pure clancy risk in miniature.
Lots of firepower, very little discipline,
and almost no concern for escalation.
As an example, in November 2025,
a swarm of unidentified drones
approached US Navy ships in the Red Sea.
The ships held fire for 14 minutes
while trying to establish positive ID.
The drones turned out to be Houthi,
but 14 minutes was an eternity in machine time.
This is a demonstration
of the robust versus loose dilemma in real life.
Thus, the more subtle danger
is the interaction between categories.
A robust competent system
facing another robust competent system
can, in theory, remain stable
if both sides care about avoiding accidents.
A robust system facing a loose system is under pressure.
It wants to deescalate,
but may be repeatedly probed and tested.
A robust system facing multiple loose or crude actors at once
risks information overload and misattribution.
Who actually did that?
You can hear the strategic worry
in how critics talk about this.
Eric Prince, in that 2025 talk, wasn't subtle.
He argued that Western militaries
have been slow to adapt to cheap drones
and that innovation is coming from small teams
and private companies
long before big defense bureaucracies take it seriously.
He pointed out that the US Air Force
initially wanted nothing to do with armed predators
and that it was the CIA and General Atomics
that pushed that evolution through.
You can agree or disagree with his conclusions,
but the deeper point is hard to escape.
If democracies move slowly and insist on high standards
while authoritarian or non-state actors move fast
and accept higher risk,
then robust and loose systems
will collide in a very unstable way.
That's not a reason to throw out our standards.
It's a reason to deliberately design systems
that are both fast enough
to survive in machine-speed environment
and disciplined enough
to prevent constant accidental escalation.
So what does that mean in practice
for different kinds of states?
Let's bring this down from theory
into concrete national choices,
starting perhaps with medium powers.
For example, France, the UK, and their peer countries.
They're not, at the moment,
superpowers in purely industrial or demographic terms,
but they're expected to operate as serious military actors
in coalition settings.
In Europe, in the Middle East, in the Indo-Pacific,
they can't afford to be free riders.
They have to help.
That means they need to build ecosystems, not toys,
not just a handful of high-end drones,
but layers of cheap drones.
Loitering munitions, naval drones, and ground robots,
all tied into NATO's command and control systems.
Their job is not necessarily
to outproduce China in raw numbers.
Their job is to make coalition warfare workable
in machine-speed world.
Now let's go to countries living at the edge of conflict.
Israel is already living in this environment.
It faces Iran, Hezbollah, Hamas,
and other actors who are willing to operate
with looser constraints,
sometimes quite competent in specific niches
and often backed by a state sponsor
that is willing to ship drones
and know-how across the region.
We saw Houthi drones reach Tel Aviv in 2024.
We see Hezbollah using cheap quadcopters
and loitering munitions near the northern border.
And we see Iran exporting drone designs and components
to proxies from Yemen to Lebanon.
Israel's drone doctrine has to be multi-mode by definition.
One mode for Iran as a state actor with strategic system,
and another for semi-state actors like Hezbollah,
and another for more chaotic environments like Gaza,
and a mode of extreme caution anywhere.
Russian and American assets may be nearby.
At the same time, Israel has to maintain
detailed telemetry and video logs
as a matter of survival in the information space.
In a region where everyone is accusing everyone else
of war crimes, having clear time-stamped evidence
of what actually happened isn't just nice to have.
It is a strategic necessity.
Israel can't afford to be loose,
but also can't afford to be slow.
That combination, fast and disciplined,
in a region full of loose actors,
is exactly the situation we described earlier
as highly unstable if you get it wrong.
Taiwan's situation is different in geography,
but similar in existential stakes.
It faces a much larger adversary
with growing drone, missile, and naval capabilities,
an adversary that has studied Ukraine, Gaza,
Nagorno-Karabakh, and experimented
with its own drone swarms and unmanned surface vessels.
Taiwan cannot win by trying to mirror
the PLA system for system.
It will always be outnumbered,
but it can aim for denial, not dominance.
Large numbers of cheap aerial and maritime drones
designed to make any amphibious assault extremely costly.
With robust fallback autonomy
for when communications are jammed,
pre-planned routes, kill boxes,
and behaviors that don't require real-time remote control,
and deep integration with civilian sensing networks,
fishermen, coastal radars, commercial satellites,
but with tight centralized control over lethal decisions.
A doctrine that carefully distinguishes
between peacetime gray zone harassment and actual war,
so automated responses to routine PLA harassment
don't spiral into something else.
Taiwan's best hope is a system that is flexible,
distributed, and hard to decapitate,
but also predictable and controlled enough
that it doesn't accidentally trigger the war
it's trying to prevent.
And finally, the United States.
The U.S. still has unmatched global alliance networks,
enormous R&D capacity, and a defense industrial base
that can scale if it chooses to.
But it also has slow procurement cycles
and a lot of legacy bureaucracy.
If the U.S. treats drones as just more platforms to buy,
it may lose the plot.
Its role now is to be the architect of stability
in a world of autonomous and semi-autonomous systems.
That means building a common software
and doctrine architecture for autonomy,
something allies can plug into with shared interfaces
for identifying friend or foe,
rules of engagement models,
telemetry logging, and escalation controls.
Making programs about standard as much as hardware
so that many vendors can support one operating environment
with many platforms,
but one logic for safety and coordination.
Creating the plumbing for machine speed crisis management,
shared black box data from serious incidents,
agreed formats for exchanging logs
when something goes wrong,
and digital hotlines that can operate
at the tempo of machines
rather than the tempo of fax machines.
And practicing for multi-actor scenarios
in which the U.S. systems are interacting
not just with a single adversary,
but with allies, proxies, and wildcards
in the same airspace.
The analysts at RAINN and elsewhere warn
that without a coherent architecture like this,
the U.S. and its allies risk fielding a patchwork
of incompatible drone and counter-drone systems
that look impressive in isolation,
but fail under real stress.
In other words,
if you don't design stability into the system now,
instability will design itself later.
For most of human history,
war was bounded by human limits.
Humans could only march so fast,
reload so fast,
process so much information.
Mistakes were made, disasters happened,
but the tempo was fundamentally human.
We now seem to be moving into a world
where that is no longer true.
In this world, machines see first
through persistent sensors and AI pattern matching,
move first at machine reflex speeds,
and sometimes misinterpret first
under conditions deliberately engineered by an adversary.
The great danger of this era is not just malice.
It's also misunderstanding with consequences.
Misunderstandings that unfold far faster
than our traditional mechanisms of diplomacy,
oversight, and public debate.
Tom Clancy imagined such a chain reaction once
at the nuclear level in the sum of all fears.
We seem to be approaching a future
where smaller versions of that chain reaction
can occur dozens or hundreds of times a day
in many theaters at once.
Some of them could escalate also to a strategic level.
That puts a new kind of pressure on us.
So the question is no longer simply
who has the most powerful systems.
It's who can build systems that are powerful,
fast, and yet constrained by wisdom.
Because intelligence is getting cheaper,
hardware is getting cheaper, autonomy is getting cheaper.
What is rare is discernment.
The ability to distinguish signal from noise,
aggression from incident, opportunity from trap.
The countries that will navigate this era best
are not necessarily the ones with the biggest budgets.
They are the ones that can combine machine speed execution
with human speed judgment under doctrines
that assume errors will happen and work to contain them.
In that sense, the hardest work is not technological at all.
It is moral and institutional.
We have to decide what kinds of systems
we are willing to build,
how much power we are willing to hand them,
and what kinds of safeguards we will insist on having
even when we are afraid of falling behind.
Unlike a Tom Clancy novel where there is an author
that writes the final act,
here there is us and the systems we choose to create.
The question is whether our wisdom can keep up.
That's it for this episode of Global Risk Profile.
If you found this useful, thoughtful, or just different,
please subscribe, share, and leave a review.
Remember, the headlines may change,
but the deeper patterns rarely do.
Thanks for listening and stay sharp.