Everyone loves an underdog story. The narrative writes itself: brave engineers in Tallinn and Riga cobbling together garage-built quadcopters, fueled by espresso and existential dread, ready to swarm the Eastern flank and stop the Russian bear in its tracks. It is romantic. It makes good television. It is also a dangerous fantasy that misunderstands modern attrition warfare.
I have spent the last three years watching venture capital and defense ministries throw millions of euros at bright-eyed founders building shiny, unproven hardware. I have seen the procurement slides, listened to the pitch decks promising low-cost autonomy, and watched smart people convince themselves that a handful of agile startups can out-innovate a mass-mobilized industrial base through sheer software cleverness.
They are wrong. And the obsession with agile, fast-moving Baltic defense startups is masking a systemic failure in how Europe thinks about deterrence.
Let us dismantle the lazy consensus.
The Garage Myth Versus The Logistical Reality
The core premise of the current defense tech hype is that speed equals survivability. If you can prototype a drone in six weeks and iterate code twice a day, you win. This logic works in Silicon Valley SaaS. It fails miserably when your hardware is sitting in a muddy trench in Ida-Virumaa and a Russian electronic warfare unit is dumping fifty kilowatts of broad-spectrum noise across the electromagnetic spectrum.
Startups pride themselves on moving fast and breaking things. In kinetic warfare, the thing you break is your own supply chain.
I sat in a briefing room last year with a defense technologist who proudly showed off an AI-guided loitering munition built by a Baltic team. It used commercial off-the-shelf components, ran custom neural nets for terminal guidance, and cost a fraction of traditional artillery shells. When I asked about component sourcing, the room went quiet. The microcontrollers came from Shenzhen. The batteries came from suppliers with obscure ownership structures leading back to intermediaries in Central Asia. The camera sensors were dependent on raw materials controlled by actors with zero allegiance to NATO.
If a real conflict starts tomorrow, these startup supply chains will evaporate in forty-eight hours.
True resilience is not about how fast you can solder a PCB in a co-working space in Tartu. It is about whether you can produce one hundred thousand identical, boring, resilient units that work with zero configuration by an exhausted nineteen-year-old conscript who has had three hours of sleep.
The Software Delusion
There is a dangerous fetishization of autonomy right now. Venture capitalists who have never heard a mortar whistle overhead are convinced that machine vision and edge computing will replace human judgment and brute-force mass.
The pitch goes like this: we build smart swarms, the AI handles target identification, and we bypass electronic jamming through sophisticated mesh networking.
Imagine a scenario where a squadron of these advanced autonomous drones crosses the border. Within seconds, they encounter localized GPS spoofing combined with directional RF suppression that blinds their commercial-grade optical sensors with cheap, high-intensity strobe interference. The brilliant neural network, trained on pristine datasets in a lab, suddenly treats a flock of birds or a burning tractor as a high-value priority target because the edge processor lacks the compute power to handle adversarial machine learning attacks at scale.
Complexity is a vulnerability. Every line of custom code is a new attack surface. Every advanced feature is a point of failure.
The startups building these systems are selling venture capitalists a software solution to an industrial hardware problem. You cannot patch a lack of high-explosive mass with an over-the-air firmware update.
The Venture Capital Tourism Problem
We need to talk about the money. Defense tech has become the newest playground for venture funds looking for impact investing with a high-octane PR spin.
They fly into Vilnius, drink craft beer with former military officers, and write seed checks for companies that have never scaled manufacturing past thirty units a month. These founders are brilliant engineers. They know how to optimize inference speeds and design sleek carbon-fiber airframes. But they do not know how to build a factory that can pump out five thousand units a day while under cyberattack, nor do they understand military bureaucracy well enough to navigate multi-year certification cycles without going bankrupt.
The result is a graveyard of brilliant prototypes. Companies burn through two million euros of seed capital, secure a vanity contract for a dozen demonstration units, and then stall because they cannot cross the valley of death into mass production.
The venture model demands hyper-growth and high margins. Defense manufacturing demands brutal, boring consistency, deep balance sheets, and the capacity to absorb massive inventory costs while waiting for slow-moving procurement officers to sign off. These two worlds are fundamentally incompatible.
What Real Deterrence Looks Like
If we want the Baltic states to actually secure their borders, we have to stop funding garage tinkerers and start investing in unglamorous industrial depth.
That means three uncomfortable truths:
- Boring is beautiful: Stop funding AI-driven swarm autonomy and start funding cheap, hardened, analog-resistant guidance systems that can be manufactured by the millions in standard automotive plants.
- Scale beats sophistication: A mediocre drone that you have one million of is infinitely more valuable than a sentient micro-jet that you have twelve of in a clean room.
- Consolidation is mandatory: The ecosystem is too fragmented. Dozens of tiny startups fighting over crumbs from national defense ministries is a waste of intellectual capital. We need aggressive consolidation into regional primes that can stand toe-to-toe with traditional defense contractors.
The current approach makes for great press releases. It makes founders feel like modern-day resistance fighters. But war does not care about your pitch deck.
Stop funding toys. Build an ironworks.
