Dear all,

when I co-founded DTM, I was working on a book project that I had to pause. I am now spending a few quiet days by a river over the summer, to get back into this book project. Working title is "North Star, The 12 Formulas Behind the World's Most Valuable Deep Tech and AI Companies."

It is a playbook that hands every builder and investor of a category-defining Deep Tech and AI company the formulas to reach a North Star narrative that compounds into capital, customers, and enterprise value. It is about levelling the field, so that the venture with the best science, not the best storyteller, is the one that wins the room.

This is one of the chapters.

On 11 May 1997, on the thirty-fifth floor of a Manhattan tower, the greatest chess player alive resigned to a machine. Garry Kasparov, reigning world champion, lasted nineteen moves against IBM's Deep Blue before conceding, the first reigning world champion ever beaten by a machine in a full match. "I lost my fighting spirit," he said. The headlines wrote themselves: the thinking machines had arrived.

Except they hadn't. In 2011 IBM's Watson demolished the two greatest Jeopardy! champions on national television, and again the headlines declared the machines had come. Again, almost nothing followed: Watson toured hospitals and call centres, and the productivity revolution failed to show up on schedule. Then, in November 2022, a small lab put a plain text box on a website. You typed a question; for the first time a computer typed back a fluent, useful, human-sounding answer. Within two months it had reached a hundred million people, the fastest any consumer product ever had. The science had existed for years. What changed was everything around it: the models had crossed a threshold of scale, the compute to run them had turned cheap and abundant, and someone had finally wrapped it in an interface a child could use. The product was ChatGPT. The moment, not the technology, was the story.

This is the truth every deep tech and AI founder eventually meets: in venture capital, being too early is indistinguishable from being wrong. The graveyard is full of the brilliantly premature. IBM shipped the first smartphone in 1994, thirteen years before the iPhone. Better Place raised $850 million for battery-swap stations and sold roughly 750 cars before going bankrupt in 2013, a decade before EV demand arrived. The engineering was finished. The inflection point was not.

Great startups don't invent the wave, and they don't merely ride it. They accelerate the one the world has already committed to riding. A team building autonomous EV chargers isn't betting that electric vehicles will happen, they already have; it's riding the gap Better Place died waiting for. A grid-storage company isn't selling the idea of renewables; it's selling the consequence, the sun and wind keep their own schedule, and the grid has no memory.

The formula for defining your Why Now:

[Shift] is happening, so the window for [our contribution] is open now.

Many investors treat a missing "why now" as a reason to pass before they reach your technology. A strong one answers the question behind every pitch: why didn't this exist five years ago, and why will someone else own it in five years if you don't move today?

A real inflection point has a tell, it is already discussed in the general media and accepted as common knowledge. You shouldn't have to convince anyone that battlefield drones, the AI build-out or European rearmament are happening. Call it the journalist test: could a reporter confirm your "why now" without your slides?

Inflection points come in four flavours:

The strongest why-nows stack two or three at once.

But naming the shift is the easy part; it's usually common knowledge. The money is in the second-order consequence, the less obvious bottleneck the shift creates downstream. EVs arriving was obvious; the charging-and-grid gap was the opportunity. The AI data-centre boom was obvious; far less obvious was that it would be throttled by a century-old business, the giant grid transformers that connect a site to power now run multi-year backlogs, and that queue is the opportunity. Don't sell the wave everyone sees. Sell the gap behind it.

Defence: downing a €500 drone with a legacy missile can cost a thousandfold more, so mass drone warfare is here, and the window for affordable, sovereign air defence is open now.

Photonics: the frontier constraint is shifting from raw compute to moving data between chips fast enough to feed them, so the window for optical interconnects inside the data centre is open now.

Agent oversight: as AI crosses from advising to acting, the bottleneck becomes trust, not capability, so the window for tooling that audits and constrains autonomous agents is open now.

Cooling: air cooling has run out of headroom exactly as chips run hotter, so the window for liquid and immersion cooling as default data-centre infrastructure is open now.

Biosecurity: the same models that can design a new protein can design a new pathogen, so the window for AI-native biosecurity and screening is open now.

The "why now" isn't just a slide in your deck. It's the bet under the whole venture, the reason to point years of your life at this problem now rather than any other decade. Get it right and the market carries you. Get it wrong and you're Deep Blue's opponent: the finest in the world, beaten less by a better idea than by bad timing.

With wishes that your timing is impeccable, and that no one ever makes a documentary about how you were fifteen years ahead of your time,

Martin

Europe’s next robotics wave is moving beyond fixed automation. A new generation of early-stage companies is building robots that can learn, manipulate, adapt and work inside existing industrial environments — without requiring factories to be redesigned around them.

→ Acumino was one of just 15 startups selected for Google DeepMind’s first European Robotics Accelerator in 2026

→ Mimic’s robotic hand has 16 degrees of freedom and is designed specifically to mirror human anatomy closely enough for imitation learning. 

→ The team says it has developed a dual-arm robotic system for under $1,200. Its modular architecture is deliberately designed almost like LEGO for physical AI. 

The defense-tech revolution has a catch: the drones, autonomous systems and expendable platforms reshaping modern warfare may grow much faster than traditional fighter jets — but they could still be significantly worse businesses. 

That is the provocative conclusion of a new analysis by Boston Consulting Group in their July 2026 report, “Are We in a Defense Tech Bubble?”: cheap, fast and software-driven defense systems are gaining battlefield relevance and investor attention, yet the biggest and most durable profit pools remain with the old-school primes.

The report was produced together with equity-research firm Vertical Research Partners and asks a question investors and incumbents increasingly need to answer: Where does all the new defense-tech money actually turn into sustainable profit?

Here’s 5 key takeaways:

1. The startups are growing faster — much faster.
BCG expects traditional high-end platforms to grow only around 2–3% annually through 2033, versus 15–20% for affordable-mass systems and a massive 35–40% for expendables. But there is a huge base effect: even in 2033, exquisite platforms are still expected to account for more than 80% of the market.

2. The money is still in the fighter jet.
In 2025, exquisite defense-aviation systems represented roughly $65 billion of spending across the US and EU, compared with around $5 billion for affordable mass and just $55 million for expendable systems. Translation: the disruptors may own the growth story, but the incumbents still own the market.

3. Sustainment is the hidden profit machine.
Traditional platforms do not stop making money when they are delivered. BCG estimates that roughly half of the lifetime profit of an exquisite system can come from maintenance, spare parts and upgrades over decades. For the F/A-18E/F Super Hornet, that means more than $10 billion in sustainment profit across a $20.3 billion lifetime program. Expendable systems, by contrast, make almost all their money at the point of sale.

4. Venture-backed defense companies carry far more risk.
Traditional primes often develop new platforms with government-funded R&D and negotiated contracts. New defense-tech players typically finance much more development themselves — backed by VC or private capital. The upside: they own the IP. The downside: if demand disappears or the technology loses, they eat the loss. BCG notes that most affordable-mass and expendable manufacturers are still not profitable despite strong demand.

5. The future is probably not old versus new — but both.
BCG’s recommendation to the primes is essentially: don’t blow up the core business chasing the hype. Instead, treat drones, autonomy and expendables as a separate portfolio layer — using partnerships, minority investments, acquisitions and ring-fenced business units with faster operating models. The battlefield needs mass and autonomy. The balance sheet may still need the fighter jet.

Europe’s full-blown autumn event circus hasn’t started yet. Good news: that means fewer packed halls, fewer “quick coffee?” messages — and a bit more signal.

Three events worth sneaking into the calendar before conference season properly goes feral:

Space for Inspiration 2026 — Copenhagen, Denmark
September 1–2
For: Space founders, investors, corporates and anyone betting on the orbital economy.
Less “look, another rocket”, more: what actually gets built, bought and monetised in space next — from orbital infrastructure to exploration and new commercial markets.

QUANCOM 2026 — Trento, Italy
September 1–4
For: Quantum builders, researchers, infra nerds and investors who have moved past the PowerPoint phase.
This one goes straight for the plumbing: quantum hardware, systems and the infrastructure needed to turn scientific breakthroughs into something remotely scalable.

MSPO 2026 — Kielce, Poland
September 8–11
For: Defence tech founders, primes, procurement teams and investors hunting for what comes after the drone hype.
A major meeting point for Europe’s defence ecosystem — and a good place to watch where demand is moving across autonomous systems, counter-UAS, EW, sensors and defence AI.