Newsletter Archive
Listen to Our Podcast
Dear Aventine Readers,
First a bit of news: this month’s newsletter is half as long! We’re switching to a twice-a-month cadence that will feature our reported feature, “The Big Idea,” at the end of the month and the “Five Ways to Think About…” feature mid-month. We’ll keep the shorter features, “Advances That Matter” and “Long Reads,” in every issue.
This month’s feature looks at a surge in spending on so-called dual-use defense technologies that can be deployed for both military and commercial purposes. Russia’s invasion of Ukraine and tensions between China and the West have lent a new urgency to defense innovation, and because militaries are not known for their lightning quick adoption of new technologies, finding a civilian market is often a first step for startups needing early revenue.
Also in this issue: A new, non-CRISPR approach to gene editing, a billion-dollar copper mine discovered by artificial intelligence and vaccines that could protect against Alzheimer’s.
More in two weeks!
Danielle Mattoon
Executive Director, Aventine
Subscribe
Subscribe to our newsletter and be kept up to date on upcoming Aventine projects
Defense Tech Is Having a Renaissance. Will It Last?
Without technology developed for military use, you wouldn’t be reading this.
The internet, cellular communication and many modern electronic systems owe their existence to research and development in the defense sector. That’s not including jet engines, radar, GPS and drones, among many other contributions. Developed and incubated as technologies to support soldiers and deliver a military advantage, these innovations were then adapted for commercial uses in ways that have transformed modern life.
While the importance of such innovation, at least as expressed in dollars invested, lost some urgency in the past decade, a rekindled interest in technologies designed to perform on the battlefield — and beyond — may change that. Increased geopolitical instability — and the prospect of more to come — combined with the lure of defense department budgets, is inspiring a new, global wave of military-inspired innovation and investment. NATO recently announced the first recipients of grants from its huge $1.1 billion Innovation Fund. The European defense AI startup Helsing recently raised almost $500 million, and the U.S. drone maker Anduril is reportedly in talks to raise $1.5 billion. Many venture capital firms are hiring former military officials. Defense technology secured almost $35 billion of venture capital investment in 2023, according to PitchBook.
“It’s picking up steam,” said Kelly Chen, a partner at the NATO Innovation Fund. “We’re seeing a lot more folks in the ecosystem, whether it is founders or other venture capitalists.”
Aventine spoke with startup founders, investors and NATO employees about the rising interest in defense technologies. They described a blossoming ecosystem, excitement about a range of technologies that go beyond headline-grabbing killer drones and explained the commercial and military benefits of this new surge in innovation. They also warned of the potential for defense tech to become another bubble if investors and founders don’t behave responsibly.
Building for a new world order
It’s hard to overstate the role fear has played in the recent upswing in interest around military technology. Experts who spoke with Aventine agreed that geopolitical tensions and ongoing conflicts around the world have been a huge driver. The war between Russia and Ukraine topped the list of concerns, but “the whole world is sort of on fire,” said Jake Chapman, managing director at Marque Ventures, a VC firm focused on defense technology based in San Francisco, California. “Northern Africa is troublesome, it's where terrorism has generally moved. Afghanistan [is] once again controlled by the Taliban. They're making friends with the Chinese and the Russians. We have the potential for a war in the Taiwan Strait … All these things have come together at this one moment in time.”
Against that backdrop, the competition to develop and own the best technology is fiercer than ever. “[Tech] is the underlying substrate that our society functions on, to a greater extent than ever before,” said Daniel Carew, a partner at the VC firm Join Capital who focuses on defense. “So people are trying to control it.” At the same time, there is concern in some quarters that China is threatening the West's lead position in defense innovation, a situation that the U.S. and its allies are in a rush to remedy. “On every single technology evolution curve up to now, the United States was the clear leader,” said Brad Harrison, founder and managing partner at Scout Ventures, a VC firm specializing in defense tech based in Austin, Texas. “We are no longer the dominant force for this next wave of innovation.”
Finally, there is money to be made. The combined defense expenditure of NATO countries in 2024 was almost $1.5 trillion, up from $1.3 trillion in 2023 and $1.2 trillion in 2022. Those nations have agreed to commit 20 percent of their expenditure to new major equipment. The average across all NATO nations for spending on major equipment stood at approximately 30 percent in 2023, up from 13 percent in 2014. With huge budgets for buying new technologies in play, “everyone sees a hot new growth area,” said Carew.
One of the early boldface names to jump in is Andreessen Horowitz’s American Dynamism fund, launched in 2023, which is seeking to deploy $600 million of capital and identifies defense and national security as key areas of focus. “You've [now] got all of these firms that are saying, ‘Hey, we want to put private capital into these things,’” said Harrison. “Otherwise, we're going to be left behind because of our innovation [and] infrastructure going through the Pentagon … just doesn't move quick enough for the rate of today's innovation.”
Beyond killer drones
It’s not news that artificial intelligence has enabled a new kind of warfare, featuring the kinds of weaponized robots that science fiction has long predicted and society has long feared. Make no mistake: This technology is real, it’s here, it’s in use and will likely proliferate quickly. And it takes many forms — from the cheap, homegrown devices typified by remote-controlled drones carrying explosives used in Ukraine, to the long-range seek-and-destroy drones being built by Anduril Industries, the U.S. startup which may soon be valued at as much as $12.5 billion.
Yet the future of defense tech goes far beyond killer drones. So far this year, only four of the 20 biggest VC investments in aerospace and defense were related to drones, robots or AI, according to PitchBook data. The others are directed at areas like space technology, internet of things, advanced manufacturing and supply chain technologies. And of the first five investments made by the $1.1 billion NATO Innovation Fund, which were announced in June, just one is based around robotics, with the others supporting computer chips, advanced manufacturing and satellite launch systems.
And of course the use of AI in defense tech goes far beyond drones. Many other technologies are also important, whether they’re enabling robust and secure communications or finding new ways to build strong, lightweight vehicles.
“I think every government agency right now is trying to understand how artificial intelligence is going to disrupt how they have to think and do business,” said Chris O’Connor, a partner of the NATO Innovation Fund. “But I think we have to have the discipline as a fund to not just look into those hype cycles, but also continue to look across this broad tech base that we're going to be investing in and finding the best in class, who may not be talked about today, but four, five, six, seven years from now.”
Compared to modern civilian technology, where software is king, hardware — from improved armor to better batteries — has an outsize role to play in military technology: “You need hardware for [defense] applications, because you have to be able to reach out and touch somebody; software can't do that,” said Chapman.
Tech that does double duty
At the same time, many investors can’t or won’t invest in the kinds of killer drones that make headlines — either because they or the organizations that provide their capital have objections to funding the design and manufacture of weapons, or they are financially obligated not to make such investments. That helps explain why an approach known as “dual-use,” which has been around for decades, is of particular interest as part of the newfound interest in defense tech.
There are competing definitions around what exactly counts as dual-use, but at its most basic it’s the idea that a technology can have an application in both military and civilian settings, and can therefore be sold to both the defense and commercial sectors. The attraction of selling the same product to multiple customers is obvious, and governments on both sides of the Atlantic are increasingly relaxed about the relevant rules.
Still, the best way to approach dual-use product development remains unresolved. One camp of investors argues that the most successful approach is to target the defense market first. Several people Aventine spoke with cited companies such as SpaceX, Palantir and Anduril as examples of highly successful companies that have adopted defense-first or defense-exclusive business models. Nicholas Nelson, a partner at the defense-focused VC MD One, argues that selling to the military provides long-term predictability and contracts that are commonly renewed. Meanwhile another camp, which includes Carew, contends that procurement cycles within the defense sector are painfully long and complex, which means few startups can wait long enough for a contract before they run out of cash. This, they argue, means that most startups are better off targeting civilian applications first.
Others believe it’s a case-by-case decision. “It depends on the technology,” said NATO’s O’Connor. “I think there are certain technologies where it makes far more sense to find success and scale in commercial before pivoting into government defense use cases. I think there are other technologies that it makes sense to de-risk technically in some government defense applications.”
The reality can be more complex still, and illustrates how dual-use approaches can deliver innovation across industries.
The U.K.-based advanced manufacturing startup iCOMAT, for example, which is backed by the NATO Innovation Fund, is developing new ways to build complex carbon fiber components for aerospace, automotive and space applications. Some of its first customers are in the Formula 1 industry, which evaluates and purchases new technology very quickly — sometimes in a matter of days or weeks, according to its CEO, Evangelos Zympeloudis. But its long-term focus is building for the commercial aerospace industry in which it can take over a decade to decide on and invest in new technology. Somewhere in between, it expects to sell to defense contractors. “Priority number one is short sales cycles, so that we can generate revenue and sales,” said Zympeloudis.
Dual-use can be a necessary and effective business strategy, and its adoption can also produce broad positive impacts. Perhaps most notable is accelerated innovation. This is particularly the case when a defense application can fund a bold technological development that has commercial benefits but would likely not be developed for commercial uses alone. O’Connor points out, for example, that the building of a spaceport might be heavily supported by government defense funding, but could quickly be used for commercial applications to help democratize space travel.
Tien Pham, the chief scientist of NATO’s Defense Innovation Accelerator for the North Atlantic (DIANA), notes other advantages too. Products developed for defense applications must be built to higher specifications than their commercial counterparts, which means that they are typically more robust and resilient. He also argues that technologies developed with defense use cases in mind often seek to address extremely difficult problems, which typically prompts fresh innovation aimed at solving the central challenge.
Don’t call it a bubble. Yet.
Despite the potential impact, several of the sources Aventine spoke to raised concerns about how quickly interest in military tech has grown. In particular, they worried that investors who have recently hopped from one hot new sector to the next — such as crypto or NFTs — may now shovel money into military technology without fully understanding the complexities of the market.
“The challenge, if you do not have a sophisticated investor base, is that [they] risk inflating valuations where they shouldn't be,” said Nelson. “[That] could give [startups] some early capital, but as in any industry it can result in an inability to raise those next rounds, [or encourage them] down the wrong route of product market fit.”
That could lead to a frenzy in deal activity followed by many highly funded startups failing to gain traction. “The really smart, well capitalized people in the space — like Anduril in particular, they're very good at M&A — will end up scooping up a lot of technologies or companies for like, pennies on the dollar,'' said Chapman. “It's not that the technologies will get lost, necessarily. There’s just going to be a lot of sad investors.”
Ultimately, much will depend on how effectively the startups riding the current wave of investment fare as they use their funding. “There's a lot of buzz, on the government side, the industry side, the VC side,” said Chapman. “But is that going anywhere? That's still very much an open question.”
Listen To Our Podcast
Learn about the past, present and future of artificial intelligence on our latest podcast, Humans vs Machines with Gary Marcus.
Advances That Matter
Mingomba Mine exploration site in Chililabombwe, Copperbelt Province, Zambia, on June 12, 2024. Zinyange Auntony/The New York Times/Redux
AI has found a copper mine worth billions of dollars. A mile beneath the surface of the small town of Chililabombwe, Zambia, close to the country’s border with the Democratic Republic of the Congo, is a ticket to vast wealth: a rich vein of copper, enough to produce as much as 300,000 tons of the metal every year for decades. But perhaps the most fascinating part of this story, as reported by The New York Times, is that this lode was identified using artificial intelligence by a company based in Silicon Valley. The company, KoBold Metals — a high-tech prospecting company backed by hundreds of millions of dollars from the likes of Bill Gates, Jeff Bezos and the VC firm Andreessen Horowitz — uses its AI to interrogate huge quantities of data, from typewritten reports in Zambian archives to radar and magnetic readings captured by light aircraft flying over promising areas. KoBold plans to own a stake in the mining operations in Zambia, which it hopes will swing into production in the early 2030s, though the site has not yet been determined. Between now and then the project will likely face headwinds — not least from the Zambiian government, which believes it should be entitled to more than the 20 percent stake in the vein of copper that it currently holds.
A more powerful gene-editing technique. It’s little more than a decade since CRISPR redefined genetic engineering, but it may soon have a powerful new companion. A collaborative effort between researchers from the Arc Institute in Palo Alto, the University of California, Berkeley and the University of Tokyo sought to identify new approaches to gene editing that enable large sections of DNA to be inserted into the genome of cells. While CRISPR is incredibly effective, it is typically used only to make very small, specific edits to strands of DNA. And though other techniques exist to insert longer chains of DNA, they typically involve many complex steps. This new approach — which uses one strand of DNA to identify its target and another that is inserted into the genome — could be used to correct multiple faults in genes with just one edit, potentially making it much easier to correct some genetic diseases. The work has so far not been successfully extended into mammalian cells — something that the team is now working on.
Highly targeted pesticides to kill some of the toughest pests. Spray regular pesticides over a field and you kill harmless bugs as collateral damage — and, perversely, you don’t always wipe out some of the tougher insects, such as the Colorado potato beetle, which is immune to more than 50 pesticides. A new approach may solve both problems by using highly specific pesticides designed to target specific genes within some of the world’s most pernicious pests. Science reports on the growing field of RNA-based pesticides, including one designed by startup GreenLight, which specifically targets that potato beetle and was approved by the U.S. Environmental Protection Agency in January after a four-year review. Similar pesticides are also in development, at various stages of regulatory scrutiny, field trial or lab development. The pesticides rely on a mechanism called RNA interference, where the cells in an insect’s gut break down carefully designed double-stranded RNA. It then prevents important proteins inside the insect from being produced, leading to its death. While the GreenLight product is cost competitive with other top-end pesticides, there are issues: Some pests have digestive tracts so aggressive that they break down the RNA before it works; the approach can have off-target effects on genetically similar insects; and lab tests show that some pests can develop a tolerance to these RNA pesticides. So, much as with any pesticide, they must be used with care.
Magazine and Journal Articles Worthy of Your Time
The strategy behind one of the most successful labs in the world, from Nature
2,600 words, or about 13 minutes
The Laboratory of Molecular Biology in Cambridge, U.K., has a stellar reputation: It has produced 12 Nobel prizewinners, and four of them in the last 15 years. This Harvard Business Review-style analysis, based on 12 interviews with senior lab members, external scientists and 60 years’ worth of archival documents, tries to understand the lab’s success. “By integrating high-risk basic science with innovative technology, the LMB facilitates a knowledge feedback loop that helps the institute to identify promising questions and continuously push scientific boundaries,” the authors conclude, encouraging “research organizations, funding bodies and policymakers to consider adopting a similarly holistic and coherent approach to managing basic scientific research.” It’s a fascinating read for anyone interested in the scientific method or grappling with how to drive innovation.
There Will Be Blood, from Science
3,300 words, or about 17 minutes
You’re on a battlefield with a soldier who’s bleeding out. There are no blood supplies for miles, but a medic pulls out a pouch of freeze-dried powder, mixes it with saline and injects it through an IV straight into the patient's arm. Moments later, the soldier regains consciousness. That might sound fantastical, but it’s the promise of an experimental blood substitute called ErythroMer, a product made from recycled human hemoglobin that has $46 million in backing from the U.S. Defense Advanced Research Projects Agency. It helps fulfil only the function of red blood cells, not all of the functions provided by normal blood, but in most emergencies that is all that is required. The product, essentially tiny spheres of hemoglobin coated in a barrier layer of lipids, is shelf-stable and universal, so it can be administered to anyone regardless of blood type. And while it’s still only in the animal testing stage, many early tests show that it could be lifesaving in emergency settings.
New anti-aging vaccines promise to prevent diseases like Alzheimer's, from New Scientist
2,800 words, or about 14 minutes
Think of vaccines, and — especially since the COVID-19 pandemic — you probably think of shots designed to provide immunity from infectious diseases. But the basis of vaccines is simply to prime the immune system to fight off unwanted threats. That’s relatively straightforward for many infectious diseases, because it’s easy to fight off foreign bodies. But it’s also theoretically possible — just more difficult — to use similar approaches to prevent noninfectious diseases, such as cancer, Alzheimer’s, hypertension, osteoarthritis, or Parkinson’s. Or, in other words, to create vaccines against many of the diseases that more and more people now have to deal with because they are living longer. As this story explains, progress in developing vaccines for some of these age-related conditions — and actually some of the processes inside the body that are more directly associated with aging itself — are being developed faster than you might expect.