Q&A with Marlinspike
Last month, I gave a webinar for Marlinspike, a venture capital firm that invests in aerospace, UAVs, AI, and other dual-use technologies. The presentation covered the status of the war in Ukraine and emerging lessons from that conflict. Below is an edited and partial transcript of the Q&A, cross-posted on Marlinspike’s Substack.
On the Ground
How does Ukraine’s Kursk offensive play into the larger strategic picture?
There were three likely motives behind the incursion. The first—and Skyrsky’s stated objective—was to relieve pressure on the Donbas by forcing the Russians to divert resources. The second was to embarrass Putin and make the war politically untenable—part and parcel of the larger campaign against Russian infrastructure. The third was to gain a territorial bargaining chip that could be used at the negotiating table.
I suspect the specific location of the offensive was not random, but was chosen because air defenses around Sumy could be pushed forward without jeopardizing that city or drawing resources from elsewhere.
Different leaders within Ukraine no doubt had different ideas of what could be accomplished, and there was probably a bit of opportunism too—seeing what could be gained as the situation evolved. But the first objective was undoubtedly the most important.
And how has it worked?
The first objective has clearly failed. Russia scraped together reserves from elsewhere to contain the situation in Kursk, while continuing to advance in the Donbas.
The incursion has also had marginal political impact. Although it’s slightly embarrassing to Putin, a few border towns are not enough to alter his calculus—the Ukrainians would have had to capture something like the regional power plant or Kursk itself to have any real effect.
The third point is more interesting. After Ukrainian losses began rapidly mounting once the Russians moved sufficient forces to that front, they started to dig in and mine extensive areas. This will allow them either to hold the small area they now control, or exact a high cost from the Russians to retake it.
Overall, however, the operation looks like a decidedly net negative for Ukraine. And now that Russian forces have been committed to the area, there’s no saying they won’t continue on toward Sumy if they push the attackers out of Kursk.
What does that mean for the rest of the front? What is happening elsewhere?
The Russian salient jutting towards Pokrovsk has created two further Ukrainian salient on either side. The western one is supplied by rail lines running through Pokrovsk, which are close to being severed south of the city—I expect the Russians to focus more on surrounding and collapsing that pocket than on the city itself.
The other salient is supplied by rail lines running further east, from the major hub at Kramatorsk through Kostyantynivka. This is protected to the east by the fortress at Chasiv Yar, which is supplied by another line from Kramatorsk. That will make it a much harder fight for the Russians, so we can expect them to advance much slower than in the west.
In the meantime, I expect that the Ukrainians are cycling their better units out of Kursk to rest and refit for the defense of the Donbas.
Technology Lessons and Future Applications
Tell us about the nature of the fighting itself. There’s been a lot of attention on drones, but what else is shaping the dynamics of the war?
The fighting continues to be dominated by traditional factors such as artillery, mines, and raw manpower. Mines have arguably been the most important: they’re absolutely everywhere, slowing down ground movements enough to make artillery, drones, and other weapons much more effective. And both sides seem to lack adequate mine-clearing equipment.
There’s been a lot of attention to production bottlenecks at the high end, such as PGMs, and at the lower end, with shell shortages. But there’s also a vast middle of specialized equipment that is difficult to produce in sufficient quantity—mine-clearing vehicles are just one such example.
How will armies respond to this seeming defensive advantage?
For the specific problem of mines, there’s obviously no substitute for producing equipment in adequate numbers, but the real question is how mine-clearance will work in the larger schemes of maneuver. The challenges are 1.) to get through minefields while minimally exposing troops and then 2.) get under cover as quickly as possible.
The first offers a promising role for UGVs: vehicles that can roll out in advance of an attack to clear lanes. This allows forces to remain in assault positions until the last minute, then move quickly through the obstacle to the objective.
The second is much trickier. I’ve argued elsewhere that any mobility will require the ability to dig in quickly when not moving, even during short halts. The likely answer is to place excavation equipment with forward units, allowing them to construct cover upon the objective within hours, if not minutes. These positions can be expanded as needed, or occupied by follow-on forces as the attack proceeds.
More generally, attackers will need so many enablers—counter-UAS, EW, engineers—that these will outnumber combat troops even at the very tip of the spear.
You caution that drones are only one piece of the puzzle, but they will undoubtedly be an increasingly important one in the future. Can you tell us how you see this evolving?
The evolution of drones is inseparable from their countermeasures. As of now, EW has been very effective at jamming drones’ GPS and command guidance. Pretty soon I think we’ll see fully autonomous workarounds: drones equipped with inertial guidance to get to the right neighborhood, terrain-reading to get on target, then AI-enabled terminal guidance. Jamming is only one kind of EW, of course: the obvious response is decoys paired with signature spoofing.
Jamming will continue to serve other important counter-UAS roles, not least against spotter drones that feed data to other weapons systems. Not every battlefield will be wide open like the plains of Ukraine, where targets can be spotted from great distances at a slant. More constricted terrain will force drones to fly directly overhead, bringing them into jammer range.
Beyond that, there’s no neglecting hard-kill answer to drones: short-range guns, counter-UAS drone swarms, directed energy, etc.
What kind of impact will these changes have on military organization?
One thing that’s not talked about so much is how drones enable much greater command and control. In a lot of cases, drone saturation is so great that higher headquarters have better situational awareness than commanders on the ground. Assuming comms are robust enough, this will inevitably mean that control becomes much more directive.
This will partly come from the natural tendency of higher commanders to exert more control whenever they can, but the greater number of enablers in forward units will also make this a necessity. There are already so many moving parts even in small-scale assaults, that things need to be tightly synchronized; accounting for growing needs and inherent battlefield uncertainty, this will have to be orchestrated from the top.
All these changes will obviously require changes to the acquisition process. What will those look like?
The lessons learned from Ukraine are only one part of the picture, and even there, things are evolving too rapidly for the traditional acquisitions process. So many new weapons are the product of soldiers and small shops tinkering, which are then ramped up to larger-scale production. It’s very hard to develop this sort of responsive industry outside of wartime. Soldiers themselves don’t know what they will need, and peacetime rules make it very difficult to allocate money appropriately. Unfortunately, I don’t see any country really solving this problem until its servicemen begin dying.
There’s also the problem that the baseline for “readiness” is so much broader than ever before. Where do you focus your limited resources? On drone warfare? On enabling ground assaults? On missiles and missile defense? On aviation, or naval assets? All that depends on the theater, the adversary, and the stakes. Most countries, including the United States, can only really prepare for one high-end fight while doing their best to adjust on the fly to whatever else comes their way.
Here, I think, we will start to see a certain type of military entrepreneur emerge, one that can provide critical capabilities that are too narrow—or too expensive—for countries to acquire in peacetime. And I expect this to happen right at that middle tier of specialized needs: engineering equipment, communications, etc. And not just the hardware itself, but operators trained in their use—contracted services being brought closer to the front line.
We’ve already seen something like this with Starlink, where Elon Musk stepped in to provide crucial comms for Ukraine. Another example might be the civil wars in Libya and Ethiopia, where outside powers were able to tilt the balance by furnishing UAVs along with the technicians and trainers to operate them. Depending on how volatile the next few decades are, entrepreneurs may be willing to risk their capital maintaining niche capabilities that will always be in demand somewhere.
As the range of capabilities expands, this will likely affect the acquisition process as well. Instead of seeking out government solicitations based on perceived needs, entrepreneurs will be more active in proposing new capabilities—in many cases, forcing serious rethinking of existing tactics and doctrine.
Anduril is hard at work on the Drones on the American side.
As for money, in America we can have 6 years to contract PPBOE and a vast Theater sized force of bureaucrats who get the 27 signatures needed from Pentagon action officer to SECDEF, and let’s not forget Congress, XOR (Exclusive OR) we can have something to fight with, a problem delayed so long that now we don’t it seems have people to maneuver and support the materials remaining BUT XOR we cannot have both.
Don’t think of the choice as binary when it has been proven Boolean. If one requires Binary,
You can have a Pentagon or you can have a military but not both.
As to the ground game; even as ground pounder and vehicle rider I propose people may wish to reconsider that the purpose of the military is either to support the infantry, or perhaps something else?
As in Moxie? Or no relation?