Welcome to this month’s first exclusive paid content. It’s another thorough, ~5,400 word article that seeks to answer the question I’ve been asked by some readers over the past few months: is there a possibility that Ukraine can still achieve victory in this war? I look at the question through the technological lens, as whatever real chance at victory Ukraine may have remains tied into its one and only arguable advantage: the Western-supplied drone and AI tech.
I leave the first ~2,000 words open to free subscribers: if you like what you read, please consider signing up for a paid subscription today, so you don’t miss my most hard-hitting exclusive content in the future.
Recently we’ve talked a lot about Ukraine’s ongoing collapse—all the myriad dysfunctions and vulnerabilities of the Kiev regime which are sure to lead to its demise in the next year or so.
But hidden beneath this sometimes overly-rosy picture are the many faults of Russia’s own military campaign, as well as the inroads being made by the vast NATO apparatus bringing up the rear of Kiev’s operations.
In this report, we’ll explore some of those potential threats to Russia’s military predominance.
Project White Stork and AI Supremacy
The first and most significant concept to understand is that the West is using the Ukraine war as an unprecedented testbed for the launch of a new era of AI warfare.
![](https://substackcdn.com/image/fetch/w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F8c9352be-610d-41a3-8b4b-5c5e9f357c5c_1246x223.png)
What was left unsaid in the more boilerplate offering above was hinted elsewhere by ex-Google CEO Eric Schmidt. Him and his cohorts believe that in a year’s time, Ukraine could achieve a technological supremacy of AI warfare that would effectively deadlock the war. Note I said deadlock not turn around; nothing Ukraine can do will be able to turn the tide so much as to allow Ukraine to retake lost territories or make Russia surrender. Offensively speaking, Ukraine is as good as done. But the question is, can Russia successfully prosecute advances indefinitely to the point where it achieves all of its battlefield objectives? Or will it get mired down in a technological bog that simply precludes all possibility of breakthrough, a la the stalemates of World War One?
Unlike his previous data-management AI system, Project Maven, Schmidt’s latest Project White Stork is specifically geared toward creating AI attack drones—ones which can operate autonomously even in a EW jamming heavy environment. What Schmidt, DARPA, the CIA, and Ukraine are doing is essentially trying to create fully autonomous drone swarm fleets that would turn the battlefield into a nightmarish no-go zone for any “topside” troops. Sure, the rear lines would indefinitely be safe because Russian forces have not only already been adept at building fortified underground structures, but have—by some reports I’ve read—even begun preemptively adapting to this future threat by taking more and more C2 HQs underground. However, the frontline is a different story. There is simply not enough time to build adequate shelter from mass drone swarms when you advance to capture a new slice of enemy territory.
A recent Forbes piece on drones gives some insight into what’s to come:
The underlying technology trends supporting this growth are less about aeronautics or material science and more a product of AI. Visual images are only part of the story, with thermal detection, geolocation tagging, 3D imaging, and volumetric calculations all adding depth to the data retrievable safely and cheaply with aerial drones. Each of these data layers has diagnostic and planning value for operations like yard management, plant maintenance, infrastructure design, and demand planning.
It goes on to explain how AI tests at Los Alamos National Laboratory have created AI applications that can infer large amounts of data from ‘sparse’ information, recognizing patterns and extrapolating them. For instance, semiconductor fabs estimating the production volumes of their competitors by simply analyzing their parking lots and attendant employee shift patterns.
I’ve long covered how such AI techniques have already been applied to automating NATO’s satellite ISR volumes of Russian strategic targets. For instance, the long-running Project Maven, a precursor to White Stork. The first half of this year has seen a spate of painful losses of rear components like prestige air defense and rare radar systems, all thanks to AI being used to sweep thousands of satellite images to detect and identify hidden systems without the need for large manpower expenditures.
For newer readers who missed my long-ago exposé, I urge you to check out this article for a much more detailed look:
Recently it was revealed that an advanced Israeli AI system called ‘Lavender’ was already being utilized, along with its sibling called ‘The Gospel’, which sift through data bottlenecks to determine both human and structural targets to be hit.
There are various such startups happening from a multitude of companies on the Ukrainian side. For instance:
French startup Alta Arès is testing an artificial intelligence system with facial recognition of personnel in Ukraine. There are plans to create a complex that can be installed on DJI's" combined-arms " Mavic 3 Pro drones, Mavic 3 Thermal and Matrice 300 RTK and similar ones in order to identify the personnel of the opposite side in automatic mode, analyze and transmit to the control center. It is planned to combine the system with Delta, the main Ukrainian application for combining intelligence data on the battlefield.
For illustrative purposes, recent unspecified tests:
Here’s a recent report from a Russian source on some of the secret developments on the Ukrainian side in this vein:
In 2023, the enemy tested a new global digital system for analyzing targets and managing the battlefield - ASMV-in some parts of the front. The system is able to track all combat operations along the front line in real time and is designed to help the command in redistributing forces for combat work along the LBS.
It is reported that this system was fully operational in 2024 and is being used by the Ukrainian Armed Forces in the Kharkiv direction, which was one of the reasons for the disruption of our offensive potential.
The system is focused on the use of NATO and will soon be deployed along the entire front line, which may increase the stability of the Armed Forces of Ukraine in defense at times.
With the advent of the automated control system, all command and control of troops is taken out of the competence of individual commanders, and each of them knows every second not only the combat situation at home and neighbors, but also receives precise instructions on where to shoot, what and what result to achieve. Thus, the subjectivity of the opinion of lower commanders is eliminated from the bottom-up chain, and the commander of the front sector receives more truthful and fast information.
The only way to disrupt the operation of the automated control system is to destroy the enemy's communications equipment - "Starlinks", the cellular network.
The author cites the new system as being responsible for Russia’s ‘stagnant’ Kharkov campaign.
And from another even longer report by Russian military analyst Alexey Zhivov, we have the following—read the frightening text in full to get a detailed first-look at the DARPA-led initiative:
A new type of war: what did we face during the Kharkov operation?
I wrote before, and confirmed from various sources, that the “North” group, after a spectacular “first round,” got bogged down in the subsequent “clinch” in Volchansk and Liptsy.
There are two types of reasons that led to this. Ordinary and extraordinary.
We include the usual reasons : the small number of the "North" group, a limited range of tasks, resources and reserves, powerful opposition from the enemy, who pulled reserves of all types of troops into the operation zone and began to use guided Western bombs.
But no one has really understood the extraordinary reasons yet . Basically, they boil down to the fact that the enemy has superiority in low air (attack and reconnaissance drones), while we lack effective trench- and army-type electronic warfare systems.
But it's not that.
Last year, NATO, together with the Ukrainian Armed Forces , began testing a full-fledged system of decentralized battle control based on special multi-layer military programs with artificial intelligence. Let's call it "military Google." DARPA is putting the lion's share of money into multi-domain warfare .
The essence of this military-analytical complex is that it is capable of “seeing the entire front” and processing colossal volumes of combat data per second. This fundamentally changes the picture of a battle with an army that has a classic (and not very reliable) military connection.
Wings, FPV Mavics and guided bombs make up a new type of attack and reconnaissance system. Many storm wounded did not see a single enemy soldier. The captured FPV operator had 50 drones, two Starlinks, and antennas worth 10 million in his landing. It was some grandfather who had been a nightmare for the month of Voznesenovka.
This is a completely different level of battlefield organization.
All units at the front with satellite communications, all artillery and mortar crews, all UAV operators, all aviation, armored vehicles and MLRS are all simultaneously connected to “military Google”. All reconnaissance means from tactical drones to American satellites are connected to it.
All movements, rotation routes, the number and composition of our military units, ammunition depots, crowds of people (hello formations), all this is immediately known to the system, and therefore to every serviceman of the Armed Forces of Ukraine.
"Military Google" takes away most of the secondary tasks of military units of the Armed Forces of Ukraine for digital outsourcing, leaving them only fire control.
The speed of transmission of a control signal in such a system is almost instantaneous, the choice of weapon is always optimal , and the effectiveness of the destruction is monitored by AI, the battle is controlled from a remote information center, which has awareness many times greater than any headquarters of any of our groupings.
Before us is a digital Goliath. And we are analog David, who must use his sling very carefully.
Groups of FPV operators are hiding in dense forests. They are guarded by fortifications with heavy machine guns in the trenches. Artillery is flying into the tanks from all barrels, the air has been fully and constantly scouted, temporary ammunition supply warehouses are the first targets after the boxes, the equipment used for destruction is high-tech, tested and efficient equipment. A delivery distance of five kilometers is an insurmountable distance. In principle, nothing else is needed to disrupt logistics.
Such reports, specifically about the Volchansk direction, have been corroborated by several other well-known, albeit not necessarily totally credible, sources like Romanov_92 and WarGonzo. In Romanov’s case, he made big waves days ago with a long cynical video, after having traveled to the Kharkov frontline, which paints a dismal picture on the Russian side.
The auto-translation here is unfortunately exceptionally wonky, but it may at least give some inkling:
Though Romanov is known for his pessimistic slants, aspects of his report should still be noted, given the other concordant sources.
WarGonzo too recently returned from the same front and painted a picture of a changing paradigm of war, thanks to total FPV supremacy. Use auto-translation to read his enlightening article: https://dzen.ru/a/ZoAa9RojxTmqmUx2
An excerpt:
A serious competition (I'd even say very serious) to ground control today is air control, not through traditional aviation, but through the new fpv-reality. That is, now the degree of territory control, at least according to internal feelings, is no longer determined not by the depth to which your assault teams were able to go, but by how far (deep) your kamikaze drones can fly. This limit of control used to be set by artillery (this was a past, sort of art-reality). Where the artillery could reach was the reference point. But the fact is that in effectiveness across a range of parameters (accuracy, ability to objectively monitor, speed of deployment, cost after all) art-reality is already far behind fpv-reality.
—
Keep reading with a 7-day free trial
Subscribe to Simplicius's Garden of Knowledge to keep reading this post and get 7 days of free access to the full post archives.