Eric Schmidt on the war in Ukraine, AI and how Russia is innovating

AI summary

Overview Richard Engel interviews Eric Schmidt, former Google CEO and former chair of the U.S. National Security Commission on Artificial Intelligence, about Ukraine’s drone war and the accelerating use of automation in warfare. Schmidt’s central argument is that the war has reached an inflection point: Russia is adapting faster, scaling new drone and missile capabilities, and Ukraine needs urgent money and technological help to keep pace. Schmidt presents Ukraine not only as a battlefield but as the place where a major transformation in warfare is becoming visible. He argues that the decisive factor is no longer simply bravery, manpower, or individual weapons, but the speed at which each side can innovate, counter-innovate, and scale those adaptations across a large war. - Relevant source moment: [00:08:23] Schmidt says the key is “innovation to scale” and that whoever plays that game fastest will win. Main argument Schmidt’s reasoning begins with the claim that drones have made much of the front line almost immobile. Soldiers who leave cover can be detected and attacked within minutes, creating what Engel calls a large “kill zone” along the border and front line. From there, Schmidt argues that Russia has moved from being a slower adapter to a more dangerous innovator. He points to jet-powered Shahed drones, mesh-network communications, expanding missile production, glide bombs, possible mobilization, and a Starlink-like Russian communications system as converging developments that could make the war much worse. The underlying assumption is that battlefield advantage now depends on rapid technical adaptation at scale. If Ukraine receives enough Western funding, engineering support, and air-defense help, Schmidt believes it can hold; if not, Russia’s growing capacity could overwhelm current defenses. - Relevant source moment: [00:34:04] Schmidt says Ukraine used to be the small, fast innovator while Russia was big and slow, but “that’s changed.” Evidence and examples Schmidt’s strongest example is Russia’s modification of Iranian-designed Shahed drones with small jet engines.

He says these engines can cost around $5,000, are small enough to hold by hand, and allow drones to outrun existing interceptors, forcing Ukraine to develop faster anti-drone systems. He also describes a strike on the SBU, Ukraine’s security service, in Kyiv. According to Schmidt, the drone appears to have flown a long route and may have been guided through a Russian drone-to-drone mesh network, suggesting a more advanced command-and-control capability than many observers appreciate. Engel adds field evidence from Kharkiv and the eastern front, describing a “kill zone” of roughly 23,000 square miles, comparable to Croatia or West Virginia. He also says Ukrainian forces are using basic defenses such as fishing line and nets to catch drones, showing the gap between low-tech improvisation and rapidly advancing aerial threats. - Relevant source moment: [00:05:32] Schmidt describes Russia adding jet engines to Shahed drones, making them faster than current interceptors. Distinctive insights One distinctive insight is Schmidt’s distinction between automation and artificial intelligence. He argues that much of what is being called “AI warfare” is actually computer vision and automated targeting: systems recognizing a person, tank, launcher, or drone and then guiding a weapon toward it. A second overlooked idea is that the war is layered by cost and range. Cheap first-person-view drones act like battlefield “bullets,” while glide bombs, Shaheds, and ballistic missiles occupy increasingly expensive and destructive layers used for different tactical and psychological purposes. A third insight is that AI may first matter less as a battlefield “mind” and more as an accelerator of the adaptation cycle. Schmidt says this war already evolves in months, not years; AI could make targeting, swarm control, design, and countermeasure development even faster. - Relevant source moment: [00:22:31] Schmidt says current systems are mostly automation, not true AI learning, though he expects learning systems to arrive soon.

Predictions and conditions Schmidt predicts the situation may worsen because Russia is increasing or preparing several capabilities at once: ballistic missile use, jet-powered drones, KAB glide bombs, manpower mobilization, and satellite-like communications through a system he calls Rusvet. He presents these as compounding pressures rather than separate threats. He expects automated targeting to become common on both sides, especially in drone-versus-drone combat. He also predicts that computers will eventually manage swarms of drones because humans cannot control a thousand drones at once. The conditions that could weaken these forecasts include effective new Ukrainian interceptors, improved air defense, disruption of Russian production or communications, stronger Western technical support, or diplomatic developments that reduce the intensity of the war. Schmidt also marks some figures as uncertain, including claims about Russian stockpiles and Patriot availability. - Relevant source moment: [00:23:34] Schmidt says humans cannot run a swarm of 1,000 drones, but a computer can. Practical implications The clearest practical implication is that Ukraine needs rapid external support, not only in money but in technical expertise. Schmidt specifically emphasizes the need to counter the next waves of Russian drones, missiles, glide bombs, and communications systems. Another implication is that Western militaries should be studying Ukraine’s battlefield innovations more closely. Schmidt argues that much of this knowledge is not being transferred to the West, even though Russia is a potential future adversary of NATO countries. A third implication is that policymakers and engineers need to clarify rules for automated weapons. Schmidt says the United States and NATO are likely to allow automation under human supervision, while he doubts Russia and China will follow the same limits.

- Relevant source moment: [00:14:04] Schmidt says Ukraine has courage and capability but needs Western money and technological help quickly. Caveats and open questions Schmidt repeatedly avoids confirming operational details about systems linked to him or Ukraine’s drone interceptions. Several numbers he cites are explicitly framed as public claims or allegations rather than verified facts, including the figure of 80,000 Russian drones before December and the reported number of available Patriot systems. There is also uncertainty about how much true AI learning is already present on the battlefield. Schmidt says he has not yet seen full learning systems in combat, though he expects them soon and notes reports that Russia may be using Chinese open-source models. A major open question is whether AI-enabled warfare will deter conflict, as nuclear weapons have largely done since World War Two, or instead encourage risk-taking because each side believes it has a temporary advantage. Schmidt calls deterrence his hope, not a demonstrated outcome. Key takeaways Schmidt’s central claim is that the Ukraine war is becoming a contest of rapid, scalable technological adaptation. Drones, glide bombs, missiles, automated targeting, mesh networks, and eventually learning systems are changing both the front line and the strategic balance. The evidence he emphasizes is specific: jet-powered Shaheds, drone communication networks, layered weapons costs, KAB glide bombs, ballistic missile pressure, and the emergence of automated target recognition. The practical lesson he draws is urgent: Ukraine needs fast technological support, and Western countries need to understand this battlefield before the lessons arrive too late.