Latent Space: The AI Engineer Podcast
Latent Space: The AI Engineer Podcast

The Autonomous Drone Tech Stack & Economics of Drones — Yaroslav Azhnyuk, The Fourth Law & Guest Host Noah Smith, Noahpinion

The future of war has been evolving before our eyes in Ukraine, yet the west still plans to fight the last war. In this special episode, guest host Noah Smith (@noahpinion) and Brandon Anderson sit down with Yaroslav Azhnyuk (@YaroslavAzhnyuk), a serial tech founder who went from building PetCube to

Featured Speakers

Latent.Space Host

Topics Discussed

Episode Summary

Executive Summary: The episode centers on Ukraine’s drone war experience and how it is reshaping modern warfare, manufacturing, and defense policy. Yaroslav Azhnyuk argues drones are now the decisive battlefield technology, that autonomy will rapidly expand, and that Western countries must reindustrialize, build resilient supply chains, and take China’s drone manufacturing threat seriously. The conversation blends wartime testimony, technical explanation, and strategic warnings.

Main Topics: Ukraine’s war-driven pivot from consumer tech to defense tech (Priority: 5/5): Azhnyuk recounts being in Kyiv on the eve of Russia’s invasion, evacuating, and then shifting from pet cameras to drone systems supporting Ukraine’s defense. He describes outreach to policymakers and startup-building as part of a personal and national response to invasion. Drones as the dominant force in modern warfare (Priority: 5/5): The speakers argue FPV drones have become the ‘god of war’ on the Ukrainian front, surpassing artillery in casualty generation and making tanks and traditional formations far more vulnerable. Levels of autonomy and battlefield AI (Priority: 5/5): Azhnyuk lays out five levels of drone autonomy, from terminal guidance to autonomous takeoff/landing, and argues that autonomy increases accuracy, mission success, range of use, and scalability. He also sketches a broader ‘eight dimensions’ framework for battlefield autonomy. Countermeasures, EW, and the economics of defense (Priority: 4/5): The discussion covers fiber optics, electronic warfare, radio horizon limits, shotguns, nets, interceptor drones, and lasers. Azhnyuk emphasizes that drone-counterdrone defense is ultimately an economic contest over cost per effect and mass deployment. China, supply chains, and industrial scale (Priority: 5/5): A major warning theme is that China’s manufacturing base could overwhelm Western drone production if conflict escalated. The conversation highlights rare earths, motors, thermal cameras, optics, and manufacturing capacity as critical bottlenecks for the West. Moral justification and the ethics of building weapons (Priority: 4/5): Azhnyuk defends weapon-building as morally necessary for a nation under attack, arguing it is immoral not to fight back. He also questions how responsibility should be assigned when makers create technologies that can be used to kill. Future battlefield: humanoid robots, AI generals, and mass autonomy (Priority: 3/5): The episode speculates about full autonomy, AI-driven target selection, humanoid battlefield robots, and the possibility that future wars will be fought through highly automated systems at scale.

Key Arguments: Ukraine’s experience shows that modern war is now a software-and-manufacturing contest, not just a contest of legacy platforms. FPV drones have become the decisive tactical weapon because they are cheap, precise, scalable, and adaptable across missions. Even partial autonomy, such as terminal guidance, can dramatically improve mission success and expand the effective kill zone. Fully autonomous drones could multiply effectiveness by reducing training needs, bypassing EW and radio-horizon limits, and enabling one operator to control many assets. Defense against drones must be mass-produced and economically viable; expensive point defenses cannot scale against cheap attackers. The West is behind China in several layers: autonomy, manufacturing scale, components, and materials/refining. A future war with a drone-heavy China would be a manufacturing war; whoever can produce and deploy more mass at lower cost may win. Western countries need urgent policy changes: procurement reform, defense industrial growth, supply-chain diversification, and tighter Ukraine-West knowledge exchange. In war, moral responsibility shifts toward protecting your own people; building defensive weapons can be ethically justified when a nation is under attack. Human infantry will not disappear soon, but drones and autonomy are steadily reducing their dominance and changing their role on the battlefield.

Data Points: Ukraine FPV drones produced in 2024: 4 million - Azhnyuk cites Ukraine’s annual FPV drone output as evidence of wartime industrial scaling. Public Ukrainian FPV target for this year: 7 million - Mentioned as the publicly voiced production goal for the current year. Estimated Russian FPV output relative to Ukraine: ~20% less than Ukraine - Azhnyuk says Russia is producing somewhat fewer FPV drones than Ukraine. FPV share of front-line casualties: 70–80% - He states FPV drones account for the majority of casualties on the front line. FPV mission success increase with terminal guidance: 20% to 71% - A brigade using Azhnyuk’s first-level autonomy reportedly improved mission success from 20% to 71%. Kill zone expansion from terminal guidance: 3 km to 10 km - The same autonomy upgrade expanded the effective kill zone. Shahd interceptor speed: 326 km/h - Azhnyuk gives the speed of one of their interceptor drones. Shahd cruise speed: ~220 km/h - He estimates the target drone’s cruise speed for comparison. Artillery shell price: ~$4,000 per 155 mm shell - Used as a cost comparison against drones. FPV drone price: ~$400–$500 - Used repeatedly to highlight drone affordability versus artillery and tanks. Reinmetall tank cost: ~$5 million or more - Used to emphasize the asymmetry between drones and armored vehicles. Raytheon laser system cost: ~$3 million - Azhnyuk cites a laser counterdrone system he saw at a conference. Laser engagement time: ~3 seconds per drone - Used to question whether lasers can handle mass drone attacks economically. Optic fiber price increase: $4/km to $32/km - He says fiber used in drones rose sharply because of AI data-center demand. Ukrainian drone autonomy levels: 5 levels - Terminal guidance, bombing, autonomous target detection, autonomous navigation, autonomous takeoff/landing. Drone dominance program range: 10 km - Azhnyuk contrasts this U.S. competition benchmark with Ukraine’s longer-range expectations. Typical Ukrainian FPV operating range: 20–25 km minimum; 30–40 km common - He says Ukraine demands much longer range for FPV effectiveness. Long-range drone strike distance: up to 2,000 km - Referenced in relation to Shahed-type drones and strategic reach. Glide bomb range: up to 80 km - Used when discussing Russian air-delivered munitions.

Pivotal Quotes: "At some point in my life, I went from making cameras that fling treats to pets to cameras that fling explosives to the occupiers." — Yarslav Ajnuk: He summarizes his transition from consumer tech entrepreneur to defense-tech founder. "It is immoral not to fight back." — Yarslav Ajnuk: He explains why building weapons for Ukraine is morally justified in the face of invasion and atrocities. "You want peace, be prepared to war." — Yarslav Ajnuk: His closing takeaway on deterrence and defense preparedness.

Implications: The episode suggests the future of warfare will be defined by mass drone production, autonomy, and resilient supply chains. Listeners should expect defense-tech investment, procurement reform, and industrial policy to matter far more in national security.

🔓 Sign Up for Unlimited Episode Search

About Latent Space: The AI Engineer Podcast

The podcast by and for AI Engineers! In 2025, over 10 million readers and listeners came to Latent Space to hear about news, papers and interviews in Software 3.0. We cover Foundation Models changing every domain in Code Generation, Multimodality, AI Agents, GPU Infra and more, directly from the founders, builders, and thinkers involved in pushing the cutting edge. Striving to give you both the definitive take on the Current Thing down to the first introduction to the tech you'll be using in the next 3 months! We break news and exclusive interviews from OpenAI, Anthropic, Gemini, Meta (Soumith Chintala), Sierra (Bret Taylor), tiny (George Hotz), Databricks/MosaicML (Jon Frankle), Modular (Chris Lattner), Answer.ai (Jeremy Howard), et al. Full show notes always on https://latent.space

View all episodes from Latent Space: The AI Engineer Podcast