The Automation of War Drags On

Peter Zeihan 4 min

Processing data is not the same as thinking. Autonomous weapon systems are simply not anywhere close.

Chapters

  1. 0:00

    The Limits of AI Processing

    Processing data is different from thinking; current AI lacks the power to think independently.

  2. 1:42

    Sensor and Power Constraints

    Robots fail due to poor sensors, unstable batteries, and inability to manipulate environments.

  3. 2:29

    Learning Capabilities of Robots

    Humanoid robots cannot learn from failure; they require thousands of attempts to master simple tasks.

  4. 3:05

    Chip Manufacturing Fragility

    Advanced chips rely on fragile global supply chains that are prone to single-point failures. Current chips cannot withstand vibration, temperature changes, or dust found in military environments.

Key points

  • 0:00

    Processing data is not the same as thinking.

  • 0:00

    Thinking requires a whole server stack of chips.

  • 0:00

    Sensors are not nearly as good as the processing department.

  • 1:42

    Lithium batteries cannot maintain charge for tracked vehicles.

  • 2:29

    Robots cannot learn from repeated failures.

  • 2:29

    It took 11,400 attempts for a robot to do a backflip.

  • 2:29

    Another 13,000 attempts were needed to repeat the action.

  • 2:29

    Autonomous weapon systems are not anywhere close to reality.

  • 3:05

    Current chips require 100,000 manufacturing supply chain steps.

  • 3:05

    Over half of these components are single-point failures.

Related episodes

Map of the places in “Can AI-Driven Robots Save Japan?” 4:44

Can AI-Driven Robots Save Japan?

Peter Zeihan

Japan must develop autonomous robots because it cannot import labor due to racism and faces a massive workforce shortage. The project looks pretty promising as it merges long-term Japanese automation efforts with new frontier technology.
4:08

The Revolution in Military Affairs: Artificial Intelligence

Peter Zeihan

AI in war will not happen as expected because the semiconductor supply chain cannot survive de-globalization. The problem will become obvious in the 2050s.
5:21

George Jetson Would Be Disappointed with Autonomous Vehicles

Peter Zeihan

George Jetson would be disappointed with autonomous vehicles because the necessary infrastructure and chips are disappearing. Individual person vehicles will not be autopiloted or electric for much longer, though truck convoys may succeed.
15:36

Why Modern Drones Scare Me FAR More Than Nukes Do

Paul J. Warburg

Western wishful thinking and failure to help Ukraine adequately allowed this terrifying technology to become a reality. Humanity must embrace this terrifying reality and take steps aligned with it rather than relying on fear or wishful thinking.
Map of the places in “My Two Cents on the AI Debate” 3:58

My Two Cents on the AI Debate

Peter Zeihan

The current debate over AI safety is actually a positive sign of open conversation. The current debate does not need to bear fruit until the 2040s when supply chains are rebuilt.

+1

AI-generated summary · model qwen3.5:9b-12k · prompt version whole-3.window-7.combine-10.s-f4be8e · video id ucjOgpSrER0 · Watch on YouTube