AI Library
The Double Structure of Digital Sovereignty
Europe’s Departure from Palantir and the Chains of American Big Tech
Kim Kyung-jin, Attorney at Law
This is a record of 2026, when European intelligence agencies and defense ministries began removing analytics tools from America’s Palantir. It covers the replacement decisions made by France’s General Directorate for Internal Security (DGSI), Germany’s Federal Office for the Protection of the Constitution (BfV), and the Netherlands Ministry of Defense; the incident in which US export controls severed an ally’s ac…
New English Edition
Artificial Intelligence in Horticulture
Kim Kyung-jin, Attorney at Law
Across five chapters and ten sections, this book examines computer vision for crop diagnosis, harvesting robots and autonomous field systems, smart greenhouses and digital twins, precision irrigation and supply-chain quality control, high-throughput phenotyping, and predictive breeding.
New English Edition
Artificial Intelligence in Food Crop Agriculture
Kim Kyung-jin, Attorney at Law
Across six chapters and eighteen sections, the book examines digital agricultural infrastructure, remote sensing, crop diagnosis, yield forecasting, precision irrigation, genomics, molecular breeding, agricultural robotics, climate-smart agriculture, and global food security.
New English Edition
The Future of Forestry and Agroforestry
Kim Kyung-jin, Attorney at Law
Driven by Artificial Intelligence and Digital Innovation
Across five chapters and fifteen sections, the book follows satellites, drones, LiDAR, digital twins, forest-specific language models, wildfire and pest forecasting, forestry robotics, agroforestry, timber traceability, and forest carbon markets.
New English Edition
Smart Livestock Farming: AI Enters the Barn
Kim Kyung-jin, Attorney at Law
Sensors listen, cameras watch, and artificial intelligence helps farmers decide.
Across five chapters and fifteen sections, the book follows precision livestock farming from animal health and reproduction to robotic milking, virtual fencing, digital twins, methane reduction, welfare, and data ownership.
Table of Contents
Han Dong-hoon, Busan Buk-gu Gap: A Record of the 100 Days Before and After the Election (Mar. 26-Jul. 3, 2026)
Kim Kyung-jin
Table of Contents and 13 sections
From March 26 to July 3, 2026, this record follows the spring after expulsion, the Busan Buk-gu Gap by-election, victory as an independent, and the first bill submitted in the National Assembly.
Table of Contents
Artificial Intelligence and Medicine
Kim Kyung-jin, Attorney at Law
AI in clinical care, hospitals, education, and research
AI in medical imaging, risk prediction, treatment planning, hospital operations, education, and research, with patient safety, privacy, and accountability.
[AI Library] Chapter 11. Tens of Millions of Virtual Deaths
Chapter 11. Tens of Millions of Virtual Deaths
Tens of millions of virtual deaths at an altitude of 25,000 feet. An F-16 controlled by AI suddenly dives in the virtual sky of the simulator. I turned towards the enemy plane, but lost speed too quickly. The aircraft stalls and plummets straight to the ground. The word “Shooted Down” appears on the screen, and the simulation is reset. This AI just died. But a second later, it appears again in the same sky. And then it charges towards the enemy plane again. This time from a slightly different angle and at a slightly different timing. This process repeats millions of times a day.
It takes thousands of flight hours for a human pilot to become a veteran. After graduating from the Air Force Academy, completing basic flight training, receiving fighter jet type conversion training, and being assigned to an actual combat unit and working for several years, you can finally do your part. And the pilot can fly for at most 2,000 to 3,000 hours in his lifetime. This is because there are physical limitations, working period limitations, and above all, life limitations. Once you crash, that's it. But AI has no such limitations. In a virtual space, AI can die at any time and be reborn at any time.
Heron Systems' AI, which won the AlphaDogfight competition hosted by the U.S. Defense Advanced Research Projects Agency (DARPA) in August 2020, has fought approximately 4 billion virtual battles as of the time of the competition. This equates to approximately 31 years of flying experience for a human pilot. Ben Bell, engineer at Heron Systems, said: “Our AI has been trained to fight against 102 different agents, making it strong enough to not break down no matter what it encounters.” The principle of reinforcement learning is surprisingly simple. It's like a baby learning to walk.
No parent teaches a baby to “bend your knees at this angle and shift your center of gravity forward.” The baby just stands up and falls, and then stands up and falls again hundreds of times. Then, at some point, they take a step, and when their parents applaud, they realize, “Ah, this is the right thing to do.” It hurts when you fall, and you get praised when you walk. The same goes for reinforcement learning in AI.
When first thrown into a virtual world, the AI doesn't even know what an airplane is. You have no idea how to move the control stick or when to fire the missile. I just do random things. Turn the control stick to the left, pull the throttle, and press the gun button. Most of them crash or are shot down by the enemy. Whenever that happens, the system gives the AI a negative score. Conversely, if you catch the enemy's tail, succeed in strafing the enemy, or shoot down the enemy, you will receive plus points. AI has only one goal: to increase the sum of these scores as much as possible. The key here is the design of the reward function.
If you simply say "shoot down the enemy," the AI may learn the kamikaze tactic of rushing around like crazy and then retreating. So engineers have to create very sophisticated compensation systems. You must shoot down the enemy but survive yourself, get extra points if you catch the enemy's tail (6 o'clock position), lose points if you waste energy with excessive maneuvers, and get a big penalty if you enter a civilian area. Lockheed Martin's Alpha Dog Fight team designed this compensation system with advice from a retired F-16 pilot. It's a pilot's decades of experience translated into formulas and weights.
In the Alpha Dog Fight finals, Heron Systems' AI defeated the human pilot "Banger" 5-0. There was one shocking scene. The AI fired an autocannon while facing the human directly. This maneuver is called the “Head-on Gunshot.” Human pilots instinctively avoid these maneuvers for fear of collision. It is also prohibited in training regulations. However, through tens of millions of virtual deaths, AI found the answer on its own: “If I fire a precision shot 0.1 seconds before collision, I can shoot down the enemy before I die.” This is not a tactic taught by humans. This is the winning equation that AI discovered after infinite trial and error.
This process is called “Curriculum Learning.” If you are stuck with a veteran pilot from the beginning, the AI will not learn anything and will just keep dying. So at first you simply learn to maintain level flight. Next, learn how to aim at a stationary target. You develop your skills by meeting increasingly difficult opponents. In the final stage, you fight with yourself or your past self. This is called “Self-Play”. The room used by AlphaGo when it conquered Baduk
It's like the law. The me of today destroys the winning strategy that I used yesterday, and the me of tomorrow uses that destruction strategy again, evolving infinitely. Ultimately, reinforcement learning begins with the “freedom to fail.” For a real-life pilot, one failure means death. However, AI in the virtual world becomes immortal by dying tens of millions of times. AI condenses the experience that humans accumulate over a generation into a few days. Beings who have been trained in a world where the laws of time are different come down to reality. This is the real reason why Alpha Dog Fight's AI was able to overpower veteran pilots.
AI is not just about fast calculations. This is because through tens of millions of deaths, humans have accumulated an amount of experience that cannot be experienced in a lifetime.
Kim Kyung-jin
Attorney · Former Member of the National Assembly · AI Policy Researcher
© 2026 Kim Kyung-jin. All rights reserved.