Now that is a post title. It is almost as good as those YouTube videos where someone builds a lightsaber or railgun. I watch some of those videos because my first reaction is, “OK, that is 100% not possible to build in a garage.”
The journey through trial and error is fascinating. Prototypes that sort of work to a proof of concept that does. Then the final test where this handheld torch lights up and looks a lot like a lightsaber. The long slide powers up and fires the aluminum projectile at speeds too fast for even a high-speed camera.
The people behind these videos all have a similar background. Either a science or engineering background with a fascination for the other side. A physicist who worked as an engineer or a mechanical engineer who studied physics in their spare time.
In high school, I worked in a physical chemistry lab. I was the smartest kid in a school full of very smart kids. Until my time in that lab, I thought I was a genius. NOPE.
The professor who took me in was that rare mind. He left rooms full of geniuses just looking at each other with wide eyes. 15 minutes of teaching took the rest of us a week to process. He was a brilliant mind with a second superpower. He could teach the most complex concepts to people without his mind.
Some can comprehend at a very high level. Few can simplify complex concepts. Even fewer can communicate their simplifications successfully to students. The scientist educator is a factory for new minds. I want to be like him when I grow up, but I never will.
He introduced me to one of his teaching assistants. She was an engineer, software developer, and scientist. She taught me C, and my aspirations of following my father into civil engineering evaporated. The professor took me in because he saw a lot of her in me. I looked up to him but followed in her footsteps. She was achievable for me because that’s how my mind is built.
Every data scientist has her in us. She was the prototype unicorn, and scientists treasured her. She was an engineer who could understand the geniuses. Her second superpower was the ability to synthesize complex concepts into the real world. She was aware of that gift and explained why it was important. Together, they gave me perspective at a very early age.
Every career has a common thread. Mine is duality. Customer advocate engineer. Leader engineer. Applied data scientist. Technical strategist.
The Duality Of AI
The professor and his teaching assistant taught me the secret to building anything. If you can create an accurate enough model, you can build a simulation. If you can build a simulation, you can engineer that in real life.
Video games are a good example. Build an accurate enough physics engine, and the game can use that model to simulate the real world. Characters jump, fall, and run realistically. The game is immersive because it feels real enough.
AI is a model and the resulting simulation. The TA was my model, and I am the resulting simulation. My father provided a different model. He worked as a pure engineer building foundations for tall buildings where the ground was not ideal for those types of structures. I adopted parts of his engineering model into my simulation.
On War is a book about military strategy. The author describes a model of warfare based on the people behind it. It is the sociology of warfare. The book explains why wars start, different approaches to executing warfare, and the exit criteria for warfare.
Let’s pretend Carl von Clausewitz’s warfare strategy model is based on causal knowledge. The causal graph’s features allow us to run one-dimensional warfare simulations and understand the decision-making process. The purpose of the causal graph is to inform decision-making based on causal knowledge about how our decisions impact outcomes.
The purpose of strategy is to inform decision-making, so strategy models are one side of an AI. Strategy without execution is the slowest path to victory. Execution is the capability side of strategy models. Strategy is science, and execution is engineering. Strategy is the model, and execution is the resulting simulation.