I put 3 AIs in the same universe and let them compete to build a Dyson Sphere. They’re starting to behave differently.

I’ve been thinking about this question: If you give different AI models the exact same starting point and rules, do they eventually converge to the same strategy or actually behave differently over time? I tried setting up a simple simulation around this. They all start on Earth with the same resources and have to deal with expansion, energy, random problems, and eventually try to build a Dyson Sphere. What surprised me is they’re already making different choices pretty early on. Curious what people here think.