I asked Claude to look up the X-Plane 12 API and try to fly a Cessna from Hainan to somewhere nearby. It kept a pilot's log. Everything below is Claude's own notes from during the flight. I occasionally had to remind it to keep the pilot log running, and had to tell it when it crashed as xplane would quickly restart to the runway. The main issue seemed to be delay from what it saw with screenshots and api data and changing course. It quickly made itself a python script to take off, then would add functions to the script as it needed to take action. It would position to land then not change angle quick enough and crash into the ground. I am interested to see how claude thinks about its own abilities and if it can reason about realtime events and how to pay attention to things. It isnt to make code that can do a good job at flying a plane. Does claude think in advanced about the fact it will be out of sync, how would it prepare and develop code with that in mind? How does it reason about its own ability? Interestingly it decided to write code to take off before even thinking about how to steer or land. It then came up with instructions to land, sent them to xplane, then took its hands of the wheel for a while, retuning to a crash. I figure this is some kind of AGI benchmark for models thinking ahead and planning what tools to develop and how to use them before take off.