Well one piece of advice that might sound self-serving, but it isn’t in a sense is that I spent a lot of time building tools that I think are really well optimized to actually explore many questions. But among them, questions and complexity the whole Wolfram Language stack is really well optimized for these kinds of explorations. Learn it as well as possible.

It’s always been surprising to me in my life because I make a point of learning tools, building tools if I don’t think they exist, and you know I see people who are like oh well you know it’s good enough for me to just do this and this and this. It’s like no you’re almost certainly wrong you know for example when I worked on A New Kind of Science which was as I say took me ten and a half years to work on this book at a very slow rate because I was figuring out everything that was in the book. But you know I try to figure out. If I hadn’t built Wolfram Language or Mathematica first would I have finished the whole project sooner or later? My very clear conclusion was I wouldn’t have finished that project in my lifetime.

In other words, even though it took me five years roughly to build the first version of those tools. Building those tools accelerated getting the project done by at least thirty years and so it’s kind of the main advice which people are surprisingly bad at taking is learn the best tools you can. You know I’ve spent a lot of effort building these tools very well optimized for this particular purpose. So, it’s kind of, silly not to use those tools and as a practical matter, you can even get them free everywhere pretty much at all the universities and even on the web and etcetera, etcetera, etcetera.

So, it’s kind of, like the only excuse is, well I can’t be bothered to learn it. One of the things for me that’s interesting is I’ve been teaching Wolfram Language a bunch to kids. Turns out by the time the middle school kids can get proficient with this so can a complexity researcher, one would hope.