> Nowadays I'd probably just ask Claude to figure it out for me
Incidentally, Mathematica + LLMs make a great combination. If you take what is pretty much the biggest mathematical routine library in the world and combine it with interactive visualization tools, and then use an LLM to accelerate things, it becomes an incredible tool. Almost ridiculously powerful for trying things out, teaching, visualizing things, etc.
(I've been using Mathematica since 1992 or so, so I'm familiar with the language, but it's still so much faster to just tell Claude to visualize this or that)
Yes, I’m sure! I had mostly stopped using WL by the time ChatGPT came out and so my only experience of their integration is a lot of hallucinated syntax by GPT3. I’ve not been tempted to upgrade to the newer, more LLM-integrated versions yet, particularly because I’ve been trying to cut down on expenses and spending more money on LLM api calls versus just using my existing Anthropic subscription for Claude code isn’t that appealing. I’ve also been going through a bit of a crisis of a lack of creativity and inventiveness and it’s hard to decide to spend money on an update when it doesn’t seem like I have any good ideas of what to do with it after I’ve spent the cash!
I bet that if they can integrate LLMs _really_ well (I’m not sure the chat driven notebook thing is necessarily the way) it’ll be a massive upgrade.
Incidentally, Mathematica + LLMs make a great combination. If you take what is pretty much the biggest mathematical routine library in the world and combine it with interactive visualization tools, and then use an LLM to accelerate things, it becomes an incredible tool. Almost ridiculously powerful for trying things out, teaching, visualizing things, etc.
(I've been using Mathematica since 1992 or so, so I'm familiar with the language, but it's still so much faster to just tell Claude to visualize this or that)