An instant interactive explainer from Claude for a frequent debate about correlation & causation. Prompt: “Create an interactive tool that explains all the ways two variables can be correlated (causation, random chance, reverse causation, etc)” Play it:

Too complicated? “Cool but too complicated for people without a stats background”

Khan Academy in a single prompt

@TomChivers

This will be so good for learning. Personalised artefacts that speak to the context of an individual learner vs. generic material.

Wow.

Great example of how AI tools can turn complex concepts into hands-on learning. Interactive explainers like this don’t just “tell” they automate the teaching process and boost productivity for anyone trying to grasp correlation vs. causation.

Useful... Bookmarking. Thanks!

Most debates die because people treat correlation like a single explanation. It’s a family of mechanisms. Reverse causation, confounding, selection bias, common cause, feedback loops, pure noise. A tool that forces enumeration ends the debate fast.

this is the kind of tool i wish existed when i first got into stats

This is a brilliant teaching tool that demystifies statistical relationships saved it to my learning resources for the next time I'm overthinking data patterns in my projects.

Nice UI, but it seems to confuse itself when trying to distinguish Confounding from Reverse Causation. In fact you could almost say it confounds the two :-)

can we get a p=np

gm fren, how's it handling noisy datasets?

education is going to have a great time with artifacts or whatever interactive learning comes next

I must say I don't understand what's going on here and sadly the example is I guess too theoretical. I just see a pile of random numbers on top of random numbers. What background is required to understand this tool? Statistics background? Math?

Definitely intriguing. But how user-friendly is it for beginners?