Youre pretty accurate wrt dark patterns but there’s other stuff like user journeys that map out for instance how much and when to deliver a casino coupon to a user based on research into addictive behavioural patterns.

Hypothetically: how does the glycemic index of typical lunch foods impact mood and when in that cycle is an addict most likely to succumb?

Put it like this: if it can be done, it is being done.

I had a promising career in Datascience very early on (2005 - this was before the CUDA framework was released) and I quickly realised that there was very little good that could come of that work under capitalism. I have friends who had aspirations in aerospace who realised that all they would end up building is weapons. Similar deal. I left it behind and I have a comfortable life in good old fashioned data management now. It isn’t sexy but I can sleep at night.