Simulating behavioral health interventions in the age of generative world models

less than 1 minute read

Published:

This is a sample blog post. Lorem ipsum I can’t remember the rest of lorem ipsum and don’t have an internet connection right now. Testing testing testing this blog post. Blog posts are cool.

  • Topic: Simulating behavioral health interventions in the age of generative world models

Sections:

  • history/motivation: what we need

  • history/RW: what we tried before…

  • history/misguided and bad RW:

  • Types of paths forward: a. create simulations solely based on true ‘real-world human interaction data’ b. create simulations only using heuristics, rules, without any direct input of “real-world human interaction data” (e.g., you can use a pre-trained LLM, but you cannot include human interactions as parameters of the behavioral simulation.) c. hybrid of (a) and (c) …… how might we do this, and when would it be more/less desirable than (a) or (c) ?

Headings are cool

You can have many headings

Aren’t headings cool?