Thanks for the post @blue2!
You’re right that these “scale-out” decisions have been deprioritized for now. We do think speed and scalability are critical components to the long-term success of the project, however, right now we’re focused on iterating quickly on experiments to determine the correct path to chose. We do keep an eye on performance and you can check out the benchmarks here: https://thousandbrainsproject.readme.io/docs/benchmark-experiments
There are also interesting scale conversations happening in the community here: Elixir-based subsystem(s) for Monty? and here: Software architecture for neural voting
