It's been writing most of my posts lately (about 60-70% of them, and/or the outlines/gist of the ideas).
It's surprisingly versatile, and I hope to share it with you all in the coming months, along with my news scraping/prediction algos (though I don't want to give the regime an advantage on that front, so idk yet).
I'm currently building a reinforcement based simulation to do short-run bootstrapping (sanity checking) of system output.
Window size, or context, has surprisingly not been that important. Neither has a graph-based backend.
Cross-referencing has been useful, and you're not the first to suggest it, which is why I'm working on a mixture of experts metasystem to scale it up.
Understanding roles within political machines I think will go a long way toward enhancing the entire predictive power of the system. Think tammany hall or the dictators handbook. Social-games theory has allowed me to make a lot of predictions manually, and I want to implement a simulation of that (see the book "games people play") for assessing individuals and small-world graphs (NGOs and so forth) in relation to the larger roles they play. Think the 'schlemiel' game highlighted in the aforementioned book. When you look at it, and then look at the field of u.s. presidential politics, you'll notice the immediate application of finding the correct social games for a given context for deriving any given playbook of most NGOs, governments, agencies, parties, individual politicians, etc.
anon 0 points 5 months ago
Feedback loops are already part of the system.
I'm currently building a reinforcement based simulation to do short-run bootstrapping (sanity checking) of system output.
Window size, or context, has surprisingly not been that important. Neither has a graph-based backend.
Cross-referencing has been useful, and you're not the first to suggest it, which is why I'm working on a mixture of experts metasystem to scale it up.
Understanding roles within political machines I think will go a long way toward enhancing the entire predictive power of the system. Think tammany hall or the dictators handbook. Social-games theory has allowed me to make a lot of predictions manually, and I want to implement a simulation of that (see the book "games people play") for assessing individuals and small-world graphs (NGOs and so forth) in relation to the larger roles they play. Think the 'schlemiel' game highlighted in the aforementioned book. When you look at it, and then look at the field of u.s. presidential politics, you'll notice the immediate application of finding the correct social games for a given context for deriving any given playbook of most NGOs, governments, agencies, parties, individual politicians, etc.