As of today, we work with AI (neural network approaches) and especially the language models by trying something and seeing what we get.
Everyone knows about Claude and ChatGPT, but evaluating the local large language models is far more difficult.
I think in a ideal world, we'd all have our own AI companion. It would be customized the exact way we'd like that is private and secure, without control of corporate agendas trying to monetize us or persuade us to use certain products.
By funding us, you help run experiments to reach that ideal world.