A local LLM runs a compressed AI model on your own device. With a local LLM, you can keep all of your interactions with generative AI on your own device, never interacting with corporate data centers after the initial download.
Although the best quality local LLMs were created using the ethically questionable processes of the frontier models, when using them you have complete data sovereignty and do not send any data or money to anyone.
We are currently researching local LLM, and comparing different models and apps. With promising results in its early stages, we are at this time able to recommend the app Ollama as the best choice for getting started with local LLMs.
We have recently formed a study group to learn about local LLM and to compare our experiences with it. We’re looking for new members, so please email us at amaranth@unm.edu if you are interested in joining.