Gregory's Server
@internetarchive If it helps pay the bills I'm all for it. Being paranoid and scared of AI isn't something we should strive for. Using public domain data to train AIs is a slam dunk. Why complain about AI companies stealing data and be against public domain data at the same time? That's insanity.
In fact using the vast amounts of public domain literature should be very helpful to LLMs as long as we use them for their original function (generative general text and not as search engines or answering questions). Hopefully the training data will be processed transparently so others can use it as well instead of making the training data absolutely proprietary and locked down.
In fact using the vast amounts of public domain literature should be very helpful to LLMs as long as we use them for their original function (generative general text and not as search engines or answering questions). Hopefully the training data will be processed transparently so others can use it as well instead of making the training data absolutely proprietary and locked down.