Post #2340505

@lritter@mastodon.gamedev.place one of the problems with “getting to know your enemy” and using local only models, is that there is a substantial difference in capability and the conclusions drawn from one might not apply to another one. I am curious how you think about that issue. I get the moral concern. Of course I have a different position and work for one of the big LLM providers. I did find most local models quite underwhelming thus far tho and not representative of what models can do (for better or worse)