Could we create a domain-specific LLM by training a ChatGPT-like model on a curated dataset of legal blogs?
Such a model could serve as a powerful knowledge base for the public, legal professionals and businesses.
We’ve got a start on the Open Legal Blog Archive backed by LexBlog with over 54,000 authors and 806,000 blog posts. We’re going to ramp that up significantly this year, with not only more blogs, but also offering single post submissions to legal professionals.
The benefit to legal bloggers? Contributing to the advancement of the law, but also being sourced by the model so as to gain significant stature in their niche area of the law.
The AP’s Matt O’Brien reported last year that AI systems like ChatGPT “could soon run out of what keeps making them smarter — the tens of trillions of words people have written and shared online.”
There may look like there are lot of law blogs – and I am not including the junk “law blogs” created solely for SEO – but there are far fewer niches being covered by law blogs than the niches yet to be covered.
I have always looked at the Internet as a way to connect the knowledge of passionate, knowledgeable and caring lawyers – especially from practitioners whose insight was never requested or captured.
The first thirty years of the net has been a start. Seems we can really get cranking now by connecting lawyers to blogs to AI. 😉