The LLM Conundrum: to Build or Not to Build?

 It has long been accepted wisdom that there is no sense in a law firm building its technology unless there is a clear strategic advantage to be had from doing so (and even then, the heavy lift of technology development will often outweigh other advantages).

In this new, post-ChatGPT world, the equation may look quite different. Instead of building technology from scratch, “building” with an LLM – especially one that has already been reinforced with legal content – involves adding a layer of either content or features (or both) on top of the existing product. It’s not so different to leveraging an off-the-shelf experience or matter management solution, then configuring all the fields according to your internal taxonomy and loading it with your data. 

With this technology, in other words, the dichotomy between build versus buy does not look quite as stark. But what are the benefits of taking a foundational LLM and building your chatbot or interface on top of it?

In our view, this is a unique situation. Building does not mean what it used to mean. Law firms will need to experiment to understand what the applicability of the technology is in the legal industry because no one knows yet. It makes sense to think that pointing advanced AI at your data might give rise to insights that an off-the-shelf product simply won’t have access to. It’s exciting that some law firms are willing to take the risk and push forward to see whether building is the right move and delivers a competitive advantage. That is the very definition of innovation.

Source: Legal technology hub





Leave a Reply

Your email address will not be published. Required fields are marked *