Isha Marathe has an interesting post today. I've read it three times and will probably read it another three. It's interesting in that it touches on multiple conversations I have in the last month or so. The first was a conversation with a good colleague at an AmLaw 10 firm (considerable larger than my firm) we discussed outside counsel guidelines and the prohibition of the use of client data in generative AI systems. The second was an internal meeting where I brought up something I had written about years before about size mattering when it comes to AI - and if a client's AI would have more of an advantage over a law firm by nature of their own data on themselves - and if big firm would have a strategic advantage over a small firm by the sheer size of the document repository - and further if small law firms would have to ban together (assuming that could even be ethically accomplished) to be able to weaponize their AI to be a player for legal business. I'm not sure I have answers yet, but some things in this post that will make you think include: "What's Making Swathes of Legal Data 'Untrainable'?" "It's unclear whether training on specifically legal data provides enough of an incentive over a general purpose model." You definitely want to read more at Legaltech news: Legal Tech's Problem: Generative AI Is Readily Available, But Not Legal Data
Published September 15, 2023.
