@willchen500 There’s a good amount of flat fee for sure but there are a ton of attorneys working hourly. Think about the small construction companies, the crim law attorneys, the attorneys working with city govs, and a ton more.
@willchen500 You have a great pov for biglaw but only ~10 of attorneys work in biglaw. Does anyone have any understanding of how legal ai will work for the people that aren’t in m&a or contract review?
@Sony91121271 We’re doing it already. Have several firms running on our software. We provide white glove service and will create any custom tool they ask for.
@OrinKerr@StanfordLaw I raised Clapper because Flock uses a similar approach to surveillance. I disagree that there’s no workable middle ground. Carpenter and Chatrie give the Court a basis to distinguish ordinary police observation from mass surveillance. I think the Court will have to draw that line
@OrinKerr@StanfordLaw Moalin (9th 2020) later rejected the gov's "ex nihilo" argument, that zero 4A interests can't add up to one. Everyone else's data makes yours more revealing. Chatrie says what matters is what the gov "can access." So the unit isn't the camera, it's what the city bought.
@OrinKerr@StanfordLaw Fair, but your excerpt says the constitutional concerns were serious enough to shape it. And the big change since 2015 is Carpenter, which seems to make the aggregation point stronger, not weaker. Where do you see the post 2018 law cutting the other way?
For law firms to fully use AI, they have to have information structured such that it can be easily given to AI.
And that's about more than just connecting a DMS to your AI tool. It's about getting all the observations, facts, and details that are outside those documents in a format that AI can easily use and act on. Current DMS and practice management systems are not optimized to make it easy to store this information so AI can use it.
But it's about more than the technology. Firms have to adjust their processes to better capture information. Facts and other observations about the client (preferences, etc.) cannot live in a lawyer's head - they have to be written down. You have to act as if you're about to leave the firm and hand the case to another lawyer. If you store that level of information, you'll be able to maximize the value you get from AI.
We are still, as a profession, figuring out what the shape of the tools should be to enable this. It's about making it very easy - automatic, if possible, to capture and communicate context. It feels unlikey that the current DMS and practice management tools are going to be the right form.
I see CEO’s of legal AI companies speaking on social about certain legal work using language that just doesn’t sound like they’re familiar with that specific legal work or they don’t really know what they’re talking about . They don’t speak the language or the pain . This is usually non-legal CEO’s. It doesn’t sound credible or relatable to lawyers and legal teams.
That’s a strong issue especially at a time when depth and expertise are strong differentiating factors for legal AI vs generic AI. The experts can smell the BS a mile away. This usually means that these companies don’t have the right people, language, and culture across the company and product. This usually translates into a lack of depth in the product also.
Legal vertical AI is about depth and expertise and about people, teams, and products that understand, live, breathe, and speak the language of the respective legal domain/problem they’re focused on. Learn the language and pain. Do the work and find the right people to instill this empathy and understanding across the company and product.