Ryanfilevine

There’s a big difference between making an AI for a lawyer and an AI for a Lawfirm

In a biz meeting, you don’t need to say: “I am ethical” “ I am honest” “I do the right thing” “I always treat ppl the right way” Anytime I hear any combination of these, my grifter radar lights up

GIGANTIC win for our property management group! Jesse Smith and team were just awarded PM contracts on two of the largest/ newest retail projects in the state of Utah! Paseo in Washington/ STG & The Commons in Herriman Thank you to Elevate Properties for choosing us for both PM & Leasing.

A common dynamic I observe with AI: it feels most impressive when you don’t know much about the subject, don’t care or don’t have a clear idea of what the you want. This applies across design, code, legal, and more. If I don’t know code very well, every piece of code it writes feels very impressive. Once you know what something should feel or look like, it becomes almost impossible to guide AI there. And you definitely can’t one-shot it.

In the same vein, writing ≠ generating text In fact, generating text is the LEAST important part of the writing process So many people mistake LLMs as good writers just because it can generate text well

I think the sequoia piece is slop. I said it. It’s slop. Coatue too. Stop this charade. The whole idea that you can sell an already outsourced service and internalize returns from AI without facing price pressure yourself is pure fantasy. Why must we endure this nonsense? Sequoia’s Judgment v Intelligence is the wrong framework and really tells on Sequoia for their views of AGI or commitment to precision. Judgment of legible things is steadily making its way into intelligence / the framework collapses into a guess about timelines — not interesting at all. It seems there really are two forms of selling services - transitional services - structural services Transitional services provide initially human services that transition to AI as AI gets better. These services—like those of Crosby—are not Load Bearing in that they are not fundamentally responsible for outcomes. The person who uses the Crosby issued contract is responsible. Crosby doesn’t list their TOS on their site but seems unlikely they offer real indemnity — traditional law firms don’t. These are transitional also because as AI gets cheaper for you, it gets cheaper for everyone else too. The firm of This Is AGI is truly not AGI-pilled. Alex Rampell’s > The TAM is Human Labor and Coatue’s > selling work is 25x larger than software are defunct for the same reason. Just because you internalize the benefits of AI doesn’t mean you’ll be the only one to do it. The other form are structural services. Here the service provider is rather an operating partner that actually absorbs liability for (and hopefully participates in the upside from) the delivery of outcomes. Even as AI commodifies the structural service is difficult to rip out because of the migration risk. The best emerging example is in AI audits. Unlike Crosby, where moving to Crosby 2.0 would face little risk, moving to a new load bearing audit firm would constitute serious risk. I’m betting on the latter.

“I thought Harvey was amazing until I saw this. Harvey is like analog in a digital age compared to you." Words that have stuck with me over the last 12 hours. LOIS’s ability to resonate with such a diverse group of legal leaders—spanning in-house teams, law firms, BigLaw, and government—is a testament to its innovation. It’s rare to see this level of enthusiasm from such a powerhouse crowd. LOIS is changing the trajectory of Legal AI! Thanks DC for an incredible educational discussion around AI in legal tech!