I think the REALLY interesting design question is what happens OUTSIDE the law firm, when the client is no longer so dependent on the firm as the organising centre of legal knowledge, process, and relationships. Of course, law firms will harness AI themselves. But the resulting gains will likely accrue to the firms, without fundamentally changing the client–lawyer relationship. And that's the opportunity we risk missing. Here's where a Jevons-style possibility gets fascinating! As the cost of legal work collapses, I suspect demand could EXPAND dramatically. But who captures that expanding opportunity? Some of the most extraordinarily competent people I've ever had the good fortune of watching operate have been paralegals. What if, rather than making them redundant, AI actually UNLOCKS them? Not simply "paralegals equipped with AI", but paralegals operating within a much richer, citizen-first collaborative meta-medium, where the client retains their context, history, and agency, and brings lawyers, paralegals, and other specialists around the work with clear boundaries, provenanced version-control and accountability. The lawyer still brings professional judgment and accountability where genuinely needed. But crucially, the new arrangement should require LESS actual lawyer involvement per matter, not more. More work undertaken by capable clients and independent specialists, with lawyers brought in precisely where their expertise and professional accountability are required. Perhaps we can also design out some of the more problematic incentives of legal services as a "credence good" - not by eliminating the asymmetry of expertise, but by preventing it from becoming an asymmetry of agency, transparency, and control. Suddenly we have the possibility of MORE human expertise becoming valuable, not less. A richer ecosystem of collaboration, rather than merely more efficient law firms. For me, the really exciting design principle is collaborative dignity: putting the citizen at the centre of relationships traditionally organised around asymmetric expertise and authority. Law is just one particularly interesting instance of a much broader problem! • More legal work • Less lawyer dependency • More human capability. 🦦 Anyhow, (bla bla bla)...nice little opportunity to consolidate some thoughts on this in my head - I'm super looking forward to enjoying "Cupertino"...fun! As always, thanks Prof! 🫶