3 ms·
Not disagreeing about the risks of AI here. But the thing is, there are correct ways to do things in court. But this guy's strategy was to improve his negotiati
by ajb 1y ago
Not disagreeing about the risks of AI here. But the thing is, there are correct ways to do things in court. But this guy's strategy was to improve his negotiating position, pre-court. That sounds like it's a lot more to do with assessment of risk, assessment of your adversaries strategy, and business position. Which are all matters that totally aren't solely the province of lawyers.
In effect, that means that the lawyers should have been doing what he got the AI to do.
- cogman10 1y ago> in court. But this guy's strategy was to improve his negotiating position, pre-court. "The court" generally just means stuff that involves a judge. Not the actual trial. There is no "pre-court" really. Whenever you and the opposing side have a disagreement, a judge will be involved to determine who should do what. That's what filing motions are all about. For example, in a deposition, your lawyer is going to ask the opposing side 1 million questions. While asking those questions their council will pretty frequently say "objection". Each time they do that, they can go back to the judge (court) and fight over whether or not that line of questioning will be admissible into evidence in the trial. The guys strategy is around the motions and briefs filed. It's not infrequent that a judge will ask each side for a brief about the current legal state and why their motion should be granted. That's where mountains of lawyer money can be spent. Every hour they spend researching the caselaw around a motion they get to bill you. But further, strategy involves how often you fight the other side. There is an endless amount of stuff lawyers can argue over which ultimately does not (or barely does) affect the final outcomes. Being able to tell your lawyer "No, don't fight this motion it won't really matter" is where money can be saved.