You. How do you think about a case?

Because the accusation is written to persuade and the record is not. The claims, the specification, and every exchange with the patent office exist before anyone knew there would be a fight, which makes them the only part of the story nobody could shape for this occasion.

Reading all of it is slow. It is also where the case usually already is, sitting in a sentence somebody added years ago to get past a rejection. Two engineering degrees and years spent writing these documents mean the reading happens without a translator standing in between.

Neither, until somebody explains it. Then it decides everything.

Every person who will rule on this, vote on this, or fund this did not build the technology. A judge, a jury, a board deciding whether the fight is still worth paying for. Complexity that cannot be made understandable becomes the other side's property, because confusion always favors whoever benefits from nothing changing. The work is to make it plain without making it wrong, and those two things pull against each other constantly.

Details, almost always. A limitation added to overcome a rejection. A sentence in an expert report that cannot live in the same room as something the inventor said under oath years earlier. One word in a claim that the applicant defined and then forgot about.

The discipline is choosing where to press. Pressure applied everywhere at once costs a fortune and moves nothing, and it is also how a matter that should have taken months takes years. Finding the one point that changes the arithmetic, and then staying on it, is a cheaper strategy than being loud in every direction.

Composure in a courtroom is a thing bought months in advance. What looks like calm on the day is usually just the fourth time somebody has answered that question out loud, and the first three times were in a conference room with people whose job was to make it go badly.

The same habit gets used on the people coming up. A junior lawyer who is going to argue for the first time gets the full mock version first, with real questions, until the argument stops being a script and starts being an answer.

Because it is the same job. Before practicing law there was a stretch of it at Missouri: undergraduate business law, graduate mathematics, and an engineering lab. Later came Patent Academy, a training program for patent lawyers built at Shook, Hardy & Bacon, and continuing legal education taught for years after that.

A jury is a classroom that did not volunteer, cannot ask a question, and is graded on nothing. Everything that works in a real classroom works there too. Start from what they already know. Say the hard part once, slowly, and then show it rather than repeating it louder.

As an amplifier and nothing else. There was a computer-services company, ProSolv, founded to help pay for college, then a stretch as an analyst at Accenture, then executive education at Wharton on the economics of blockchain and digital assets. Artificial intelligence is in daily use in the practice now.

All of it makes a prepared lawyer faster. It makes a careless one wrong more quickly, and more confidently, which is worse than slow. No tool has yet read a prosecution history and noticed the thing that mattered, and until one does the reading is still done by a person.