I've spoken previously about the project-crushing development known variously as technology-assisted review, computer assisted review or predictive coding. I have previously been part of projects that were brought to a conclusion in short order was TAR was put to use. I think it often doesn't work particularly well -- it depends on who sets up the parameters for the software to base decisions on. Garbage in, garbage out. But it sure as shit works fast. It can turn a 4-6 week project into 2 weeks. Not a temp attorney's friend.
Well, TAR has come to my project. Up until now, official estimates have been that the project likely would last at least until the first of the year, and probably beyond. Oddly enough, there are no new estimates on project lifespan. When they announced today that they were going to be introducing what they're calling "computer-assisted review," the project managers were quick to say "No one is losing their job." Which brings us to Rule No. 1, of course: they're lying. Or maybe they're just playing semantics: after all, it isn't "no one" losing a job, it's "everyone." But that's why you front-end load the overtime every week of every project with overtime, because of Rule No. 3 (or so): Every project ends tomorrow. This one isn't ending tomorrow, but there are some folks in the room today who have already been on this project more than a year who had forgotten all about Rule No. 2. Not sure when "tomorrow" comes for this project, but it will come a lot sooner now.