Originally posted on July 28, 2012.
When the original DocReviewMD website lost its content, I was disappointed. We posted some of the first commentary on many of the cases that were breaking in the predictive coding space in 2012. Now I am reposting these blog posts with updated introductions. This was the first big conference we did after Global Aerospace became a big case. We thought we were changing the litigation landscape. As a prominent judge in the e-discovery space told me after an ESIBytes podcast, common law takes a long time to evolve and that has served us well for 500 years. In its way predictive coding is common law as lawyers need to build experience with these tools. This is one reason we have embarked on a national tour in 2013 called the Predictive Coding Thought Leadership Series to help lawyers learn more about these new tools. Enjoy the write up below done over a year ago……
Now that I am three days removed from the Carmel Valley E-Discovery Retreat (CLEDR 2012), I wanted to blog about my observations. First of all, many thanks to Chris La Cour for providing a great atmosphere to meet and discuss complicated E-Discovery topics both in sessions and informally on an outside patio overlooking the ocean. The later discussions were among my favorites as lawyers sometimes disagree on e-discovery topics so it is amusing to hear a third voice in the background of the debate.
About the conference, I spent most of my time in the E-Discovery panels and missed much of the discussion on the cloud. However, because my judges panel required working with the AV staff to ensure the hook ups and microphones enabled the audience to hear three judges calling us from Colorado, I spent a moment in one of the cloud sessions. There was tremendous irony in hearing some lawyers describe why the cloud is risky and should be carefully used. Then walk up to a review panel with lawyers saying technology assisted review improperly used can risk turning over work product. The presentations raised good points but in addition to the commonality that in both cases these were lawyer groups pointing out why technology that most business groups are embracing might cause problems, it is interesting that it is lawyers who are saying things like this. As a group, we need to figure out how to offer advancements in our space instead of offering advice that the use of technology can cause bad things to happen. While this is true, so is crossing the road without looking both ways. But we aren’t getting rid of cars because of the risks of accidents.
Now for the technology assisted review programs I moderated and worked on putting together nine months ago. The judge’s panel was fantastic. The discussions between Judges Peck (SDNY) and Fischer (WD of PA) who were live in Monterey and Judges Facciola (DC), Maas (SD of NY) and Waxse (Kansas) were always entertaining and full of great observations. Rather than summarizing the content, I’d encourage you to download the show which will be posted in its entirety on www.ESIBytes.com. But we had some great discussions on why key words have not worked in past cases, the future potential of using advanced analytics such as predictive coding, and the role of experts in these matters. I have started listening to the show again and the sound quality is very good thus far. I can’t wait to post this program which I am also making available to distribute to the Federal Judiciary Center.
The second panel of comprised of power users of advanced analytics we may or may not post live on ESIBytes. We had Maura Grossman from Wachtell Lipton who has been running TREC and publishing lots of great academic articles demonstrating how these tools out perform traditional linear review, Tom Gricks, the head of E-Discovery at Schnader Harrison, and a chemical engineer educated trial litigator who argued the Global Aerospace case in Virginia. Bennett Borden, head of e-discovery at Williams Mullen who has published articles benchmarking the results he has achieved using analytics and software. Herb Roitblat, the Chief Scientist from OrcaTec, a predictive coding and analytics software program and Chairperson for the E-Discovery Institute which has published ground breaking academic papers on the utility of using predictive coding, and David Lewis, a computer scientist who has many patents in the field and is an expert in the Kleen Products case. My goal in moderating this panel was to use my background studying and working in the space to get Maura Grossman and Tom Gricks to engage in a “friendly fight” between using seed sets versus random sampling to train a predictive coding program. The dialogue was spirited, the audience had beer and 4th C Wine to drink, and I announced each round. This could have been held in Las Vegas ringside. At the end of the day, these tools are GPS like and they get you to results which can be defended using a number of different approaches. All the participants agree on this point. We also recognized that we are early into the process so it is worth examining different types of approaches. What showed this panel really worked well is we were scheduled for 4PM and we had the close to the same number of attendees which I would estimate as over 200, as we had for the judicial panel after lunch which was done without any competing panels. We also ran 20 minutes long and everyone stayed till we decided to shut it down.