A year or two ago, any discussion surrounding the future of predictive technologies in e-discovery was usually met with a high degree of skepticism. How could a machine be trusted to replicate the judgment of a trained lawyer?
Fast forward to last month's LegalTech New York conference and attitudes have clearly evolved (although there still doesn't seem to be consensus on what to call predictive products). Skeptics have largely come around to recognize the value of predictive analytics in e-discovery or, stated another way, have come to recognize the futility of applying traditional methods to analyze exponentially growing data volumes. Meanwhile, a number of lawyers are now actively engaging with the software and beginning to apply it in real cases. As further evidence that the e-discovery world is embracing the predictive movement, several experts spoke of moving the technology outside the review bubble and applying it to earlier e-discovery stages (something the E-Discovery Beat has covered regularly).
It's fair to say that everyone involved in e-discovery is a bit curious what the conversation surrounding predictive technologies might sound like a year or two from now. Only time will tell. In the meantime, here's what a handful of e-discovery experts had to say about the future of predictive technologies at LegalTech:
*click on the speaker name to access a longer interview
I think that (predictive) will replace keywords in terms of the full search. Keywords still may be useful in coming up with the seed set that one uses to train the predictive coding, or TAR (Technology Assisted Review), software. Also, predictive coding, even when it gets much greater acceptance, which I hope it will, may not be appropriate for every case. There are still likely to be cases where, because of lawyer comfort levels or the amount of money in issue, loading (documents) onto a vendor platform to do a TAR project can be too expensive.
The truth of the matter is that this is a disruptive technology. It disrupts the current establishment of vendors, the current establishment of law firms, because it's so much better. The more you learn about it, the more you realize that this is really big; it is changing everything…Big Data requires artificial intelligence. It's the only way to do it. And the good news is we are now learning how to do it… I do think that the application of computer assisted review or predictive coding at the front end in the organization of documents is going to be something that we'll see more and more in the future. This is just the very beginning in a long term wave.
A lot of attention has been focused on (predictive), with a lot of talk and arguing going on. I think that's good because it's on people's minds, meaning that acceptance of it may be coming faster… I think the nice thing about predictive coding is that you can utilize it in so many different ways. If you want it for automated review, great, but you can also use it to QC (quality control) your documents as well. So you can still have human beings driving the bus and at the end of the day use predictive coding to see how well things measured up. A lot of utility can come from predictive coding, with different flavors depending on what fits your workflow and your culture.
I think we'll see a little bit more (adoption) in 2013. But as we've seen with any new technology, it's the lawyer's role to be skeptical of everything that comes around. I think the fact that we have a couple cases now showing judges acceptance of it will lead incrementally to a little more adoption but I think we're still maybe three or four years away from wholesale adoption.
Having been through and worked in the Global Aerospace case and trying to get people to use predictive coding, I see a lot of significant barriers still out there but I also see continued growth. The biggest barrier is not having people knowing how to use the tools. Even if they have some idea it's usually not very deep, but they know that the person on the other end or the judge doesn't have a good understanding either.
I think three years from now it will be extremely commonplace. Everybody is moving in that direction. People will understand and appreciate the cost savings associated with it and it will become completely mainstream just like the evolution of so many others tools and applications in this space.
To learn more about the future of predictive technologies, watch Exterro's on-demand webcast Best Practices for Leveraging Predictive Technologies.