“AI” can ascend, but not without “HI” (human intelligence)—especially in the legal realm.
Artificial intelligence (AI) is the new headline-grabbing buzzword these days as we come to appreciate the extent to which it has permeated our daily lives—not just in conversation—but in reality. Everyone’s talking about it and it really is everywhere.
Oddly, although we’ve been using AI in eDiscovery for years in the legal community, we’ve never gone out of our way to call it that. Perhaps some intentional marketing misdirection to temper fears of rampant innovation, we’ve come up with other terms—such as predictive coding and technology-assisted review—that have firmly attached to machine learning and other AI-driven processes.
It’s time to lift the veil. Machine learning is a type of AI, and finally, the cultural permeation of the term is driving the legal community to allow the proper nomenclature to emerge.
No worries! We can all rest easy. It has been demonstrated repeatedly that machine learning AI systems, if designed, executed and measured for performance by experts with the appropriate scientific competencies, can perform document review better than human attorneys can.
So, does calling it AI make a difference? Maybe. Familiarity can breed comfort as well as contempt, after all, but the question is: what will this recognition foster, and how soon? Will a widespread acceptance of the term applied to what we’re already doing make acceptance of AI, in all of its growing incarnations, more likely?
As is usually the case with so-called “disruptive innovation,” there is a fear factor to address, real or imagined. In the legal community (and most others), the mere mention of AI will usually drive the conversation to the number of jobs that will be lost as the “robots” ascend. This is unfortunate. Rather than recognize AI as enormously powerful software tools that could enable counsel to predict, analyze, and assess in insightful new ways aspects of the legal matters they handle (as well as their own businesses), it is anthropomorphized instead and plopped into the poor, unemployed lawyer’s still warm chair.
It’s the same conversation that goes on in the face of any disruptive technology (yes, cars displaced blacksmiths, but, hello—car mechanics and gas stations?). Yes, there will be job disruption, but there would be anyway; other evolving technologies will make that unavoidable, AI or not. Some legal processes (e.g., the document review that is now commonly “technology-assisted”) must be addressed with at least some kind of technology to be done at all—data volumes have just become too massive for page-by-page manual human review. And anyway, these are not the most sought-after human jobs, to be sure.
Forward-thinking law firms (the ones that will likely be in business in the future) are already boldly considering ways to use AI and data science to leverage, not displace, the crucial human intelligence that is at the very heart of their work. Taking the cue from their own clients, who are beginning to harness the power of their data for business intelligence, these firms are looking beyond traditional legal workflows to explore how AI and data science, augmented as they must be by human expertise, can reshape both their client relationships and their own roles as legal practitioners. According to Legaltech News, many firms are bringing data scientists on board or have begun to build tech-savvy teams that focus on data analytics. Still others are growing by cultivating practice expertise in technology, including the business ethics and obligations that attend to its use, in order to guide their clients through the implications of their technology decisions and strategies.
Some top-level legal service providers are on the same path, upping the ante on their offerings with hybrid teams of legal and AI/data science experts that can provide customized, technology-driven solutions to legal challenges that are becoming more complex and risky every day.
To face the future without fear, we must cultivate a greater understanding of the technologies we create and use. AI and other tools and processes derived from data science carry enormous power, but there must be a rigorously principled approach to their development and execution. They must be created and advanced with the requisite human expertise and oversight to ensure appropriate, accurate and valid use, with transparency, measurement, and accountability top of mind. If there is any kind of “intelligence” to be applied here, it is that.
Learn more about a principled approach to using AI in an article by H5 CEO, Nicolas Economou.