One of the first key takeaways from Vanderbilt Law School’s conference on Thursday about artificial intelligence is that the term doesn’t carry much value in the scientific community.
“A.I. is whatever we can’t do this year,” David Lewis, a speaker who holds a PhD in computer science, said in between panel sessions.
Lewis estimated we’re currently experiencing the second or third wave of “A.I. hype,” in which everyone uses the term to describe their technology. That’s happened before, he said, and then it went out of style as a marketing buzzword. “By 2020, it’ll have a negative connotation again,” he predicted.
Still, hype aside, there have been real advances in machine learning and other computing power. Predictive coding, the “flavor of the day,” as Lewis put it, relies on supervised machine learning algorithms to reduce the number of documents that lawyers must review, and probably does reduce the cost of a project, he said.
But less important than what you call a technology is whether it’s having an impact. Deduplication — the tool that removes duplicate documents from a data set is simple technology, Lewis said, likely the most effective tool on the market for culling the number of documents that need review.
“Does it matter what you call it? Well I guess it matters for one reason: how much you can charge for it,” he said. “What really matters is whether it works.”
Others at the conference agreed that artificial intelligence is being bandied about for all sorts of technology.
“Nowadays, that term gets thrown around a lot,” said Andrew Arruda, the CEO and co-founder of ROSS Intelligence, which is a building an “A.I. lawyer” based on IBM’s Watson technology. “I like to think about it in law as kind of a monster under the bed. We all get scared about it.”
He defined A.I. pretty loosely, as teaching a computer to do “something intelligent,” a task that was traditionally only associated with something humans could do.
As an example, ROSS Intelligence can field natural language queries on bankruptcy questions. It’s homepage has a single search query, to which Arruda typed: “Are oil and gas leases executory contracts?”
ROSS surfaces the relevant case law documents and highlights the most pertinent passages. They hope to scale out ROSS to other practice areas and on Thursday, Arruda announced Baker Hostetler is one of its first paying customers.
But one of the audience members, John Tredennick, CEO of Catalyst, said he’d used similar natural language queries 20 years ago as a lawyer at Holland & Hart. The real difference is just that there’s been a lot of breakthroughs in developing machine learning algorithms, and ROSS can handle more complex queries, much faster, said Aruda.
By far, the main theme of the conference was that the new technology on the market is already changing how law is practiced and demanding new skills of lawyers.
Clinton Sanko, a Baker Donelson shareholder who heads up the firm’s e-Discovery unit, said that he believes many lawyers will need to acquire a base layer of knowledge about document review in order to be effective, in much the same way that medical malpractice lawyers need to understand medical technology to ply their trade.
The keynote speaker, Richard Susskind, who’s written books predicting the rise of artificial intelligence will drastically change the legal profession since the 1980s, acknowledged that the technology has changed in ways he didn’t necessarily anticipate.
As to law firms seeking to adapt to the changes, Susskind said, “I don’t think this is urgent … I don’t think this is, ‘If you don’t do something in the next six months, you’re done.’”
During the event, there was no single idea or advice that every speaker agreed law firms must take to stay competitive and adapt to the coming technological changes. But nearly everyone took aim at the billable hour as the underpinning of the law firm business model.
Law firms are unlikely to adopt technology that makes them more efficient, so long as they’re being compensated based on how many hours they bill, said Tredennick.
“One of those classic points of wisdom is that disruption almost never comes from inside,” he said.