A recent New York Times article explores whether artificial intelligence is replacing lawyers or whether it will in the future. The basic conclusion is that humans are necessary for legal work for the time being. As an attorney who does litigation, with an emphasis on appeals, habeas, sentencing, motions practice, and some limited trial work, I see this as an optimistic piece. Steve Lohr, who wrote the piece does a good bit of technology writing for the Times and has done a great job here identifying trends in A.I. and the law. So, let me take a minute to tell you why I think that this piece is ultimately optimistic.

In any law firm, or any business,, there are very few things that only a lawyer should do. And to become more profitable and efficient, lawyers should limit their work to what only the lawyers can do. What are the things that truly only I can do in my practice?

• Only I can stand up and make an argument in court.
• Only I can approve and sign pleadings and briefs that are submitted to the court.
• Only I can represent a client, which essentially means that only I an give legal advice and ultimately execute on legal strategy.
• Only I have the ability to connect with a client, opposing counsel, law enforcement officials, and judges as a human being in relationship with others as human beings. A.I. would have to surpass the Turing Test, to begin to replicate this function. And the lawyer’s stock and trade is his relationship with others in the system.

But there are things that others in my practice can do as long as I supervise. For me that might mean the gathering and scanning of all the documents in the case that were left by the lawyers and clerks that go before me (I’m seldom the first lawyer on a file). Document review can be done by others in my office as well as the preparation of internal memoranda, witnesses, and court exhibits, and some legal research. Indeed, it’s to the client’s benefit in terms of cost and efficiency if the lawyer does what only the lawyer can do. And the better news is that we live in a time where a good bit of the non-lawyer stuff can even be outsourced to contract vendors who can do an excellent job remotely, which can cut down on the amount of office space necessary to run a law firm. And ultimately, the client should be happy that the client no longer needs to fund such in infrastructure.

So, back to the article. Mr. Lohr identifies some of the areas that A.I. is improving. I’ll highlight a few.

• Legal Research. Ross is one of the A.I. services highlighted in the article. Ross intelligently engages in legal research and generates a rudimentary legal memo. Again, this sort of thing is what lawyers often get drawn into but is not one of the things that only lawyers can do. There’s a great Ted Talk about Ross with its founder. The talk is kind of inspirational.
• Scouting Opposing Counsel and the Judge I think that ultimately there is value in making some calls and using Listserves to scout out this information. But I’m intrigued by Lex Machina and Ravel Law. These services provide analytics of opposing counsel and judges – sort of like a Moneyball for the law. I haven’t looked deeply into these services, but I suspect that they are pretty good for civil litigation and not so great for criminal law.

I’m not concerned that these services will replace me. I think these services will allow me to be more of a lawyer and to compete with bigger firms to provide great legal services against the resources of the government.