I Talk to Lawyers about Machine Learning

Question: What is the difference between a lawyer and a mosquito? Answer: One is a blood-sucking parasite and the other is an insect.

OK, everyone loves to hate lawyers, but I actually met some cool ones recently. I was invited to speak to a group of very senior lawyers, at a small mini-conference organized by a large tech company. My talk was “Understanding Deep Neural Networks for the Law Profession”. The audience of about 50 people included senior lawyers from some major law firms, and lawyers who worked for the Legal departments of some large companies, and a few CIOs from law firms.

Let me cut to the chase and say I was pleasantly surprised by the audience. These people were very bright. In particular, they had a knack of quickly identifying and understanding the key ideas I presented, even though most had little technical background.

After first describing some vocabulary (machine learning, artificial intelligence, etc.) I explained in some detail exactly what standard feed-forward neural networks are. They grasped FNNs right away. Then I moved on to explaining various deep neural networks, including regular DNNs, convolutional NNs, and recurrent NNs. I emphasized RNNs (specifically LSTM NNs) because I felt that since RNNs deal with text information, they were probably most relevant for the Legal profession.

I made a few observations that machine learning and AI can apply to Legal in both an internal way and an external way. For example, ML/AI can be used to analyze complex contracts (internal). And ML/AI can be involved in issues such as liability when a self-driving automobile has a catastrophic accident (external).

Moral: I suspect that the Legal profession has traditionally been a late arriver to technology advances. But there are some, such as those at the mini-conference, who are forward-looking and want to be at the forefront of ML/AI. Even if they’re lawyers.

Story: A lawyer, a Jewish rabbi, and a Hindu priest are on a bus that breaks down in the middle of nowhere, in the middle of the night. The three of them walk to a farm house and knock on the door and ask if they can spend the night there. The farmer says, “Sure, but there’s only two beds here so one of you will have to sleep in the barn.”

The rabbi says, “I will volunteer to sleep in the barn,” and he heads towards it. About five minutes later, there’s a knock on the farm house door. It’s the rabbi who says, “I’m sorry but there’s a pig in the barn so I cannot sleep there.” The Hindu priest says, “OK, I will sleep in the barn.”

About five minutes later, there’s a knock on the farm house door. It’s the Hindu priest who says, “I’m sorry but there’s a cow in the barn so I cannot sleep there.” The lawyer says, “OK, I will sleep in the barn.” About five minutes later, there’s a knock on the farm house door. It’s the pig and the cow.

Just kidding.

Advertisements
This entry was posted in Machine Learning. Bookmark the permalink.

One Response to I Talk to Lawyers about Machine Learning

  1. First of all I do no have the intellectual ability to understand the context of your story. However, Hindus in India until 1970s atleast and in some rural areas even now sleep in the barn where they rear their domestic cattle / cow. Hence, your joke invokes no humour.

Comments are closed.