Should lawyers still learn to code in the age of AI?

For years, I have been advocating for getting more lawyers to learn programming. Has this recommendation changed in light of ChatGPT, GPT4, Bard, and other artificial intelligence (AI) tools?

Should lawyers still learn to code in the age of AI?

Originally, I wrote nine reasons why lawyers should learn to code here.

Does my recommendation stand in the age of the almighty ChatGPT?

Why would you learn to code if ChatGPT can give you all the code you might ever need (and then some)?

Because there is a lot more to coding and algorithmic thinking than just the plain syntax and getting from point A to point B.

So absolutely, learning programming is still a great idea for lawyers – and here are seven AI-specific reasons why.

#1 It will help you get better at prompting

Prompting GPT effectively and getting good results takes work. Spend some time with it and you will realise that there is more to it than just texting it like it’s a group chat.

The reason for this is that while your input is text-based, you are still prompting a computer model. This means that it really matters to be clear, succinct, and very explicit.

There are many hacks on how to get better at prompting the AI (see my selection here) – but I believe that getting some coding practice in can help you conceptualise this process beyond plain dos and don’ts.

#2 Learning to code is one of the best ways how to understand what is going on

There is just SO MUCH going on in the AI world right now. And so many new words, such as neural networks, deep learning,… these concepts are hard to grasp and if you want to go beyond empty words at a dinner party, you better get your hands dirty.

Here is a rundown of 10 key things about AI to get you started.

In essence, expressing the fundamental blocks in code yourself can help you understand tech more broadly and give you the confidence to navigate bigger challenges.

Wondering what the key programming concepts are? Check out my Coding Basics for Lawyers series.

#3 We need to understand AI to set its legal base

There is already AI-related regulation popping up, and this is unlikely to stop for the next couple of years. This means a lot of work for AI-savvy lawyers.

Without taking the time to understand it all, we cannot really navigate (or especially draft) any regulation in a productive manner. Understanding its building blocks can help us be more precise.

Read more: Rep. Don Beyer is pursuing a machine learning degree at the age of 72.

#4 You still might have to debug

Let’s do a thought experiment: you actually will just use AI to generate all your code – but you get some code that just does not work.

As an experienced developer, you will know where to look, and what the process of debugging your code looks like. Yes, you can ask GPT to debug it for you, but it can send you down some random rabbit holes.

If you want a visual representation of this, I recommend this video about building a website using Midjourney/Figma/GPT combo. Just watch how the developer uses the GPT – as a tool to enhance, not replace, their process.

#5 You can learn to build something with AI yourself

With the leaps in technology and convenience of OpenAI API, building an application powered by the most powerful large language models so far has never been easier.

Already got some javascript down? Try this free 5 hours long intense OpenAI API course from FreeCodeCamp.

#6 There is a world outside the GPT prompt window

While parts of legal work may be automated by AI sooner than many realise, there is still a lot beyond its envelope.

With some use cases (such as a simple template or process automation), using AI for everything would currently be a huge overkill. Think cutting toast with a chainsaw.

You can make a lot of buzz with much simpler (and cheaper) coding hacks – and explore some of the additional skillsets of modern lawyers.

#7 It is a fantastic thing to do to your brain, your logical and algorithmic thinking

Programming thinking is a fantastic way how to strengthen your cognitive abilities.

The coding part (putting your programming thinking into the proper syntax) may be the less interesting part of it for you. But it is a brilliant way how you can build something and learn a new way of thinking.

Final Provisions

Artificial intelligence has been on the go to thoroughly disrupt the legal profession. But by being proactive, you can focus on leveraging, instead of fearing this transformation.

Building things with code and algorithmic thinking is also incredibly empowering and lots of fun. And, as I argue here, it can be kind of easy for lawyers.

Are you on board? Check out how to start coding as a lawyer here.

What do you think?
Are you going to start to learn to code in 2023?

Baru

By Baru

A blogger and teacher from Big Law with proclivity for computer science and good design.

2 comments

  1. Great post, thanks. That’s actually what I’ve been asking myself lately. What programming language would you recommend to start with? I’m thinking Python.

    1. Hi Lukas and thank you for being here! As to your question, I am a big fan of starting from more low level languages, as is the approach of CS50. It is a little more demanding at the beginning, but it gives you a really deep understanding of the entire landscape and what is happening “under the hood”, and you can later make more informed decision as to what you would like to learn next. Although python is definitely super useful! I documented my decisions in this respect and a TON of resources (including CS50 and python-specific ones) here: https://attorneyatcode.com/how-to-start-coding-as-a-lawyer/ Let me know what you pick! 👀

What do you think?

English

Discover more from Attorney@Code

Subscribe now to keep reading and get access to the full archive.

Continue reading