Start Cheap and Easy  

Tips For Getting Comfortable With AI — Because It’s Here To Stay  

I do a lot of speaking at CLE programs and other seminars, mostly on legal technology and practice management. As of this writing, I am in the midst of a CLE series for Lawyers Mutual Liability that is taking me across North Carolina giving the same program seven times in different locations. Most of the time when I teach a CLE, it is a one-off, so it’s an interesting change to give the same program multiple times. 

There are some upsides to doing it this way: I get more comfortable with the material, my delivery improves and I also learn which of the technology tips I have prepared resonate with the audience and which do not. One of my tips this year is about generative AI. When I get to the slide for that tip and start to introduce it, I can feel the air leaving the room like a deflating balloon. Eyes roll. Focus shifts to phones. Everybody is a bit tired of hearing about, reading about and thinking about AI.  

I get that. I am a little fatigued with it myself and I have to pay attention to this stuff for a living. It’s absorbing a ton of our collective attention, and we end up whipsawing back and forth between the breathless boosterism and groundless pessimism. It makes it hard to calibrate our reaction (or non-reaction) to this latest technological advance. Is AI just the latest incarnation of whatever current tech tulip mania is in vogue (RIP FTX Arena), or is this truly an important technological change that lawyers will be well-served to begin to understand? 

The answer, I think, is both.  

We are absolutely in the midst of an AI gold rush, with every legal technology company that can figure out an AI angle for their product pivoting hard to capitalize on the moment. It is no wonder why: venture capital investment in AI went from 4% to 21% of total VC investment between 2012 and 2020, and ChatGPT has done nothing but pour jet fuel on the fire over the past 12 months. 

The pace of change makes it tough to separate the wheat from the chaff; to figure out which parts of this are worth some time in a lawyer’s strained calendar and available brain space, and which are, well, mere puffery. 

Despite the over-the-top frenzy for all things AI, underneath it all, the technology (I’m referring here to the large language model chatbots, such as ChatGPT, Google’s Bard/Gemini, and Microsoft’s Bing), is highly likely to have an outsized impact on law practice, among many other professions, in the coming years. There is a tendency to see coming change as threatening, but for the moment, let’s try to focus on the opportunity side of the equation: some things are going to be possible as AI models improve that are not possible today. The law firms that figure out early what these AI tools do well (and, importantly, what they don’t do well) are going to be in the best position to exploit those opportunities. 

I think the best way to think about where we are with respect to AI is one I picked up from a panel discussion at a conference: AI right now is like the rise of the internet in law practice in the 1990s. We’re at the beginning of a curve that is creating new ways for lawyers to find and do work, and we don’t yet have a line of sight to where that curve ends.

If you’re, ahem, seasoned enough to remember practicing law in the 1990s, you will remember watching as some firms embraced the new technology immediately and others proclaimed it a fad. Thirty years later, and the need for law firm websites and email, to name just a couple of changes, is no longer debated. They are just part of our ecosystem. 

All of which is to say, it is worth some fraction of your time and energy to start to understand AI and the role it will play in your practice.  

I know that approaching AI, especially for the less technologically oriented, can be daunting. But it can be done slowly, methodically and without meaningful risk if you start with just a few basic steps. 

Start Small (and Cheap) 

Law firms often adopt technology in big, expensive chunks of investment. Buying practice management software, developing a new website, etc. With AI, I’d recommend instead that law firms start small, very small. And cheap, very cheap. There’s just too much potential to waste your money and the opportunity otherwise. 

One starting point I recommend for skeptical but AI-curious lawyers, is to start using generative AI in their internet searches. Whether Google or Microsoft’s Bing, it is now easy and free. Both search engines place pop-ups on their main search pages that allow users to try AI in their searches – all you have to do is click the button to indicate you’d like to try it for a given search. 

Next time you are going to search for something, try an experiment: run the search conventionally, and then run it using the generative AI support and compare the two results. Observe the differences between the results. Keep a couple of notes of what you saw. Repeat that for your next 10 searches, and then see what themes emerge. This is the absolute simplest way to dip your toe in the water of using AI to see what value it can add in your life. 

Start Internal 

If your search engine experiment leaves you curious about next steps, you can start kicking the tires on using the AI chatbots to generate some ideas and text. 

I usually recommend that firms start with low stakes, internal-focused projects. Nothing mission critical, nothing you have to sign your bar number to. Think of using AI like learning to use a chainsaw (or how I would imagine one learns to use a chainsaw, anyway, I can barely use a screwdriver): start with the small, easy cuts where you are unlikely to, you know, cut your leg off or whatever. 

One of my favorite ways to use AI is as an aid in brainstorming: developing lists of interview questions, newsletter topics, etc. Most of the time, out of 10 ideas brainstormed by the AI, I will have already thought of 8 of them, but two will be things I hadn’t considered, and one of those two will be really good.  

The key here in the early going is to use generative AI for things where you can instantly spot places where the AI has returned results that are problematic (hallucinations, as they call them). In a list of interview questions, there’s no chance you are going to accidentally ask a candidate an insane question generated by an AI; you will be able to immediately and effectively sort the lists for the best options. 

As you get more fluent in your use of AI, you can start to ratchet up the complexity and importance of the projects you use it for. But what you’re doing first, is developing your competence and confidence in using these tools safely, ethically, and without undue risk. 

Become a Prompt Wizard 

One of the primary competencies you will be developing as you learn to use AI effectively, is to get better at asking it useful prompts. The quality of your prompts more than any other variable will determine whether you get value from AI in your firm. AI will reliably get better at answering questions over time, but there is no substitute for you getting better about the prompts you feed it. 

Think of AI of as a very bright but not very experienced law clerk. If you are vague in your requests, the work product you get in return is likely to be uneven. The more specific you are in delegating the work, the more likely you are to receive something back that matches what you wanted.  

You can experiment with many different prompts to find what works best for you with various projects. You will have graduated to ninja level AI user (well, if not ninja, at least 201 level) when you start asking the AI a prompt to help you write better prompts. And you will be at the tip of the iceberg. 

Conclusion 

Can AI be ignored at this moment in time? Sure. But I hope this article has convinced you that there is enough opportunity there to make it worth at least a little sliver of your time to start figuring it out and thinking about what these new tools might allow you to do in your firm.