We best be careful, or we will find ourselves in a closet talking to ourselves too much.

Once upon a time, I had a good client who was fond of saying, “We best be careful, or we will find ourselves in a closet talking to ourselves too much.” Meaning, of course, that you get into trouble if you don’t get diverse viewpoints from people who perhaps see the problem and the world differently than you.

My client’s wisdom was recently brought home to me in connection with the Gen AI hoopla. The last two months have been a whirlwind of conferences for me. During that time, I attended three technology conferences. One was CES, which was generally directed toward consumer electronics and technology. The other two, LegalWeek and ABA TechShow, were both directed at legal technology in particular. 

At all three conferences, there was considerable discussion of Gen AI and the predicted revolutionary changes it portends. At LegalWeek and TechShow, the clear message was that Gen AI will disrupt legal as nothing before and that lawyers and law firms are rushing to adopt and use the technology like nothing ever before. 

After these three conferences, my view of the world was that Gen AI was revolutionizing the legal world and that lawyers were constantly coming up with new ways to use Gen AI tools in their everyday practice. This worldview was seemingly confirmed by a plethora of vendor studies.

A Dose of Reality

So when I recently attended a conference of top trial lawyers (and only lawyers), I fully expected the same: lawyers gushing over what Gen AI could do and how they were using it every day. After all, these are the soldiers with boots on the ground, lawyers who actually practice and try cases. If anyone were going to be adopters of this revolutionary technology, it would be these guys. Or so I thought.

Most of them were genuinely frightened about using Gen AI

Boy, I was in for a dose of reality. Most of the lawyers I talked to at the conference had not adopted Gen AI tools. Most had little familiarity with what it could do, how it works, and how they could be affected by it. Most of them were genuinely frightened about using Gen AI, a fear fed by the publicity of hallucinations and lawyers being sanctioned for citing cases that didn’t exist that Chat GPT provided. 

Many of the firms of which these lawyers were members had banned the use of Gen AI with little analysis. Few firms had offered training about how to use Gen AI or even what large language models were. Very few firms had invested in closed systems that promised to reduce the hallucinations about which they were most concerned. Indeed, many of the lawyers didn’t understand what closed systems were and why they might be more trustworthy than public models. Augmented retrieval generation was not a term with which many had any familiarity.

These lawyers were genuinely concerned that their clients might demand that they use Gen AI tools, and how that could interfere with and detract from how they served their clients. When a speaker asserted that Gen AI tools could disrupt the billable hour and leverage models, he was met with blank stares and even by some with outright disbelief. The lawyers don’t yet accept the notion that their competitors or adversaries might adopt Gen AI tools and gain an advantage. There seemed to be little acceptance of the notion that change would occur any time soon.

Instead, the lawyers focused on the hallucination risks. They also saw judges issuing orders requiring certification of use and review. To the lawyers, these orders further auger against the use of Gen AI for a host of reasons, not the least of which was that certifying the use of Gen AI to many judges might prejudice the lawyers’ overall arguments. Finally,  the notion seemed to be that if a GenAi tool were used, the time to check its accuracy would end up being more than the time saved by its use. The fear of Gen AI was palpable.

Why the Disconnect?

Clearly, there was a disconnect between what I had been hearing from the tech community and what I was hearing from potential users, the lawyers. Part of the reason for this disconnect was that the three conferences I previously attended were primarily platforms for vendors promoting Gen AI products. These conferences were also showcases for speakers who, by definition, were some of the most tech savvy in their respective areas of expertise. 

Perhaps part of the reason is that so many surveys suggest widespread adoption of Gen AI may be that the vendors doing the survey inadvertently started from the notion that Gen Ai was already accepted and in widespread use. Or perhaps those who responded to the surveys were self-selected early adopters, didn’t want to sound behind the times, or just didn’t understand enough to accurately respond.

Perhaps the disconnect is because the lawyers I encountered at the most recent conference were curmudgeons who consistently resist change and the use of technology. 

Maybe, but I don’t think so. Trial lawyers, especially good ones, must be familiar with technology and how to use it to persuade effectively. I don’t think this group of lawyers at the conference I attended was uniquely behind the times.

What About Lawyers?

I do think, though, that adoption of Gen AI tools by lawyers will take longer than what many pundits and vendors like to think. The lawyers I talked to are cautious; they understand the need for accuracy better than most. They understand that your credibility in front of judges and juries depends on this accuracy. They know and understand the ethical rules that require the vigilant protection of information about the representation of their clients. They are concerned about being sure they represent their clients competently and professionally.

They are right to proceed with care with a technology that can do so much but yet may pose threats to their clients

So they are right to proceed with care with a technology that can do so much but yet may pose threats to their clients, threats some of which are not yet known or understood. They are right to take a wait and see attitude as this technology unfolds and develops. Add to this a skepticism borne of having all too often seen legal technology overhyped and underperform.  This skepticism is heightened by the fact that good lawyers don’t have the time to do the in-depth investigation that many vendors and consultants do. Nor is there any trusted source or consensus yet available on use cases and validity. 

And I get the sense they often don’t appreciate the condescending attitude of those who don’t practice and who travel unencumbered by clients and ethics telling them how behind the times they are. Lawyers know BS when they hear it and they feel like they are hearing a lot right now.

The Good News

To paraphrase Bill Gates, we overestimate short-term change but underestimate long-term change. I’m not selling these lawyers short when it comes to long-term change. I have witnessed them make changes over and over again. They will get there with Gen AI, but it will take time. Let’s not get ahead of ourselves.

In the meantime, the rest of us need to, well, stay out of the closet.

Photo attribution: Photo by Darwin Vegher on Unsplash