“60 Tips” Rapid-Fire CLEs – Slow Death or New Potential?

Episode Notes

Transcript

Announcer:

Web 2.0. Innovation, trend, collaboration. Got the world turning as fast as it can? Hear how technology can help, legally speaking, with two of the top legal technology experts, authors and lawyers, Dennis Kennedy and Tom Mighell. Welcome to the Kennedy Mighell Report here on the Legal Tok Network.

Dennis Kennedy:

And welcome to episode 416 of the Kennedy Mighell Report. I’m Dennis Kennedy in Ann Arbor.

Tom Mighell:

And I’m Tom Mighell in Dallas.

Dennis Kennedy:

In our last episode, we talked about the AI projects Tom and I are actually working on right now for ourselves in our own work and where AI is helping us most now. We might have to do a show coming up on where AI is not helping us most, but last week you definitely heard where it’s helping us. The spoiler alert for me, it’s absolutely not legal research. So anyway, that show is highly recommended in case you missed it. In this episode, Tom and I are going to take a hard look at one of the longtime staples of legal technology education, the classic 60 Tips in 60 Minutes Session and what that means in 2026 and beyond. Tom, what’s on our agenda for this episode?

Tom Mighell:

Well, Dennis, in this edition of the Kennedy Mall Report, we are going to compare notes on whether the traditional legal tech CLE format of rapid fire tips has outlived its usefulness. We’re going to debate whether the rise of AI and the fact that lawyers are all using different versions of different products makes the 60 tips approach more confusing than helpful. In our second segment, we’re going to move to a question from one of our show’s super fans about the risks of agentic AI. And as usual, we’ll finish up with our parting shots, that one tip website or observation that you can start to use the second that this podcast is over. But first up, Dennis, it says in the script that I thought it would be useful, but I think it’s really you thought it would be useful to look at an old legal tech topic with fresh eyes.

And what raised this was that I just finished a 60 tips session for the Virginia State Bar. They have a Virginia State Bar tech show. I’ve been going to it for years and since COVID, it’s all been virtual. It used to be in person. I love going to see my legal tech friends there, and it’s one of the things that I’ve enjoyed. And people still seem to love these presentations. But Dennis, you’re asking the question, is it still working? Between AI changing the game, gap between early adopters and those on legacy software, does this approach still make sense? I feel like we have a pardon the interruption episode coming up, but I’m going to throw it to you first, Dennis. What is your take on the current state of the 60 tip phenomenon? Although I’m just going to ask you, how long has it really been since you either presented or attended a 60 tip session?

Dennis Kennedy:

I would think it’s approximately 350 years. No, I mean, it’s been a long time. So it surprised me, Tom, I thought about this topic was that you told me that you were working on preparing a 60 tips presentation. And I started to think, how does that make sense these days, given kind of the pace of change, what’s happening and the wide gaps between what people do? So I don’t know, I would say it’s been at least 10 years since I’ve given one or attended one. And I usually go there just to hang out with my friends anyway, but I just always found it difficult to present these kinds of sessions and to attend them, frankly. They’re fun. I mean, they can be fun, but for me, and I expect for you, it’s actually pretty rare to come up with anything that really useful, especially as you kind of suffer through the jokey portions of the 60 Tips program.

So I’ll stipulate right away. I’m not the right audience for these things, but I do want to hear about your recent experience at the Virginia Bar and maybe give us a description of what a 60 tip sessions is and how this approach got started.

Tom Mighell:

Yeah. So quick history lesson on 60 Tips, although I’m not sure I know the full history, but this is something that has been going on since I’ve been doing legal technology. So the early 2000s, if not earlier, and I think frankly at ABA Tech Show, it was going on for much, much earlier than that. But the concept is pretty simple. One tip per minute for 60 minutes, 60 tips. It is rapid fire. It’s designed to be entertaining, but also useful to the audience, but it requires that people, that the presenters have the ability to both deliver that information quickly and also have something useful to give. For me, when I’m preparing for a 60 tips presentation, I’ve really got a built-in shortcut for that. And that is this podcast. That is the one tip website or observation you can use the second this podcast is over.

They become my tips. I really do only one 60 minutes, 60 tips session a year. And the first thing I do is I go back to all of our parting shots. I pick up the ones that are worthy of a conference audience and that becomes the spine of my list. From there, I supplement. I like to put in Microsoft 365 because that’s where I live on a daily basis and that’s where I think most of the attendees live. And then there’s always something … This year, like everybody else, the list skewed very heavily towards AI. And I think it’s hard to do a 2026 session that doesn’t include AI in some part. I then, frankly, turned over the whole list over to Claude to build the slides. And it was just about the best slide building experience I’ve ever had, better graphics, more creative layouts, more visual variety than I could come up with, and it took a fraction of the time than me having to think it out and think about in creative stuff because one of the complaints I do have about 60 Tips presentation is that some tipsters will just throw up some bullet points and that’s all they’ll do.

And it’s really just not very interesting to look at. And I think if you’re going to give an interesting presentation, it ought to look interesting too. If you’re presenting and you’re not using AI to build your decks yet, there’s your tip right there. The session itself, I think was great. I like presenting with these folks. They’re friends, they’re fun, they’re curious about technology, they actually want to be there. And it was at the end of a long day of CLE. So we’re hopefully sending people out on a high note. Downside was, and I guess the irony and what we’re talking about today is we didn’t actually finish. We only got through maybe 48 tips total. We were all spending too long on our tips and all of a sudden they just decided to end it. And so my best tips were at the very end, so I didn’t even get to do my tips.

But one of the rules of 60 tips in 60 minutes is you actually have to do 60 tips in 60 minutes. The format demands that discipline and we definitely failed at that. But all in all, it was a good experience. And I’m going to kind of talk more about why I think it was a good experience. But my overall impression, I still enjoy doing it because I like to talk about legal technology to people.

Dennis Kennedy:

And I think you hit on a couple of things why it became this staple of technology conferences over the years, and especially ABA Tech Show, is that when you have a multiple day event focused on technology, you can wear people out really quickly. And a lot of stuff goes over their head and they don’t know what to do first. And sometimes they come in feeling that they know no less leaving than when they came in, although that’s typically not the case. It just sort of feels like they’re drinking from a fire hose. And so the idea was the 60 tips, you could do something where everybody could leave the room and with one or two or three tips that they could go implement right away and it would be helpful to them. And even if there was stuff that was not helpful, then it was over in a minute and you went on to the next thing.

And as you had this great gap between lawyers who knew nothing about technology and lawyers who knew a lot, this was the way to kind of bridge that gap. And it could work really well, but I think the gaps have come greater. And so I would say what kind of soured me on the approach is that, and it’s not just because people don’t invite me to do it anymore, but what soured me on it is that I sort of felt it was like one size fits none, that this gap between the power user and people who didn’t know much of anything made it really difficult to come up with tips. And so if you would do a tip and you’d go like, “Oh, here’s a great thing in the current version of Windows and everybody in the room or the majority of the room were two versions behind in Windows.” Or you’re talking about, “Here’s a great thing you can do in Microsoft Word with Copilot and people are on Word 2016,” the tips don’t work very well.

And so that I think was a stressor that the tips thing was designed to overcome, but actually I think that it kind of emphasizes that gap even more so. And then as you said, Tom, with AI, the AI features are changing so fast that if you had me do a set of slides today to deliver on Friday, I could have a tip that I did today that was obsolete, or at least to me, by the time the audience actually watches the CliE. So those were always the things that I struggle with plus the traditional thing that you did is that in the typical case, you never got all 60 tips.

Tom Mighell:

So I think that when we talk about the one size fits none problem, I think that you’re pointing at something real, but it’s maybe labeling it in a way that I wouldn’t label it. The version fragmentation problem is not a format of the, it’s not a problem with the format, it’s a curation of the tip. A good presenter picks a tip that travels. If a tip is version specific, you either skip it or you use it to pitch and upgrade. If you’re on the old version, here’s your reason to move to the new version. If you’re not using Copilot, here’s why getting to Copilot’s going to be great. I would make the argument that the average tips audience is probably skewed 80 to 90% low knowledge about technology and 10% Power user. So I’d still argue you’re getting most of your bang for your buck.

I think though that the job is to find the mix. Tips simple enough that beginners can use them tomorrow and fresh enough that the power users haven’t seen them. It’s trying to figure out what’s the balance. It’s hard, but it’s not impossible. And then I think that talking about AI product for fragmentation, that’s, I think, an interesting thought. Now everybody’s split across Claude and GPT and Gemini and Copilot, a third of them aren’t using anything. I think that’s harder to design for than version fragmentation, but even still, that’s a curation problem. You pick tips that work across tools or you pick concepts that can translate. Don’t disagree. AI tools are changing and that was an issue that we had in our session. I had a tip about Claude Cowork and it wasn’t that it had changed. It’s that in the meantime, since we had done our, and I’m a little skeptical that in two days worth of time, something would change meaningfully with an AI that would change a tip.

But anyway, it had changed for us. It had been about a month since we created our deck for the program. And during that time, Microsoft issued its own version of cowork that it was licensing from Claude. But I used that as an opportunity to bring that up during Slide. I talked about it. I updated it in real time. So I think that’s really a problem with either recorded CLE or doing your slides ahead of time.

Dennis Kennedy:

I think back to the people I associate with creating the 60 tips, the 60 gadgets, the 60 websites type of approaches, which you can point to Ross Codner definitely is probably if you were going to pick one person. But I think that what happens is that, and you talked about this, that the sessions often prioritize the speaker’s speed and their presentation ability and flashy outputs and coming up with these really unusual things. And so I think we did a podcast once where we talked about tips that we wanted to share with the audience, but our requirement was that they were things that we actually used. Because you go to 60 tips things and somebody go like, “Oh, here’s this thing that just came out and I don’t even own it yet, but I think it’s going to be really cool.” And you go like-

Tom Mighell:

Useless tip.

Dennis Kennedy:

“So what? ” So I think that it puts a premium on performance and so it can be great, especially the Saturday morning of tech show, you’re tired. So having something fun is a great thing. I’m not against fun.

And then I think there is this notion that even if you can’t use 57 of the tips, finding two or three makes the hour worth it. And you can find one thing. I noticed something on my desk today or in a desk drawer today that came from a 60 tips session that is not quite as useful as the presenter said it would be, but it’s definitely something that I found useful and still do. So I think it’s good for that. And then possibly even better is it gives you a map of the landscape of what’s possible out there. So if you hear about something you had no idea it existed or you could do these things, I think that’s great. And then that gives you some new ideas that you might not have considered. And then maybe it’s just inspiring, which I think is a plus as well.

You would say like, “Oh, I was hoping to do something.” And that’s what we do with our parting shots too. It’s like, “I didn’t know you could do this. So I’m inspired now to track this down and do something with it.

Tom Mighell:

” I agree with all of that. I think that if a lawyer walks out with two or three tips, like you said, that they’ll actually use the hour pace for itself for them. Nobody implements 60 tips. That’s never going to happen for anybody and they were never supposed to do that. I think that when you talk about the map of the possible, a 60 tip session is a fast tour of a landscape that most lawyers don’t have time to survey on their own. They don’t have time to go find it all on their own. They might not use the tip today, but it’s going to lodge somewhere. It’s going to sit there and six months later they’re going to remember it when they hit the problem. I think that where you talk about inspiration, the value really isn’t in the specific button click, but it’s seeing somebody else solve a problem that you didn’t know was solvable, that reframes how you think about how your own workflow.

I think that my AI tips that I used in my presentation was to bring up use cases that the viewers might not have thought about on their own. They had just gotten an entire day of how to use AI to do legal things and how to research and how to do litigation and all that sort of stuff. And I said, “Why not just use it to get control of your email or to plan your day or to summarize a meeting or to be more productive?” And so that’s where I think the inspiration comes in.

Dennis Kennedy:

Tom, it’s interesting why you’re saying that is that, and this sort of illustrates all facets of this, is that I think the last time I did 60 tips, I talked about the designer tool in PowerPoint, and I think that’s a really great thing because you say like you spend a lot of time just like trying to get stuff formatted onto PowerPoint slide. And the designer tool in there actually does a really good job and gives you a number of choices and it looks really good. And you could be in a 62 session and you go like, “Oh my God, this is amazing. I learned this great thing.” And then you go back to your firm and you go like, “Oh, designer, we’re using a version of PowerPoint that that doesn’t even exist in. ” And then I think that brings us to, I think, a topic for another show, but we revisit it from time to time, which is what happens when the, as a lawyer, the technology I’m using at home is way better than what I’m using in the office, but that’s the topic for another day.

And I think we need to take a break, don’t we, Tom?

Tom Mighell:

It is, but I just want to say, that’s why the tipsters should say this is available to anybody who has this version, which will hopefully inspire that attendee to go back to their firm and say, “You know what? We need to improve on that. ” Whether they’ll be successful is another question, but plant the seed and maybe they’ll be inspired to go and get their IT department to improve the software they use. I know pie in the sky, but we can all hope. All right, we need to take a quick break. We’ll be back in a little bit, but we need to take a break now from a quick word from our sponsors. And now let’s get back to the Kennedy Mall Report. I’m Tom Mighell.

Dennis Kennedy:

And I’m Teddy Kennedy. We wanted to remind you to share the podcast with a friend or two that helps us out. Tom, I have a radical proposition here and it’s based on your comment before the break about one inspiration. So I think tips are great and I think that to me, one of my objections, they’re sort of low engagement. So I feel like, “Okay, I’ve spent this money on a conference and they’re giving me tips.” Is that enough of a takeaway? So what if instead of 60 tips, what if we said 60 annoyances to alleviate or 60 pains to eliminate where we’re saying like, okay, I’m walking away with something where I have actual irritations and you’re not giving me pie in the sky stuff of things that I can’t, that I don’t have the money to buy or they don’t make sense for me.

But what if we just identify the stuff that is like a real pain and the tips session becomes about getting rid of those? And maybe instead of two or three things, I walk away, maybe I walk away with 10, 12, 15 things that make a difference the next Monday I’m back in the office.

Tom Mighell:

So I’m open to hearing you out on this. I have opinions about that. So tell me, how would that look in practice for the tipster?

Dennis Kennedy:

Well, I think that it’s going to be like a lot of things that we do these days. And when I look in legal education and elsewhere, it’s going to say to the presenter, you’re going to have to up your game in a completely different way and you’re going to have to consider the audience and you’re going to have to learn about the audience in a new way. And AI will help with that. But I think that we’re kind of doing these tips things because a lot of people, especially back in what I call the real heyday of 60 tips, people did the same tips from show to show. But I think what becomes interesting is if we use, and in my case I would use AI to say like, what are the things that lawyers are really concerned about? Can we get some survey information from the people putting on the shows and we can kind of attune what we’re doing to what people care about and plus there’s all kinds of information out there about what lawyers care about, tons of surveys and stuff.

So maybe as a presenter, I can say like, “Let’s identify that stuff that matters.” Plus as a presenter, I’m going to go, “Man, there’s like a hundred annoyances I have I would like to eliminate. So let me just dig into my own stack.” So Tom, I guess in the classic word of these days, I’m looking at a pivot and saying, let’s stop looking for solutions in search of a problem like, “Oh, here’s this gadget that no one might not have even about yet or like is in some kickstarter thing that nobody knows about or something that says, I’m hoping to buy this one day.” But let’s say, instead of saying solutions and search for problem, let’s start with the actual friction that lawyers feel every day and say like, “Okay, so let’s identify those things and then suggest actual things that we can do.

Tom Mighell:

” I mean, I agree with you, but I think my argument would be that the pivot is really the traditional format done the way that we should have been doing it the whole time. And that really the argument is not with the way the tips have gone, but with the way that people have pivoted themselves to giving something flashy or talking about something fun. My take is this, is that if you’re doing 60 tips right, you’re already alleviating annoyances, you’re already eliminating pain. Every tip should in some way solve a problem that the audience may not even know that they’ve had, that they’ve got. So whether we call it 60 tips or 60 annoyances, to me, it’s a rebrand of what a good tip session is already supposed to be doing. So just examples from my presentation, I gave a meeting, I gave a tip on using smart meetings in Microsoft 365, transcripts, facilitator, all of those things.

Yes, I did need to tell people you need to have Copilot to do this and Copilot is really useful, but I didn’t view it as much a feature tip as solving a problem of bad meetings and the cognitive load of taking notes while also participating and being able to take that away. When I talk about, my one gadget tip that I gave was on the new remarkable Paper Pro move, the little small one that fits in the palm of your hand. To me, that solves the problem of walking into a conference room or sitting in front of a client and not having a good way to take notes without dragging a laptop into the conversation. So that’s a way to solve that. I gave a tip on the fact that I, and I’m assuming everybody is receiving lots of emails or text messages that say, “Click here to reset your password.

We got a password reset request from you and we’re all getting it and it offered tips on how to protect yourself when that happened.” So I think that solves a problem that people are actively experiencing. They just don’t know that there’s a clean answer to it or they don’t know what the actual steps are to deal with it. So none of those, I think are, here’s a cool button to use in Word that you may never have used before. I think they’re all problems first and tools second. At least that’s how they should be addressed. Whether good tips presenters are doing that, I think is the real issue.

Dennis Kennedy:

And as you’re talking, the other thing that I’m now seeing is that you typically have a panel of people who are incredibly knowledgeable and they’re churning through like one minute a tip. And in the audience, you don’t have a chance to ask questions or anything. It just kind of moves on. And I’d almost like to say like, let’s just like stump the panel. It’s like, “Hey, let’s get questions from the audience and just have the panelists answer them.” We’re never going to get to 60, right? Knowing that people are panelists, we’ll probably get to, as the classic Jim Calloway comment was the one time when there were technical problems, welcome to six tips in 60 minutes, but it could be really interesting where if you say like, “Oh, we have this high powered panel. Let’s have the audience ask them questions of like what’s going on and do some follow up.” And that would be great.

And then also the thing I always miss, and I can tell that you do this, Tom, is to say, “Here’s something that you run into and if you’re on Windows, here’s what you would do. ” And if you’re on a Mac, it’s similar, but it’s a little bit different, but you could still do the same things. If you’re in ChatGPT, this works a little bit differently than in Claude, here’s the reason for that. And you just go ahead and do that because otherwise if you’re a Windows person, somebody says, “Here’s, I have three great Mac tips, they don’t focus.” So I see those things and then I think there’s potentially this great 80 / 20 rule if we can in a session identify 20% of the annoyances that cause 80% of the headaches, that’d be awesome.

Tom Mighell:

Yeah. I would say I love your idea of stump the panel, but I would say that’s a different, you’re just advocating strictly for the elimination of 60 tips in its entirety by doing that. Why can’t we have both? I mean, to the Virginia State Bar’s credit, they did. They had during the lunch hour, the viewers were invited. Every one of the speakers of the conference were on the call. I wasn’t able to make it. Everybody who could make it, but during the lunch hour was on the call just to answer questions. And it was a virtual stump the panel. It was very similar. I think it’s a great idea. I think it exists separate and apart from a 60 tips session. Totally agree on being version agnostic. I think that the whole point of a good tip session is to make sure that as much of the room as possible can use as much of the advice as possible.

And that was why I had a tip about finally the fact that iPhone and Android can send documents to each other and can use AirDrop or Quicksave with that. And I gave tips on setting up AI tools that worked no matter whether you were using Claude or Copilot or Gemini. I wasn’t giving anything specific because I agree. If they say, “I’ve got these three great Mac tips,” I’m tuning those out because I’m not a Mac person and that just won’t work. So I totally agree with that. I’m not opposed to a rebrand. I think that 60 annoyances is probably a sharper marketing pitch than 60 tips. I think that tips is not right, but I think that by doing that, we’re putting a label on the discipline that should already be there, that should have been there in the first place. And maybe that’s a valuable thing.

Maybe naming it that way forces speakers to start with the pain instead of the flash and sparkle.

Dennis Kennedy:

And I also was thinking about how when you and I and our friends are looking for tips or we have our own questions, that what we want is a little bit different. So you could say, remember a couple months ago, like everything was whisper flow, like you got to use whisper flow, but really what you were saying is like, “Hey, I would like to do dictation and voice recognition.” And I got a bunch of people saying that this is the thing, but there are other ways to do it. And if I can sort of express what it is, I actually, not to be chops to be done all the time, but like once I can figure out what it is that I want, I would go like, “You know what? Whisper Flow isn’t exactly what I want, but this other thing is. ” Or, “I don’t want to go all in on this.

” So if I’m in Mac and you say, “Oh wait, I can just dictate into Apple Notes, I just need to see whether this even makes sense for me. ” So I think some things like that could be helpful, but as you say, I think it’s just this great opportunity that maybe we lost a little bit, but you identified here in the post COVID era is that this sort of open Q&A online where people can submit questions in chat and stuff like that is actually a great format.

Tom Mighell:

And I think we’ve rung about as much out of 60 tips as we can for this episode. We have more to talk about in our second segment, but first we need to take another quick break for a message from our sponsors.

Dennis Kennedy:

And now let’s get back to the Kennedy Bio Report. I’m Dennis Kennedy.

Tom Mighell:

And I’m Tom Mighell. Dennis, before the break, we were rethinking the 60 Tips approach and to help us dig deeper, we’ve got a question from show super fan Jerry Lawson. We’ve played plenty of Jerry’s clips over the years. We’re thinking about bringing him on as a guest soon instead of just playing his clips, but let’s listen to his question right now.

Caller:

This is Jerry Lawson calling from books yet another question. Agent AI is definitely a hot topic today, but I can see some problems with it. It is very powerful, but it seems to me like it’s also dangerous. It’s the very potential to give you so many benefits that also makes it dangerous. People who get a little benefit from it would probably be tempted to try to get even more benefit and this could lead them to give the AIM too much free reign. And for example, I saw recently where some people have been recommending that you give an Angentic AI app access to one password together. What is your thinking while the security risks of Angentic AI? Thanks,

Tom Mighell:

Mike. Okay. So Jerry is asking a question about a Agentic AI, where the AI just isn’t answering questions for you, but is also taking actions on your behalf. And what are the potential issues? Security issues, ethical issues, other risks with giving a computerized tool called an agent, more autonomy over the technology that you live with every day. I think, Jerry, that’s a great question. And I think that you are putting your finger on the right tension here. Agentic AI has enormous upside, I think. But the second that you let an AI take actions on your behalf instead of just answering questions, the risk profile changes completely. This past week, as we’re recording this, there’s a story on the news about a software company that used AI agents and it completely wiped away its entire online development environment. So people lost access to an app that they were using and it did it in less than nine seconds without anybody knowing.

I think there are tremendous problems with the security that it gave these agents, but the fact is it can happen. So it’s something we need to think about. And let me start with where I am personally, because I think it grounds a little bit of what I want to talk about. I’m using a lot of agentic tools. I’m using Claude Code to build … We talked about this in the last episode, to build an application. I’m using Claude and Chrome to go out and look at certain websites and things, but I’m directing what those websites are. Claude for Excel. I’m using Claude Cowork. We talked about that. So I’m not coming at this as a skeptic from the outside. I find real value in these tools, but I’m also pretty conservative about what I let them touch. Right now, I’m giving Claude access to only a handful of specific areas on my computer.

I’ve connected it to my Gmail and Google Drive. Okay, fine, but I have limits on that. I give it limits on what it can do. It can read, it can’t delete, it can’t send. And do I 100% trust it to honor these limits? Honestly, I don’t know. And that’s part of why I’m trying to keep my quote perimeter small on this. It’s interesting in your message that you mentioned one password because of all the companies, one password’s actually taking and thinking a lot about agentic AI because it realizes that people are going to want to be able to do it. Here are some features you may not know about that one password is rolling out. They’re starting something called agentic auto fill, where the agents are going to use a secure encrypted channel to fill credentials in a browser, meaning that the agent is never going to handle or store the username or password.

There is a human in the loop aspect where the user can receive a prompt to approve or deny an agent’s request to sign in that ensures that an agent’s not going to access credentials on their own, especially for sensitive things. There’s something called just in time credential delivery, which means that credentials are only provided to agents only when required for the task, and they’re never hardcoded, which means that they’re less likely to be stolen or leaked. And then if you’re familiar with the security concept of least privilege, they now have what is called scoped access, which means that one password allows you to limit an AI agent’s access to only specific vaults in your one password. So you can basically block it off from anything you don’t want it to have access to and only give it the passwords that you need to. Because in order to make agentic AI work well, sometimes you’re going to want to have it log into specific sites, but only ones that probably present less issues for you.

But I think frankly, the bigger question, what are the real security risks have? I think let me hit a few very quick security risks. There’s prompt injection that’s at the top of the list.This is the attack where someone, it goes to Claude and Chrome goes to a website and there’s instructions there that the agent reads and it follows the instructions instead of my instructions. That’s not theoretical. That’s happening right now. If you’re going to use the tools, you need to know that that risk exists in doing it. There is a concept known as over permissioning, which is giving the agent more access than the task needs. And when it comes to lawyers, you have to be careful about this because if you connect an agent to your email and your email has client matters in it, that agent is now in your client files. And rule 1.6 didn’t go away because you’re using Claude.

You need to pay attention to that. There’s another one, irreversible actions. The agent sending an email or buying something or deleting a file. We used to get the, are you sure message or please press here to confirm that prompt you, that protect you for 30 years. It’s gone. So you need to find a way to design that back in and keep the human loop. And I think that the one thing that’s also we need to pay attention to is that there’s an ethics dimension here that doesn’t probably get talked about enough, or at least maybe I don’t see it getting talked about enough. That’s rule 5.3 of the model rules of professional conduct. Rule 5.3 says that you are responsible for supervising non-lawyer assistance. And I believe that there is precedent out there to say that that includes AI. That is non-lawyer assistance. So if your agent does something on your behalf, sends a message to opposing counsel, accesses a document, fills out a form, you’re on the hook for it.

Can you actually supervise an autonomous agent? That’s a great question. I am not sure, but you want to make sure that … I don’t think that the bar has fully really understood that or grappled with it yet. So my answer, Jerry, to you in one sentence is yes, you can use these tools safely. And yes, the risks are real and both of those things can be true at the same time. So I’ve been talking a lot. I had a lot to say. I’m going to shut up. Dennis, your thoughts.

Dennis Kennedy:

I just find the whole area of agentic AI in April 2026 to be incredibly problematic. I don’t even see how it can work. I think people who think they can use agentic AI to save time and do things efficiently don’t understand that they’re probably going to spend more time trying to supervise them, work with them, figure out what they’re doing than any amount of savings that they’re getting. And I just don’t think that the legal profession especially understands that. I mean, I joke around that law schools should go back to just teaching a required class in agency law, because I think this is going to be a big thing. But I just see these taking so much work and anything I’ve done that’s in the nature of Agentic AI is just been spectacular failures. So I use this example because this is agentic in concept, certainly, where I have AI go out, ChatGPT go out and scour the web news sources and give me a summary of the news that’s important for the day.

And this weekend, ChatGPT and I agreed that it had reached the point of total failure. It wasn’t even going to the web. It was telling me it’d done things that it hadn’t done. It had scored like what it had done as it had done 80% of the directions I had given it. And then we kind of reviewed what it had done and I asked it to rescore and I said, “Well, actually it was more like a four out of 10.” And I said, “That’s a fail, right? That’s an F.” And they go, “It is an F.” And I said, “So what should we do? ” It’s like, “We should terminate the experiment.” So I’m just concerned that the security stuff, security is really hard. And if you just say, “I’m going to just have vibe code some things that go out, ” then your attack surface and your exposure has just gone up dramatically.

As Tom says, there’s ethical rules and I just don’t know there is even close to enough reliability and success on these things to make them usable, except in the very smallest instances. And if I’m doing that and I have to have that much supervision, then I think I just need to have software do this stuff and not AI. So I’m bearish, I think would be the word on agentic AI. So Jerry, I think there’s potential out there, but I think the potential is down the road. And boy, there is a lot of work that has to be done before I could see advising anybody to rely on this.

Tom Mighell:

I’m just going to have a small rebuttal to that because I think, Dennis, you shared with me the instructions and the tool that you used for your daily update. And I will tell you, I used Claude to create … I used all three tools. I used Claude, Gemini and ChatGPT to create that. And I will tell you that Claude and Gemini were not good experiences at all. They were very poor experiences, but Claude works almost flawlessly. I have to start it on my own. Well, at the time that I did it, it wasn’t doing scheduled events. It’s now doing scheduled things. And I have a few scheduled things that are firing at the right time and are doing all of the right things for me. But I will tell you that my news thing came back pretty successfully. There were very few failures that I would notice.

The one thing that … And I stopped doing it because the news tended to be … I tended to be getting the same stories over a couple of days. So it was more in the quality of the content and not in the failure of the agent, I would argue. It was that it was not the right thing to ask it to do. So I guess my message here is, and like I said, I’m now using agents to run email checks for me and to look at my email and summarize it for me, to give me a daily review on what my day looks like, to look at my task list, to go out and look at the weather and give me a weather report and to do all that. And I will say it’s working well. I’m not seeing that. So I guess that I’m just going to be the counterpoint to say that yes, there are failures, but there are also successes.

Dennis Kennedy:

And that is the trade off. So I would say it’s absolutely not set and forget, and that if you have something that’s searching for the news in the morning after there’s an attempted assassination on the president and it doesn’t show up in the headlines, there’s a problem with that news agent.

Tom Mighell:

There’s a problem with that. I agree.

Dennis Kennedy:

So we could debate this for a while, but it’s time for our product shots. Tom, that one tip website or observation you can start to use as second this podcast is over. Tom, take it away.

Tom Mighell:

No AI in my parting shot this week. It is about an internet tool. I typically, whenever I am on the road and I connect to hotel wifi and I just think that the speed is miserable, I’ll try to go to an online speed test site and the quality of those speed test tools is questionable. I am now using a tool called Orb. Orb is continuously monitoring the quality of your internet connection on any platform, on any device. It’s an app that you have either on your computer or on your mobile device, doesn’t matter. And you’ve fired up and it will generally let you know how good your speed is, your download, your upload speed, how reliable the connection is, how responsive it is. It’s absolutely free if you’re only using it on like one or two devices. If you want to use it on more, it’s going to cost you some.

But I will tell you, it has become a very reliable tool for me to use to just see how is my connection, how does it happen to be? I’ve been having some questions about how my mesh wifi is in the house, and this has been helping me out a lot to tell when it’s having issues and when it’s actually performing well. So orb.net, the tool is orb. Dennis. I

Dennis Kennedy:

Feel that that would just make me sad.

Tom Mighell:

Well, it’s about having a good connection.

Dennis Kennedy:

Well, we are living in the United States, not a country known for its broadband. So my parting shot is a move I made recently toward precision and better ergonomics. So forget the AI stuff and all these other things. I recently moved away from a track pad to try to get more control and have picked up the Logitech Lyft vertical ergonomic mouse, currently on sale for $59 at Amazon and it’s a vertical mouse and I really like that it gives me both better precision in moving the cursor on the screen. I can scroll better and it just alleviated the end of the day risk strain. So who would have thought the traditional mouse would make a comeback?

Tom Mighell:

And so that wraps it up to this edition of the Kennedy Mall Report.

Dennis Kennedy:

Big thank you to the Legal Talk Network team for producing the show. You can find show notes and transcripts at the Legal Thoght Network website.

Tom Mighell:

If you like what you hear, please subscribe to the podcast in your favorite podcast app and leave us a review.

Dennis Kennedy:

You can also connect with us on LinkedIn with your questions.

Tom Mighell:

So until the next podcast, I’m Tom Mighell.

Dennis Kennedy:

And I’m Dennis Kennedy and you’ve been listening to the Kennedy Mighell Report, a podcast on legal technology within internet focus since 2006.

Announcer:

Thanks for listening to the Kennedy Mighell Report. Check out Dennis and Tom’s book, The Lawyer’s Guide to Collaboration Tools and Technologies, Smart Ways to Work Together from ABA Books or Amazon, and join us every other week for another edition of the Kennedy Mighell Report only on the Legal Talk Network.

Continue Reading