The Geek in Review closes 2025 with Greg Lambert and Marlene Gebauer welcoming back Sarah Glassmeyer and Niki Black for round two of the annual scorecard, equal parts receipts, reality check, and forward look into 2026. The conversation opens with a heartfelt remembrance of Kim Stein, a beloved KM community builder whose generosity showed up in conference dinners, happy hours, and day to day support across vendors and firms. With Kim’s spirit in mind, the panel steps into the year-end ritual: name the surprises, own the misses, and offer a few grounded bets for what comes next.

Last year’s thesis predicted a shift from novelty to utility, yet 2025 felt closer to a rolling hype loop. Glassmeyer frames generative AI as a multi-purpose knife dropped on every desk at once, which left many teams unsure where to start, even when budgets already committed. Black brings the data lens: general-purpose gen AI use surged among lawyers, especially solos and small firms, while law firm adoption rose fast compared with earlier waves such as cloud computing, which crawled for years before pandemic pressure moved the needle. The group also flags a new social dynamic, status-driven tool chasing, plus a quiet trend toward business-tier ChatGPT, Gemini, and Claude as practical options for many matters when price tags for legal-only platforms sit out of reach for smaller shops.

Hallucinations stay on the agenda, with the panel resisting both extremes: doom posts and fan club hype. Glassmeyer recounts a founder’s quip, “hallucinations are a feature, not a bug,” then pivots to an older lesson from KeyCite and Shepard’s training: verification never goes away, and lawyers always owed diligence, even before LLMs. Black adds a cautionary tale from recent sanctions, where a lawyer ran the same research through a stack of tools, creating a telephone effect and a document nobody fully controlled. Lambert notes a bright spot from the past six months: legal research outputs improved as vendors paired vector retrieval with legal hierarchy data, including court relationships and citation treatment, reducing off-target answers even while perfection stays out of reach.

From there, the conversation turns to mashups across the market. Clio’s acquisition of vLex becomes a headline example, raising questions about platform ecosystems, pricing power, and whether law drifts toward an Apple versus Android split. Black predicts integration work across billing, practice management, and research will matter as much as M&A, with general tech giants looming behind the scenes. Glassmeyer cheers broader access for smaller firms, while still warning about consolidation scars from legal publishing history and the risk of feature decay once startups enter corporate layers. The panel lands on a simple preference: interoperability, standards, and clean APIs beat a future where a handful of owners dictate terms.

On governance, Black rejects surveillance fantasies and argues for damage control, strong training, and safe experimentation spaces, since shadow usage already happens on personal devices. Gebauer pushes for clearer value stories, and the guests agree early ROI shows up first in back office workflows, with longer-run upside tied to pricing models, AFAs, and buyer pushback on inflated hours. For staying oriented amid fractured social channels, the crew trades resources: AI Law Librarians, Legal Tech Week, Carolyn Elefant’s how-to posts, Moonshots, Nate B. Jones, plus Ed Zitron’s newsletter for a wider business lens. The crystal ball segment closes with a shared unease around AI finance, a likely shakeout among thinly funded tools, and a reminder to keep the human network strong as 2026 arrives.

Sarah Glassmeyer

Niki Black

Marlene Gebauer

Greg Lambert

 

Transcript

Stephanie Wilkins (00:00)
Hi everyone. I’m here to talk to you for a minute about the market maps that we do here at Legal Tech Hub because we’re going to do a refresh of a lot of them. You may have seen them going around LinkedIn or you’ve seen them in your inbox in a newsletter. But what we’ve been doing since the spring of this year is mapping out the marketplace in terms of AI use in Legal Tech tools. We started with our Generative AI map for legal, which started with 400 solutions. Now quarterly we’ve been updating it and are up to over 700 of them.

And we’re going to be doing a quarterly refresh, which coincides with the end of the year for all of the legal tech solutions that use generative AI that we are aware of. We also did an agent AI that we did over the summer that we will be refreshing for the end of the year. And now we in coordination with the Blickstein group, we’re mapping out the ecosystem for solutions specifically for corporate law or in-house lawyers. And so every time we publish these maps, we get a great response. But one of the biggest ones we get is.

Why am I not on it? How do I get on your maps? And that’s what I’m here to tell you. It’s actually an easy answer. If you’re not on it, it’s because we don’t know about you. So what we need you to do, if you want to be included in these market maps, is to create or update your profile on Legal Tech Hub, and that’s free. Every company can a free profile. And we need to know if you are using AI, if you’re using generative AI, if you’re using agentic AI,

specifically for law firms or corporate legal. So the answer to everyone’s question of how do I get on this map is we just need to know what you’re doing. So this is your PSA to update or create your profile on LegalTech Hub so that you will be included in our next data sweep to be in on our maps for the end of the year. We’re currently working on our map for corporate legal.

And so if you want to be on any of those maps because that’s what you do, go to legaltechnologyhub.com, and there’s a link on there about how to create or update your free profile. Or if you have any questions, you can email curation at legaltechnologyhub.com. So.

Let us know what you’re doing so we can let the market know what you’re doing and you can be on our next map if it’s appropriate. Thanks.

Marlene Gebauer (02:20)
Welcome to The Geek in Review, the podcast focused on innovative and creative ideas in the legal industry. I’m Marlene Gabauer.

Greg Lambert (02:27)
and I’m Greg Lambert.

Marlene Gebauer (02:29)
So Greg, before we begin, I want to acknowledge the passing of a dear friend from our KM community, Kim Stein. Many knew Kim from her work with Lexis, Thomson Reuters, and Upland Software. Others probably knew her from her many conference dinners and happy hours that she hosted. She had a big personality and a very generous spirit. She was really fun, she was loyal, and deeply present for people that she cared about.

I’m going to miss her deeply and today’s episode is dedicated to her.

Greg Lambert (03:01)
Thanks for doing that Marlene. So, well, it is that time of year again where we’ve brought in a couple of experts to help us look back on 2025 and look forward into 2026. they were with us last year. So we’ve got a scorecard to kind of look at how we did. so you can then use that to determine how much you depend on us to predict the future.

Marlene Gebauer (03:22)
Yeah.

Greg Lambert (03:33)
So we are returning for our annual State of the Union address on industry. We have our good friends Sarah Glassmeyer Black. So Niki and Sarah, thanks for joining us again.

Marlene Gebauer (03:31)
See how good we are.

Sarah Glassmeyer (03:32)
You

to be here.

Niki Black (03:49)
Yeah, looking forward to it.

Marlene Gebauer (03:51)
So a little bit of background in case anybody out there doesn’t know, but Sarah Glassmeyer is the Director of Data Curation at Legal Technology Hub and the industry’s resident truth teller, or as she might say, the neighborhood busybody of legal tech. And Niki Black is the Principal Legal Insight Strategist at 8AM. ⁓

She brings data-backed perspective of the solo and small firm market, tracking what lawyers are actually doing versus what the hype says they should be doing.

Greg Lambert (04:21)
All right, welcome guys.

Marlene Gebauer (04:22)
Welcome back.

Sarah Glassmeyer (04:24)
I was so nervous when I saw this and was like, what did I say last year?

Niki Black (04:25)
Thanks.

Marlene Gebauer (04:26)
I was like, really surprised after

the last time that they’re like, okay, we’re going to do it again.

Sarah Glassmeyer (04:31)
you

Niki Black (04:31)
Hahaha!

Greg Lambert (04:32)
I like the neighborhood busy body. think of Mrs. Kravitz from Bewitched. And I only watch those in reruns. I’m not old enough to have watched them when they came out. All right. Well, before we dive into the absolute chaos that was 2025, look back and take some receipts on last year.

Sarah Glassmeyer (04:35)
I know.

Marlene Gebauer (04:37)
Mm-hmm.

Sarah Glassmeyer (04:38)
Yeah.

Marlene Gebauer (04:43)
So he said.

Greg Lambert (04:59)
In December of 2024, our consensus was that 2025 would be the year that the novelty wore off and the boring became useful. Sarah, did the boring become useful or are we still, ⁓ what do you think?

Sarah Glassmeyer (05:16)
No, don’t think,

Marlene Gebauer (05:17)
Meh,

meh.

Sarah Glassmeyer (05:18)
I was like, that’s what I was thinking. There’s gotta be like either like a French or German word to explain the feeling you have. it’s, you know, I was thinking about like other tech adoption cycles. There’s like two kind of like wavy charts to think of. And one is the Gartner hype cycle where it’s like inflated expectations and then trough of disillusionment And then, you know, that it becomes like useful tech. But you also, that is overlaid or like.

Marlene Gebauer (05:25)
You

Sarah Glassmeyer (05:41)
Same thing with the tech adoption cycle where you usually have like early adopters and like the real innovator type people. And then it kind of like spreads through the organization and then people start. like, but like the problem with AI and like LLMs and generative AI, it’s like everything was thrown at everyone all at once. And it’s also like.

ubiquitous because like there’s no you you weren’t like watching a Super Bowl and seeing the ad for a document automation tool or a timekeeping tool but it’s like here’s generative AI and you can play with it and here’s how changes your life and it’s also like not just work it’s also like playing your kid’s birthday party or figure out how to cook dinner you know I mean it’s like everywhere

And I think we’re like overwhelmed with possibilities still. Like it’s just like, I don’t even know where to use it. And that’s the thing. It’s not like, here’s one finite tool that does like contract review or here’s one finite tool that you can use. You know, it’s the Ginzu it slices and dices, it does everything. And it’s everyone’s like, why don’t, where do I even begin? Like, so that’s the thing. It’s like, it’s everywhere. And like every, like every vertical legal tech has an option pretty much somehow to use Gen.ai in some facet of it, but it’s like,

it’s still not being adopted because we also have to go through traditional adoption. It’s not that special. You still have to do use case and have your campus and have your team.

Greg Lambert (06:56)
Yeah. You you said the, the Gartner

hype cycle. I haven’t heard anyone reference the Gartner hype cycle in months. Now that I think about it, maybe, maybe it’s just the crowd I run in, run into, but I, I think you’re, I think, you know, what you were alluding to is right. That this is kind of off the scale as far as yeah. Yeah. We’re, yeah, we’re in the trough.

Sarah Glassmeyer (07:05)
Yeah. ⁓

Marlene Gebauer (07:16)
It’s a constant hype cycle, you know, because it’s like new things, like, squirrel, ooh, squirrel, you know, and,

Sarah Glassmeyer (07:18)
Yeah.

Greg Lambert (07:23)
disillusionment for you know, chat GPT, you know, 4.1 or 03, but now we’re on 5.2.

Niki Black (07:33)
Yeah.

Sarah Glassmeyer (07:34)
And it’s also just like…

Marlene Gebauer (07:35)
And I want to piggyback

on what Sarah’s saying. it’s, it’s, it’s like AI is, is the, is, is advertised as the tool, not like the tool is powered by AI, which makes that much more hard to figure out. Okay. What do I use it for? Cause I mean, you could have like, okay, this is the, this is the contract aspect that it does. It does contact drafting. Okay. Everybody understands that. But if you’re saying, this is just an AI tool and it does everything. Then I think it becomes.

astronomically hard for people to say does this apply to me and how.

Sarah Glassmeyer (08:05)
Yeah, mean, it’s just like, the other thing is like, as part of the conversation about it, there are some people who like have made their whole personality hating AI, and there’s some people who made their whole personality being like, AI is going to change everything. And it’s also like, so it’s like not helping the conversation and trying to actually figure these things out. But it’s like, it’s so fascinating to like see on LinkedIn how some people are just like.

And we’ll probably talk about this a little bit later, but like the Reddit sub comments, legal tech sub Reddit and just like how people there just like Harvey, hate it. But you go on LinkedIn and there’s like the real like fanboys of certain tools. And it’s just like really like, huh? No one’s getting this hype about a private tool, maybe they should. mean, that’s like, did you care about like the underlying tech before this? So just stop worrying about it. It’s really weird, I don’t know.

Marlene Gebauer (08:44)
so invested,

Sarah Glassmeyer (08:55)
It’s kind of broken everyone’s brains.

Niki Black (08:57)
I tend to look at things from sort of a bird’s eye view and then from a practical view. So by bird’s eye view, mean, in the context of everything that’s happening, I think that you can’t really look at AI without looking at sort of, there’s this feeling of chaos within technology and AI and legal tech even, but there’s just chaos across the board.

And I think that that like, it just feeds into that. Like nothing is predictable, whether it’s the, where we are in the AI hype cycle, whether it’s what’s happening on a geopolitical basis, what’s happening within our own country. There’s just so much unpredictability. And I think that that makes it so that the AI adoption and hype cycle feels even crazier than it actually is.

It does feel like there’s just something different being spoken about at every conference. mean, agentic, everyone’s talking about agentic AI, but very few people have implemented it. All the legal tech companies are incorporating AI into their software. mean, without fail, think practically every platform has done that at this point in one way or another. they’re doing, they’re having deeper and deeper integrations over time as they start to figure this out. So people are just able to use AI in a tool that they’re already using.

on that more practical or granular level, even though it feels like we’re in a hype cycle, the data shows, at least the data that we have, that the adoption levels are exponential, unlike anything that I’ve ever seen before. I tracked the cloud computing adoption. I wrote the Cloud Computing for Lawyers in 2012 that the ABA published. And I’ve been tracking cloud adoption ever since, really since 2010. And we…

we’re stuck at about 30 % for cloud computing adoption for years, starting in about 2015. And it wasn’t until the pandemic that pushed us over 50%. And now, as we all know, all the even the software companies are sun setting their server based stuff and pushing people into the cloud. So now we’re kind of at the, but that took a long time. I mean, we’re talking the cloud came out in 2007. Here we are in 2025, finally seeing cloud adoption at a large scale level, with generative AI, which

came out in the fall of 2022, three years ago. Last year, our data, I’ve write the 8 a.m.’s legal industry report, and this is my fifth year doing it. And last year, we had a whole section on generative AI, and the year before, we did as well. And so it’s interesting to see over time the adoption level. And these were lawyers across the board, but I would argue that it was more, not argued, the data shows that it was.

more solo, small and mid-sized firms versus large. But our data showed that the adoption went up significantly. was close to three quarters of the individuals surveyed were using it in their daily legal work, general purpose tools, compared to 31 % last year and 27 % the year before. And then when you talk about law firms, we found 46 % of law firms

had adopted these tools. And that was up 21 % last year of legal specific tools that law firms were adopting. So the numbers are going up significantly and it may feel like there’s not a lot of adoption, but there really is. That’s why we’re seeing so many headlines about all the hallucinations lawyers are using it.

Sarah Glassmeyer (12:15)
I don’t really remember this happening before, but like, FOMO, like, people are definitely having that fear of missing out, that they like, I wanna have it, you know, and I wanna try it at least. I was kind of thinking, I was speaking to a group a few weeks ago, and they said, we really wanna, we can’t wait till we get our hands on Harvey. And I was like, well, what are you gonna do with it? And they’re like, we don’t know. don’t have a plan. Like, that’s not.

Greg Lambert (12:34)
No idea. Gotta play.

Marlene Gebauer (12:37)
But we

have it!

Sarah Glassmeyer (12:38)
But like, they’re like, that’s the thing. It’s like the cool kids get Harvey and we want Harvey. Like we want, you know, and it’s just like this weird, like, I don’t know. It’s just like stay in culture coming to legal tech and like people just really wanting to be able to say they have it. And it’s like not great for actual business productivity. But yeah, I mean, that’s the thing. Like it took, you know, that’s thing. Why we have to Reynen Court That was the thing that we had this like safe cloud because still the cloud was still dangerous and it was scary. And that was 2021 for a lot of big law. Now it’s like, yeah, you just put it up there. We don’t care.

Niki Black (13:06)
Yeah.

Sarah Glassmeyer (13:07)
But yeah.

Greg Lambert (13:09)
Yeah, well,

one of the things that I think is significantly different now this year in 2025 than it was in 2024, and this is a very hard metric to nail down, but I tell people that I have never, I’ve been in the legal industry since 1994 when I started law school,

I have never seen lawyers excited about the work that they do. And this year it was like a ⁓ switch got flipped and they’re just like, ⁓ you know what, this boring thing that I’ve been doing for years, now I got this, you know, got a chat GPT to do this. And then of course, you know, then I rushed into the room and say, no, you can’t use chat GPT here. Use Harvey, which uses chat GPT.

Sarah Glassmeyer (14:02)
you

Greg Lambert (14:05)
⁓ But just this excitement level is something that I’ve never ever seen before. I don’t know how we can maintain it, but it makes it a lot easier to work with lawyers this year than it’s ever been.

Niki Black (14:05)
Right.

And another really interesting thing I think that is happening really recently, like in the last few weeks I’ve noticed this, when you’re talking about chat GPT versus Harvey, a lot of people, especially in the solo and small firm space that are considered experts in this space are suggesting that the business versions of chat GPT, of Gemini, of Claude, are sufficient for, have sufficient protections for the vast majority of legal work. And that you don’t, you can use those for most

and input most data unless it’s particularly sensitive and use those. And they’re much more affordable. mean, know, Harvey’s not cheap and it’s way above the budgets of most small and solo law firms. And so that’s, you’re starting to see this sort of trend where people, ⁓ those companies are understanding the business needs of different professions and they’re incorporating controls in the business levels, you know, the business tiers.

that provide the levels of security and confidentiality that people need. And so that’s gonna be interesting to see how that impacts things. And the other thing that we noticed in the industry report is that there are more law firms investing in chat GPT and those other general purpose tools compared to legal specific tools, at least more at the midsize and small firm level. But it’s interesting to see sort of these trends over time.

Sarah Glassmeyer (15:42)
That was actually one of my predictions for 2026 is that, you know, the specialness of legal tech is not going to be always there anymore for two reasons. One, kind of building on what Greg said, you know, back in the old days when I was working at digitization of library services, we always talked about our, like the four horsemen that we were competing against on Facebook, Amazon, Apple, and Google.

Greg Lambert (15:57)
Hoo hoo!

Sarah Glassmeyer (16:04)
that you needed your online versions to be as easy to set. Like it’s checking down an ebook as easy as getting an ebook from Amazon. And that’s the thing, there’s now like all these commercial versions of specialized tech, why use a Harvey or a Legora when you could do OpenAI? And that’s the other thing is that OpenAI is coming for legal tech. mean, they have our professional services. I think it was them that just like whipped up this little kind of contract maintenance tool or contract review, contract drafting tool that…

Um, it was one of the main ones, either them or Gemini. But yeah, I mean, so it’s a thing like they see them, they see lawyers now as a customer, the general purpose tools that so, I mean, that’s the thing you, what’s like, like, what’s Iltacon going to look like now? Like OpenAI has a booth right there next to Harvey. mean, so I think we’re going to see a lot more, you know, that’s the nice past 20, 30, 40 years, Legal Tech’s been our own nice little.

Enclave a Weirdos, where we kind of know what lawyers want and all of our special needs. But now I think we are going to see some of the bigger players kind of coming into this space. And that’ll be interesting to see how that all changes everything.

Niki Black (17:07)
Well,

Marlene Gebauer (17:05)
celebrity death match at Delta.

Greg Lambert (17:06)
the?

Niki Black (17:07)
and that was

a bunch of your Legal Week three or three years ago, like right after ChatGPT came out. I actually got in a really heated conversation during dinner with a legal tech founder who told me that respectfully, I didn’t know what I was talking about when I suggested he was a little rude to me and I got a little I responded in kind, but I suggested that

I thought that in a few years, OpenAI and some of these other tools would have the ability to produce low hallucination rates and legal specific tools because once they train it on that data, they may be able to produce output that is at least as good or close to what the legal specific tool they’re going to be producing. And he ardently disagreed with me with that. Again, he was the CEO of a tech company.

But I think a legal tech company, I do think that where that’s happening now, we’re getting to this point where these tools may be able to, and they have a lot more spending power and they have a lot more investment and they have a lot better ability to advertise and reach people than a lot of these smaller legal tech companies in comparison. So it’ll be interesting to see how this all plays out, but it’s a really strange time to be covering legal tech.

Marlene Gebauer (18:17)
We’ve, we’ve,

we’ve talked about on the podcast about like how maybe, you know, open AI would, would buy an information company, know, and, and, you know, if that happened, you know, God knows what, what would happen after that. So.

Greg Lambert (18:30)
Yeah,

I to pitch Microsoft buying Thomson Reuters, but now it’s OpenAI. I will say just one thing, just to kind of close this and then we can move on to hallucinations. thing that OpenAI is doing that seems to say they’re not going after legal yet is just the amount of investment.

that they’re putting into the legal industry for them to use their products. So it’s like the circular, it’s like, well, we’re supporting you in order for you to support us. But yeah, think, Niki you’re right. Once the legal industry isn’t necessarily, or at least the legal vendors aren’t big enough to support open AI, they could go direct, you know.

cut out the middle man and take the whole thing for themselves. So that will be it. Do you think 2026 will be that year or do think it’ll be later if it happens?

Niki Black (19:26)
I think 2026 you’re going to start to see signs that will happen, but I predict it’s going to be mid to late 2027 where that may start to… Maybe not in large law firms, but large law firms are not the majority of legal practitioners. So think you’re going to see it in the mid side market down happening a lot more because they can’t afford these other products. They’re going to have to use OpenAI and Gemini.

And a lot of them are already, or Microsoft, you know I mean? Because these are the tools, the ecosystems are already in, and it’s going to make the most sense for them to use them. And then use the tools that are on a smaller scale built into the products they’re already using that only have like siloed feature sets, you know, that work specifically within those products. So, and it’ll be, I don’t know. It’s a really interesting, I keep saying interesting. I’m really repeating myself here, but I think that’s what I keep landing on too. It’s a weird times, but we’ll see.

Greg Lambert (20:17)
It’s definitely weird times. let’s jump into hallucination. Last year, Sarah, were talking a lot about, or the last couple of years, how the vendors were really hanging, or the legal information vendors, Westlaw, Lexis, all of them were hanging their support on the retrieval augmented generation as the savior for ⁓ hallucinations.

Niki Black (20:19)
Hahaha

Greg Lambert (20:43)
Some were even coming out in their promo materials as hallucination free. ⁓ So, we still hear almost weekly of some lawyer getting busted for going to probably the free version of ChatGPT and asking it to do legal research. What’s your view on the industry now as far as at least the legal information?

Marlene Gebauer (20:48)
You’re free.

Greg Lambert (21:10)
vendors on how well they’re doing in preventing this and thus the reason that we still need legal information vendors in AI.

Sarah Glassmeyer (21:20)
Yeah, it’s interesting. So I talked to lot of legal tech founders and developers, and there’s a couple of different kinds. And some are lawyers who had like a particular problem and they decided to like hook up with some tech people and solve that problem. And then you have other people, especially now in the gen AI world, that they were a data scientist of some sort, and they were just looking for a new challenge. And they’re like, hey, law has a lot of words. I can attack that. And so was talking to one of those such founders.

And his, was really, it really just blew my mind. I don’t know why it did, but he said, ⁓ yeah, hallucinations are a feature, not a bug. And I was like, my, my, are you sure about that? But he’s like, he’s the front, the way these tools work and from the way like my mind says a data scientist, it is a feature. Like we want it to like kind of extrapolate and make things up. And it’s like, do we though? But so with.

Greg Lambert (21:54)
You’re welcome.

Marlene Gebauer (21:58)
We’re here to help. There you go.

Sarah Glassmeyer (22:11)
RAG you know, that is a thing that it does. You know, I know like VLex goes through and cites everything directly. And I just had this, you know, after like probably the 10th, you know, thing that came across my social media feed of a lawyer getting busted for submitting a hallucinated case. You know, just reminded me so much of back in the old days when I taught KeyCiting and Shepard’s and like telling my students, okay, it’s a yellow flag, but you need to still go read the case. And they’re like, yes, Professor Glassmeyer we will absolutely read the case.

And I just realized they didn’t read the case I bet now. I think they were lying. And so really, for a lawyer’s duty of competence and duty of work, you should treat it just like KeyCiting And also, PS, you should have been reading the case when you were KeyCiting or Shepardizing stuff. mean, you are always going to have to do some due diligence. You shouldn’t have been trusting these tools in the past anyway. I remember it’s Aaron Kirschenfeld, I think is his last name. He’s at UNC.

Greg Lambert (22:40)
Hahaha!

Sarah Glassmeyer (23:05)
He did this really interesting study of like looking at the differences between TR and Lexis and their flagging systems and what it actually meant and realizing that it really wasn’t a firm science behind it. So I think it’s just kind of a continuation of what we’ve always have been dealing with. We maybe not realized we were dealing in a world where you had to always verify everything coming from these magic boxes, but we really have had to always do this. so I’m not like…

overly worried about it. fact, it’s like the fact is that you have to you have to verify. You’re never going to not be able to verify and PS you should have always been verifying. But yeah, I mean, think that is going to be at least and also one of the things with LLMs and gender of AI. There’s so many tools or so many like options of like what they actually do that legal research is actually just a small part of legal of it is a small part of lawyer work and it’s a small part of you know the capabilities of these tools. So like

Niki Black (23:46)
Okay.

Sarah Glassmeyer (24:01)
Don’t worry, like if you’re doing like straight on legal research, I would stick with a legal research company because word that has like a backup of a corpus that it will do rag on and you can go back and read the cases to make sure everything’s being cited correctly. Other

workflows, sure, like don’t worry about it so much. Like use, I mean use those other generalized tools or more like non-TR or Lexis, or VLEX based tools. That’s my thought.

Niki Black (24:21)
Well.

Marlene Gebauer (24:21)
I’ve always wondered

why research is like the holy grail for, for gen AI. Like that was like the one of the first things I gravitated to. It’s like, why, why that? Yeah.

Greg Lambert (24:27)
Yeah, and that’s super hard.

Sarah Glassmeyer (24:31)
Like it’s one of those things because it’s like

You know, having talked to other people like about this, you know, like it’s a process and sometimes the right answer is there no right answer that it’s something that it just hasn’t been settled law yet. And so, and also the fact that the corpuses are so hard to get a hold of, you know, there’s basically like four companies now that have like all the law. Yeah, so there’s like so many other things you can do with generative AI and LLMs. that, leave legal research alone. It’s just like, it’s like, and that’s actually might be one of the problems of why this adopt, because it’s so, such a hard

and such a skill and a process. so like, obviously if it’s an error, you know, why start at that as your entry point to this tool? Like, yeah, do something like reading narrative, trying to create timekeeping narrative or something like that. Like something that’s not so black and white or gray as a case may be.

Niki Black (25:11)
Wow.

Well, what am I?

Greg Lambert (25:22)
And Niki, what

Marlene Gebauer (25:22)
Nick,

Greg Lambert (25:23)
are you seeing on the small side of things when it comes to hallucinations? Are they abandoned in that or are they still going head strong into the winds there?

Marlene Gebauer (25:23)
I was gonna say with Niki.

Niki Black (25:28)
Well…

I mean,

I don’t even know if it’s a small law perspective. I mean, all lawyers are using this and they’re all ending up in the, so many of them are ending up in the headlines. One of the most interesting things that I’ve seen very recently though was one of the recent orders that was a big sanction order where the person was sanctioned significantly. When they described their research process, it was kind of fascinating because I think this fear of hallucinations and everybody saying double check, triple check.

make sure it’s accurate. This person used six different tools, some of which were legal specific. Like they use, like ⁓ OpenAI, I think they use Claude, I think they use like LexisNexis vLex, and some other, I don’t even know what they might’ve been like. I don’t wanna like throw certain ones onto the bus. I’m not even sure, but it was definitely some very well-known legal specific tools. One, we have been like a brief drafter or a contract drafter, and then some right… ⁓

Marlene Gebauer (26:10)
takes longer.

Niki Black (26:25)
consumer tools and they used all of them to cross check and triple check. They’re working at that point. It’s like a game of telephone. You you enter it into one, ask them to check it, pop it into another, and by the time it comes out of the eighth tool, of course it’s not accurate. And I feel like people are so terrified of screwing up that this is what’s happening. Like they’re just doing site checking in one, but then it changes some words and they don’t realize they did it because that’s what these tools do. And so I think that it’s like becoming a lose lose situation where

When you’re talking about research and drafting briefs, really don’t think you should be using, for the initial draft, any of these tools. I think it’s okay to like, for like editing the text or the grammar and possibly double-checking your argument, but keep your original that you wrote, that you did the research on in the traditional tool, that you drafted on your own so that it doesn’t get all screwed up and then use one of those tools that cross-checks to make sure there’s no changes.

I forget what they’re called, but you know, they, when like opposing counsel submits some, like a new contract, a new cross check to see, I don’t know what they’re called, because I never did contract law. But yeah, but I do feel bad for practitioners right now because this just wasn’t a, when I was practicing law, this was not a thing. And it’s got to be kind of scary. You don’t want to be a headline. You don’t want those 10,000, 20,000, $30,000 sanctions they’re starting to hand out now. But.

Greg Lambert (27:29)
brief analyzer.

Marlene Gebauer (27:46)
And there’s also

like a, a, a financial aspect to it. Like, you know, is the client going to pay for all of that? did they know, did they, they understand that it’s necessary? Is it necessary to cross check it five different times? And, and so that’s, that’s a discussion too.

Greg Lambert (28:02)
Well, I do want to say, yeah, no, not yet. I do want to say something positive though, that I’ve seen, especially maybe in the last six months. and I, it was noticeable enough that when I was at the, the Clio conference, I had a conversation with, with Ed Walters on the side and I asked him, I’m like, cause I was like, Ed, I’m, I am seeing a significant improvement.

Niki Black (28:03)
And does that save on productivity? No. ⁓

Marlene Gebauer (28:07)
No, no

it doesn’t. Let us explain why.

Greg Lambert (28:31)
on the answers that are coming back from the legal research tools. And I just kind of let him know straight up, I was like, you know, for two years, it was crap. It was not good with the results. And then all of a sudden, maybe this summer, I started noticing that the answers are coming back. They were much more coherent. The cases and statutes that they were referencing were much more on point.

Still not perfect, but a hell of a lot better and I asked him I said is it you know, is it the reasoning models? Is it because the foundation models have improved and he pointed out and I’ve Since have done like two months of research on this That it’s like yeah. Well, that’s part of it. But the other part is that one the the information vendors are better understanding how to

gather the information and group the information, whether it’s in a rag or vector database for the rag technology or into a knowledge graph for the hierarchy of how legal is set up. So now it’s essentially set up in two categories. It’s put in the vector data, the data is put in the vector databases and chunked and put in so that the contextual information can be understood.

And then it’s also put in, so there’s two set of data, storage that they’re putting in, and it’s also put into the knowledge graph so that it understands the hierarchy of the courts, whether it’s overturned, whether it’s been cited. And so it goes from one back to the other to compare those and get the best matches before it’s brought back and the answer’s generated. So the retrieval.

part and the augmentation part has advanced quite a bit in the last six months. then the generation, thus, is being a little bit better. So I think when they jumped on RAG, it was just throw everything into the vector database. We can do contextual, which really finally gave us what we were promised in 2002 of natural language search.

fruition 23 years later, but it was, just not getting that, you know, that second level of stuff. if they finally are figuring that out. So kudos to, to some of them, because I think if you, if you did the same research a year ago and you do it now, it is significantly better than it was. It’s not perfect, but it, but it is much, much better.

Sarah Glassmeyer (31:14)
Yeah, mean, again, one of those Ed Walter comments things is like, this is as bad as AI is ever going to be. It’s always going to get better. And it’s also kind of a shame, like how it was such a splash and everyone got their hands on it two and a half years ago, and they’re like, this sucks. Like, I don’t want to use it again. Without realizing, like, well, just give it time to improve. it was like too much of an explosion that too many people got to try it instead of letting it work its kinks out and then trying it.

Greg Lambert (31:25)
It was terrible.

Marlene Gebauer (31:39)
Yeah. So I want to talk a little bit about mashups because this was definitely the year of mashups. had big law tech and mashing up with small law tech, and then you had vendor mashups. So I think Greg, you were calling the Clio acquisition of vLex their big swing at big law. And I’ll direct the question to Niki. What do you think this consolidation means for mid-size firms or for large firms? ⁓

you know, are we moving into the Apple versus Android of law? ⁓

Niki Black (32:11)
Well, this, so I don’t know if you recall, I used be part of a company called App Folio that my case was owned by them for eight years. And a lot of people that worked from at App Folio came from Citrix from the server-based universe back in the day. And when, at the time, know, law practice management had, there were like whack-a-mole with law practice management. There were just so many different companies popping up that were small startups. And you had like the big three at that point, which I would argue are.

my case, Clio and Rocket Matter. And then all these other ones started popping up and there were so many. And one of those people that came from Citrix said, listen, they didn’t say, they didn’t, it wasn’t a Battlestar Galactica reference. That wasn’t their intent, but it was. This is all happened before. It’s gonna all happen again. We’ve seen this in the server-based world. All these companies pop up, there’s massive consolidation. Then there’s just a few big players. And then you saw that with practice management and then the cloud in general, right? Like all the cloud, legal cloud products.

You see a saw it in eDiscovery as well, right? All of the like a whack-a-mole, consolidation, a few big players. And so I think that what we’re seeing now is AI has kind of created a new category of this, but that’s the same thing that’s happening. And I don’t think it’s all just going to be consolidation. don’t think it’s going. And I think what Clio did and vLex this acquisition was very interesting. And I definitely think that it was notable.

But what it really did was it made everybody around them, the LexisNexis and Thomson Reuters, sit up and take notice about this comment. And they’ve tried to do this before to combine the practice of law and the business of law, but pre-AI, that didn’t go so well. And the couple of people, think Ken Crutchfield is one person that wrote about that. But AI changes the game. so now everybody’s looking. Just today Aderant announced Harvey integration.

So it’s not just going to be acquisitions, it’s going to be integrations. It’s not all about M&A it’s integrations between the business of law and the research, like the actual work of law, the practice of law. And so you’re going to see acquisitions like that that continue along that vein, but you’re also going to see tons and tons of integrations where I don’t necessarily agree that it’s going to be the single platform that everybody works in. Some people may choose that route, but others are going to be like, no.

I like my practice management software or I really like Harvey. I don’t want to switch to vLex or I like, you know, some other tool like in my document management, I like NetDocs, you know, and I don’t want to switch into some other environments. So they’re going to look for those ⁓ integrations that are going to give them the functionality they need, especially as you move up into the larger firm. it’s, again, there’s a, it is definitely consolidation, but I don’t think it’s going to, if there’s going to be some big major players, it’s going to be.

I’ve been thinking this all along. It’s going to be open AI, Google, Microsoft or something. I don’t think there’s going to be huge players in our space because I think that big players in our space are going to get dwarfed at that level as things level out across the software in general, not just legal tech.

Greg Lambert (35:09)
Well, how do we keep from going into what ⁓ Corey Doctorow calls the enshitification of everything, which basically you get four or five companies that run an entire industry. what happens is they don’t necessarily look at what they can do for the customer, but rather what can they get from the customer to then pad the bottom line of

of their company. it’s like you see less and less really good features coming out and you see more and more gathering information from their clients and then spending that into whatever the next product it is that they’re going to sell, is half as good as what you originally bought. I always worry when we get into that and especially if it’s like you potentially project

Niki Black (35:54)
I.

Greg Lambert (36:01)
Niki and that it’s not even legal companies that are going to be doing this is going to be these broader, you know, global behemoths of AI that are running the world. Well, I mean, after a while, we’ll just let AI do everything and then we can just get our universal income and, ⁓ you know, just go and we’ll be in Star Trek world.

Niki Black (36:23)
Well, mean, listen, that’s capitalism, like late stage capitalism. That’s just what happens across the board. And I think the only solution is like a reboot. think that, you if you operate, I like to operate on the assumption we’re in a simulation, so we just need a reboot. We got to like try this all over again, because at this point, everything, the stress test has not worked and everything’s becoming shitified. So they just need to start over again. So I think that’s the solution. ⁓

Greg Lambert (36:43)
All right, so Niki’s projecting a revolution.

Sarah Glassmeyer (36:46)
Hahaha!

Niki Black (36:48)
I’m rejecting just a reboot of the simulation. Not even revolution. I just think everything needs to be restarted.

Greg Lambert (36:51)
Okay, take the blue pill. the blue pill.

Sarah Glassmeyer (36:54)
I mean, so that’s one of the, yeah, so literally 15 years ago now, I made a chart. It was one of those things I did like spontaneously, because I was kind of bored one day and it turned into like this big thing that showed the consolidation of the legal publishing market. And there used to be in starting back in the late.

late 1800s, 19, dozens of companies. And then came down to basically just Thomson Reuters and Lexis and then to extent, Wolters Kluwer And I always show, I still like talk to law schools classes. I always show that and I’m like, well, we learned our lesson. That’s why I’m in legal tech. Cause I was so mad because legal publishers treated librarians so badly. And it was like, we had no, no choice, no like real power in the relationship. And, then I showed like this new slide of just like M&A and

in legal tech. I have great concerns about this as well, because once again, we learned this lesson in libraries. Will we learn it in the broader legal tech world? I hope so. But also, one of the things that I’ve always been very hopeful for and promoting is the idea of integrations. Keep your small company and integrate. Use APIs, use Opens, and use standards like SALI or SOLI or whatever they’re called, and that whole mess.

I don’t even know what’s going on there, but like, you know, are like options to do data standards so your, you know, different tools can talk well with but as far as like the vLex Clio thing, you know, it’s so funny like to compare that to TR, cause that’s also.

You know, there’s kind of the one is running jokes and in the nerdy legal tech world is like, what’s your favorite product that TR bought and ruined? You know, they’re kind of notorious for doing that because they are such a huge corporation. I’m very hopeful though about the vLex Clio thing for many reasons. A, I think because there was like personal relationships between the various product teams and owners that there will be slightly more care done that it won’t be like a huge conglomerate and shitification, but also

Greg Lambert (38:27)
Yeah.

Sarah Glassmeyer (38:47)
I don’t know, maybe I’m just fooled by Jack Newton’s pleasant Canadian exterior that he’s… I’m like, well, let’s just trust Jack Newton. He’s fine. He’s not gonna screw us all over. But also just the idea that because Clio or vLex has what I consider world-class tools, they’re up there with all the big guys.

Greg Lambert (38:54)
Yeah.

Sarah Glassmeyer (39:06)
And now with Clio’s general customer base, like I could care less about them kicking it up to Biglaw. They’re making a play for Biglaw. I think it’s really exciting that the traditional Clio customer base will now have access to these kinds of AI tools, which they may not have been able to get to before because Harveys won’t sell the Solos like at present, you know, but now they can try to start experimenting with these various types of tools. So I’m pretty excited about that, both just for new opportunities for some of these smaller midsize law that they can really take advantage of.

of

the AI efficiencies and either for access to justice purposes or for profitability purposes, they’re really going to be able to benefit from these tools, which they maybe not have been able to do before.

And also, I said before, what I saying, like don’t be a fan girl or fan boy to Legal Tech, because that’s weird. The scream I scrubbed when I saw on the news come through about Clio and vLex I always keep calling it that the Taylor and Travis of Legal Tech, but it was like, it so exciting that they’re like, oh, good for them. They’re getting together.

Greg Lambert (40:06)
Yeah, I saw the size of that ring. It was pretty massive.

Sarah Glassmeyer (40:10)
Yeah!

Marlene Gebauer (40:11)
hahahahah

Greg Lambert (40:15)
All right,

well the next time I see Jack and Ed, I’ll ask which one’s Taylor and which one’s Travis. ⁓

Sarah Glassmeyer (40:21)
I know, I know.

Greg Lambert (40:25)
let’s skip to the next part where we’re talking about whether or not we trust the infrastructure with AI and also the issue of shadow AI, which we’ve kind of woven through this entire conversation without actually saying shadow AI. Marlene and I had spoken with Nam Nguyen from ⁓ Truth Systems back in October.

Um, and you know, there seems to, uh, in, in his whole thing was like, all right, let’s keep honest people honest, but let’s also verify that honest people are staying honest with the, with the AI system. And in fact, I just saw today with, uh, on Harvey, when I, when I logged in was, uh, with the advent of chat GPT 5.2, which is agentic

at default now. Now it’s, well we have 5.2, but it’s with the Harvey guard rails on it now. So now they’ve got to actually put in guard rails against the systems as they advance. So it’s this really interesting dynamic between the builders and the foundation model. So Niki, me, me.

Pop it over to you. Do you see this as, we’re going to need much more hard policies and surveillance on the usage of AI as these models get more complex? Or do we rely upon the vendors to do that work for us? What do you think?

Niki Black (42:04)
You mean surveillance on behalf of the firm, on behalf of the…

Greg Lambert (42:07)
Yeah, I mean just what your

people are doing.

Niki Black (42:10)
I mean, they can take out their personal phone and use ChatGPT and you don’t even know what’s happening, even if you give them a business phone. People are going to use these tools regardless of whether you tell them to or not. There’s rules about whether you can enter company confidential information, but they’re going to use them for, hopefully they’re following that rule and understand the implications of entering confidential company or client information into those.

general purpose tools, if they’re not in a tier where that’s covered and if it’s forbidden by firm policies, but they’re still going to be using them. mean, at the end of the day, don’t think surveillance is the answer. This is too big of a, it’s like a tsunami or a tidal wave or something. You can’t surveil that. You’ve got to work to do damage control once it’s already hit. You just got to be doing damage control. You can’t be predicting where it’s going to go or what’s going to happen. This thing is already here.

and surveilling it is not going to be the solution. And at the end of the day, you’re going to be spending more resources and money surveilling what everybody’s doing than actually just making sure it’s implemented properly and tracking the changes that are being made by the larger vendors through which the legal tech vendors are connected. I I just think it’s a waste of time and resources to be constantly surveilling everybody and then following through on those and having all these HR incidents. Where are you going with all that?

I think that’s unreasonable and people that are doing that are making a mistake.

Greg Lambert (43:37)
Well, we need a bigger security ops team.

Marlene Gebauer (43:40)
It also goes

Greg Lambert (43:40)
Sarah,

what’s your thoughts on how to have, you know, we obviously can’t control it. So how do we contain it?

Sarah Glassmeyer (43:47)
Yeah, it’s so funny, just during what Niki was talking, I was reminded of, it was like this slide that was very popular. You saw it all the time in presentations about five, six years ago where big data is like teenage sex where everyone says they’re doing it, but no one really knows how to do it. so they’re, yeah. And then something like, maybe it is also like this is because like basically they’re gonna do it anyway. So you might as well make sure they’re doing it the safest way possible. So I think that’s probably my, and also just coming from my general professional development education,

Greg Lambert (44:03)
Nobody’s doing it right.

Sarah Glassmeyer (44:16)
pro background that I think that you’re going to just have to do a lot more training and a lot more. You understand they’re going to try it. They’re going to want to do it. So make sure if they, they understand the risks and they understand like what they’re doing and why, what they’re looking at may or may not be valuable and why, what they’re doing may or may not be safe. You know, just kind of give as much like user training and opportunity and like safe spaces for them to experiment. So it’s not like.

them other folks. have, I can’t remember if I even like mentioned this last year during this present, this episode that I was actually doing, I was moderating a session and a question from the audience is a younger associate saying that they had a friend who always just brought his iPad into work and just was using ChatGPT privately. And you see like all the other, like the regular law firm people in the crowd is going like.

my God, I have to go make a phone call. It is like a panic, like, you know, yeah, I mean, again, it’s like, well, it’s gonna happen. So let’s make sure it happens in the safest, most responsible way possible and not just them feeling like they can’t tell you that they’ve made a mistake. That’s the other thing I always, you know, creating like an open door, kind of open communication thing. If it doesn’t become like this forbidden fruit, then they’re going to be, if they do screw up.

Marlene Gebauer (45:10)
Hahaha

Sarah Glassmeyer (45:31)
they’re not going to tell you. Like, make it so it’s not going to be the end of their career, the end of their world, if they do tell you they made a mistake.

Marlene Gebauer (45:38)
Also, I had like some of this ties into actually getting the right tools that are useful for people to do their job. And I think there’s a challenge doing that because hat like, again, how are organizations capturing that type of value? Like what’s the, what’s the value prop? Like what are people doing? I don’t know that everybody has a good grasp on all of that, you know, on different workflows and things like that. So you bring in something which.

may work or may not work. And then if it doesn’t work, then they’re going to these, these shadow opportunities. So I think it’s a, ⁓ that, organizations need to get a handle on sort of what problem are we trying to solve here? What are we trying to do here and make sure that the tools actually, ⁓ we’re up to snuff.

Greg Lambert (46:25)
Marlene, brought up return on investment here, or value. I’m just wondering, because I think we mentioned this last year as well, well, we also have to start showing some return on our investment in 2025. I don’t think that that happened. I haven’t seen a Skadden or a Kirkland or any of these step up and go.

Marlene Gebauer (46:31)
Yep.

Greg Lambert (46:53)
Oh, well, you know, we reduced our costs by 20 % and thus our revenue is, you know, has jumped by 40 % yet. Unless I miss that one. But, but well, well, no, we never keep it quiet. My God, we’re the only industry that goes out and tells what everybody makes. it’s

Marlene Gebauer (47:04)
unless they’re keeping it quiet, which they might be.

How?

I don’t

know, I feel like that the-

Greg Lambert (47:19)
And we have

a damn list of every firm and how much money they make. they’re not in their private companies, these private companies, these are not public entities.

Marlene Gebauer (47:25)
I just feel like it’s been a little more closed.

They just feel it’s like it’s been a little more closed about sharing. And I know we’re a big sharing community, but I feel like it’s been a little more closed about here’s what we’ve developed and here’s the impact on it. I I could be wrong. will, I will toss it out to, I’ll toss it out to our guests and see what they said.

Greg Lambert (47:45)
I’ll take the other side of that. I’ll

take the other side of that because the answer is we have put a ton of money in this thing and we don’t have Jack coming out the other side.

Marlene Gebauer (47:58)
You

Greg Lambert (47:59)
Sorry, we went on a little rant there.

Marlene Gebauer (48:00)
We took over like, let’s.

What do you guys think about this?

Niki Black (48:04)
Well, I think that when they’re actually going to start to see the return on their investment is when, and I think this is coming, when the billable hour gets discarded. Because right now, any increase in efficiency undercuts billable hours. And that’s the bread and butter of large law firms are these $2,000 hourly billing rates, right? And so I think that once you start getting more AFAs implemented,

you’re going to start to actually see that return on investment. I think, I don’t remember if I said this last year, but I have been seeing this across the board in multiple outlets. I think that that’s also where AI comes in. When you have AI that is built into either your practice management system or your legal billing system, and it’s a sophisticated generative AI tool that can analyze all the law firms’ historical data, billing data, and case-related data. And then you can ask it.

We just, have a potential client for this kind of case. This is the complexity. These are the issues involved. This, know, give it all the information, look at all of our past similar cases and provide us with an AFA arrangement that will be both competitive and profitable, you know, and, and it’s able to crunch that in a way that was never available before. And I think that they can actually get really good output in that way. So I think that you’re going to, and what I think, and it’s already happening because corporate legal counsel.

are some of the most sophisticated consumers of legal services. And corporate legal counsel are investing in these tools. They know that I can do half the work ahead of time and hand it off to them and get it. They’ll have them do half as much work as they used to do by using these tools. And if they give me something back and charge me 15 hours that I know should have taken four, because I know they have Harvey available and these other tools, they’re gonna push back and they’re gonna, I’m not gonna pay for this. And you’re also gonna start seeing that with insurance companies. I used to do insurance.

I used to do plaintiffs and defense litigation work. Some of it was the defense work was insurance funded and they’re the same way. That’s another very sophisticated buyer of legal services that pushes back on the billing, ⁓ billable invoices that come to them. They cut line items. They won’t allow you to bill for certain types of work and they’re going to be doing the same thing. So once you start having these more sophisticated legal purchases of legal services pushing back and it’s already happening, I think you’re going to start seeing it.

Force, especially when boutique firms that spin off and are able to compete and do AFAs and flat fee arrangements, they’re going to start undercutting the large law firms. So I think that it’s already happening and you’re going to see it happen a lot more next year. And I think that’s when they’re going to actually start to see the return on investment. I don’t think you’re going to see a return on investment with billable hour but as Steve Embry would say on Bob’s show, and he comes from the big law background.

There’s a lot more to it than just hourly billing. The entire system of a law firm is set up on that hourly billing framework, so it’s easier said than done. I’ll acknowledge that as well, but we’ll see.

Greg Lambert (50:56)
Yeah.

It’s also the only system that the general counsel tends to know. so, and that’s the only system they’re rewarded on because it’s not what kind of AFA did you get us, it’s what kind of discount did you get us off the billable rate? So the incentives needs to change across the board. I will take my privilege here and say that I do think the first return on investment is gonna be back office.

And I think you’re going to look at, because you see it in the Fortune 500 companies right now, the ones that are coming out and are announcing, and they’re doing it this quarter, announcing benefits that AI has brought to the bottom line for the company. One of them was a transportation company that was using AI to determine how best to ship back empty containers.

that have gone from one location to the other. And they said they used to take a person three hours to do it. And now the AI can do it in seconds and give an answer. It can be all automated. the people are cut out of the loop. And the efficiencies are astronomical. And it’s increased their revenue by 20 % in this one area because they were putting AI in it. So I think.

you’re gonna see more of that back office just because the billable hour is such a monster to fight. ⁓

Sarah Glassmeyer (52:22)
Yeah, that’s

exactly what I was gonna say. Lawyer work tools, ROI not so much anytime soon. Backoffice tools, absolutely, yeah.

Marlene Gebauer (52:30)
So we’re talking about information and exchange of information and keeping informed and, want to talk a little bit about sort of how we’re doing that as an industry. So, you know, the other day we, we were discussing the 20th anniversary of the CLAWBIES and sort of fragmentation of, of social media. And, you know, Sarah, I’ll direct to you first, like you’ve been very active in the migration to blue sky. And so.

Is the legal tech community still a cohesive thing or are we, are we sort of fracturing, into sort of, different areas and, sort of different discussion groups, you know, and does it, does it make it harder to kind of call out the BS that you like to identify?

Sarah Glassmeyer (53:14)
I unfortunately, I do think the legal tech community is not yet a blue sky. I mean, it’s definitely getting better like as a platform, because I mean, that was the one thing during the World Series that was a real like, there are like lot of people here now and it’s becoming more active. It’s kind of unfortunate that I feel like part of the problem is, is that so many people got used to being kind of just free writers that, you know, the law Twitter community, there had.

You know, people joined and it took, again, kind of same thing. It took several years to build up and it was be able to be a slow burn. And some people were able to just kind of just watch and they were like, I’m on Twitter and I’m participating, but like they never posted actually. They maybe sometimes only just like interacted and liked, but that’s the thing about social media. Everyone has to contribute, especially at the beginning to make it seem like a very interesting ecosystem and for there being exchange of ideas.

I know it’s one of those problems that I really wish there was, like, don’t know if we, again, cause I’m old that we needed to bring back like message boards or something, like some sort of way. Cause like LinkedIn is terrible. I can’t find anything on LinkedIn. I never see anyone I work with. I see like posts that are like three weeks old, or it’s just like clearly generate AI slop from like marketers and stuff. There’s the, this legal tech sub Reddit. And that’s just basically like who hates Harvey this week. it’s like, you’re not that useful for me.

Marlene Gebauer (54:29)
The haters.

Yeah.

Sarah Glassmeyer (54:30)
Yeah, I refuse to look at Twitter because every time I do, I almost get recruited into the Aryan Nation that there’s always some posts like that, like right in my feed. I’m like, I don’t need to see that. Yeah, I mean, there’s no real good, there was like a kind of a semi-active link. You had a Facebook group. I think you actually ran it out. I was like talking about that out loud. I’m like, I think actually Greg was the moderator of that. was like a Facebook group of like legal tech or law librarians and tech.

Marlene Gebauer (54:52)
It’s legal marketing, legal

marketing, right? Yeah.

Sarah Glassmeyer (54:54)
Yeah, but like, really

wish there, I mean, honestly, what I really do wish is that people will get back more into blogging, says the girl that hasn’t published a blog post in like six months. But like, I think there’s so much going on that requires, because I mean, legal tech or social media, even like LinkedIn posts and use short form social media. And I’m actually pro, I wish I had more time to experiment with like TikTok or Reels on Instagram, you know.

All those quick bites of like 35 seconds, 40 seconds, 140 characters, 280 characters, whatever. We can’t really cover what’s happening in that amount of time and have like real.

substantive discussions that we need to have. And so I wish more people would just start blogging again, because I mean, again, I feel very fortunate that I kind of came up at the time when people first started with blogging. You know, we all use like geek and law blog, and there’s a whole group of law librarian bloggers that we all then, you know, kind of migrated Twitter. But we had that base first of kind of more and we went back and forth between our long form thoughts and our short form thoughts. And now everyone’s just like spitting out hot takes without any sort of context or follow up or

I’m packing. It’s just like…

Greg Lambert (56:01)
That’s what the algorithm likes though.

Marlene Gebauer (56:03)
Mm-hmm.

Sarah Glassmeyer (56:03)
Yeah,

I know that’s the other unfortunate thing. this, I mean, I absolutely believe the world is where, one of the reasons the world is where is that because Google reader got taken down.

Greg Lambert (56:11)
March 2012,

the world has gone to crap since.

Sarah Glassmeyer (56:15)
Yeah, it’s just like, without that, I mean, it’s really shocking, like how it’s so hard to find other people. I mean, it’s, you know, you’re really luck of the draw and then like, at least probably once a month, I’ll find someone, I’ll start reading their stuff on LinkedIn, you’ll go back and read other, but I’m like, where’s this person been all my life? I want to read everything they’ve ever said.

And then, yeah, and it’s hard to find these people again. There’s no, and then maybe it’s also just because the world is so big that there is no like, you know, LinkedIn again, there’s no dedicated stream of just legal tech content. can’t do hashtags. We can’t do groups. It’s just all everything.

I mean, I guess if I win the lottery, I’ll just start like a legal tech social network. But until then, just plugging away at my job. But yeah, I really feel like we are losing out as an industry that we don’t have a place to like centrally share ideas. Because that’s the thing, like, so much of what I miss about the initial social media landscape is not only just being able to like spout my mouth off and get to share my ideas, but I really did learn from other people and just like seeing new types of like ways of thinking about things or even like hands on getting your hands dirty.

Niki Black (57:10)
Okay.

Sarah Glassmeyer (57:19)
trying this? I mean, we have no like really good place to exchange ideas like that anymore.

Niki Black (57:25)
And I don’t think that we ever will again, because back then, I started blogging in 2005. back then, there was such a small number of the legal space who were tech savvy enough to be online. So you really did have this really interesting combination of tech savvy people that had a legal background. And as everybody ended up online, Twitter with the hashtags was okay, but even before

Greg Lambert (57:40)
That is true.

Niki Black (57:54)
Musk bought it. I had gotten off the platform because I no longer received… You had to create your own list because I have a first world problem or maybe that’s, don’t know the right word to call it, but I had 22,000 followers on Twitter. even if you follow 8,000 people, I didn’t even look at my main Twitter feed because it was just so much stuff. So you’d have to create lists and then I forget what list I created that meant what. And so I just kind of gave up and got off Twitter.

And I went to LinkedIn for any number of reasons. and I have like 206,000 followers on LinkedIn. So I kind of have a, I’m, but I’m very lucky because first of all, if you click on the bell so that you’re following people, they will show up more often than not in your feed and also in your notification. So you need to follow them specifically. And that helps a lot. So you can kind of curate a list, but also I’m fortunate enough that I have enough.

Followers that are interested in what I write, like a lot of what I post on LinkedIn is just a summary of an article that I’ve written or a podcast that I’ve been on. So it’s some kind of content that I’ve created about something I’m thinking about because I still write even though I’m not blogging. the people that follow me, the ones that actually follow what I say all the time versus the 206,000 that follow me are very active in my comments. So they’re constantly commenting, making substantive comments. from all across the legal tech sphere.

You know, you actually usually find more comments than likes on all my posts. There’s a significant amount of engagement there. So I’m kind of lucky because there’s a community of sorts that’s sort of involved in my own posting and partly because of the amount of content that I put out, I think, I don’t know. So for me, that’s kind of filled that gap, but I don’t think we’re ever going to be able to have a dedicated legal tech space because there is knowledge management. There are librarians, which is distinct. You know, there’s eDiscovery, there’s ⁓ the AI people, then there’s just…

Legal tech in general, there’s consultants, there’s marketing. And it’s almost like they all need their own spaces with a little bit of overlap. you just can’t do that anymore because everybody’s online now and everybody didn’t used to be online. So it’s an unfortunate success story, I think.

Greg Lambert (59:53)
Yeah.

All right. Well, before we jump to the crystal ball question, let me ask ⁓ real quick, since we’re on this Sarah, what’s one go-to resource that you’ve been using in 2025 that you find indispensable?

Sarah Glassmeyer (1:00:13)
I was going to take a page from Time Magazine, first of the year. I want to say librarians. Listen to librarians, but especially there’s a blog called AI Law Librarians. Pretty much everyone who writes for that blog, find, it’s in a good mix of very product overviews or thoughts or even just how to use AI in your general life. But it’s really thinking about AI from someone who comes from a knowledge world perspective.

And very, again, not overly hype, but also not overly negative. I feel like it’s a really good balanced take on AI in the world.

Greg Lambert (1:00:46)
Tell us that one more time. was the title of it?

Sarah Glassmeyer (1:00:48)
AI Law Librarians.

Greg Lambert (1:00:50)
All right, easy. Niki, what’s your go-to resource?

Niki Black (1:00:53)
I would say it’s two things. First, I really enjoy being on Bob’s panel. And I think that me being on it aside, it’s a great place to hear from the people in legal tech space that cover it. What’s trending? That’s every Friday at three, Bob Ambrogi’s Legal Tech Week panel. It’s just a great place to get a sense of what’s trending and what everyone’s talking about and the big issues. But then in terms of a single person, I saw that this was a question on the list of things you were going to talk about. And the person that immediately came to mind for me with AI, especially

for smaller firm lawyers was my co-author Carolyn Elefant She consistently, she doesn’t blog anymore. Again, like you said, Sarah, everyone’s kind of abandoned their blogs and she was one of the OG bloggers in legal tech or legal. But she consistently posts on LinkedIn about the way she’s using AI, the way lawyers can use AI. She has like videos, she has analysis, she has white paper. She’s constantly creating, like doing these deep dives that are how-tos and…

Like most recently, she was the one that had got me thinking about, ⁓ she was talking about Google, how she used Google Notebook LM to create an analysis of her position that lawyers can start to use the business tiers of the general use platform, AI platforms, because of all the protections involved in this Google Notebook LM video.

analyzed exact took her analysis and put in this really user friendly way and it’s attached to her LinkedIn post. But so I love that because Google notebook LM is a fantastic tool. just wrote about it myself for my daily record article. And also, I should just provide some great information about practical use of AI.

Greg Lambert (1:02:29)
Marlene, going to ask you, what’s your go-to resource this year?

Marlene Gebauer (1:02:34)
seeing you put me on the spot, I wasn’t ready to answer the question. So let me see. I mean, I do use LinkedIn a lot. I have been listening to a number of podcasts.

Greg Lambert (1:02:36)
Yeah.

Marlene Gebauer (1:02:45)
it’s a bunch of sort of futurist thinkers and they talk about all kinds of things, not just gen AI, but they’re talking about like quant computing and sort of where things are going and they look at it from multiple perspectives. And I like that because it’s sort of outside of the, ⁓ outside of legal. And like, I, I find that helps me because I like to connect.

Like different ideas from, from different environments

Greg Lambert (1:03:12)
The one that I’ve been listening to a lot and reading is there’s an AI guru kind of guy named Nate B. Jones. And if you Google Nate B. Jones, you’ll come up, but it’s same as with you, Marlene. It’s much more about what the foundation models are doing.

⁓ what, what’s working, what’s not working. And he’s really good about as new versions come out to talk about how to get the most out of them, what they’re, what they’re doing and what you can do in a chat GPT 5.2 that you couldn’t do in 5.1. and it’s amazing how you can then apply that to your day to day, and go, yeah. And I was like learning that chat GPT 5.2 is.

agentic at default makes you understand, okay, if I know I can do that, then I know it can go out and do, it can go out and search websites, but also verify that the information that it’s getting is right and can go back if it’s not. it makes the user experience for me a little bit different by knowing the underlying, kind of features of what these things can do. So, all right.

Marlene Gebauer (1:04:22)
I found it. It’s moonshots.

It’s moonshots.

Greg Lambert (1:04:24)
Moon shots. All right. We’ll put that on the show notes.

Sarah Glassmeyer (1:04:29)
I thought of another one if I can pop in with a third. Okay, this is more about general tech industry as a whole, like the business side of things is called Ed Zitron’s, it’s a newsletter. Ed Zitron is the author and it’s “Where’s Your Ed At?” I think if you just Google Ed Zitron, it’ll come up. But I mean, it’s terrifying. You would talk about, are we in a bubble? Why hasn’t the bubble burst?

Greg Lambert (1:04:30)
All right. Well, Niki got two, so you get to get two.

Marlene Gebauer (1:04:31)
Yeah, sure.

Sarah Glassmeyer (1:04:50)
looks at, know, like, hey, OpenAI is still into Nvidia, Nvidia is still into OpenAI, like, how, like, really shows how maybe a little bit of a house of cards, if not a bubble. But it’s just, it’s really interesting for someone who’s not necessarily business minded to hear, like, that perspective and understand, like, what’s all going on on the business, because that’s, you know, all of this, if OpenAI goes, a lot of legal tech goes as well. So.

Greg Lambert (1:05:00)
It’s a little Enron.

Marlene Gebauer (1:05:02)
Yeah ⁓

Sarah Glassmeyer (1:05:15)
Yay. That’s what I thought. That’s like cheery thoughts that I every now and then think about right before I’m going to bed like, oh dear.

Marlene Gebauer (1:05:23)
So, all right, sorry, you teed me up perfectly for the crystal ball question. So thank you for that. I’m going to ask both of you, speaking of bubbles, like what is the single biggest bubble that’s going to burst in 2026? Niki.

Niki Black (1:05:36)
I definitely think it’s some sort of financial bubble related to AI. I don’t know what it’s going to look like. But I think that on a broader scale, of a across the board recession that’s potentially caused by some sort of AI bubble that’s just going to trickle down into our industry. I don’t think 2026 is going to be pretty. I’m not sure why, but I have a very foreboding feeling about 2026.

So, and more on a, again, a broader view, not just legal tech, but I think it’s definitely gonna filter into legal tech and it’s not gonna be pretty. I can’t be super specific, I just think it’s gonna be bad.

Sarah Glassmeyer (1:06:18)
Yeah, mean, again, from the Ed Zitron newsletter, was one of the, I forget which company released the new image generator, and he like kind even did the math like, okay, each one of these is costing the company this much money to make because you’re just based on the compute power. And I mean, there’s so much that can go wrong with the general AI industry from data centers and like not getting those built far long fast enough and.

Greg Lambert (1:06:18)
Sarah.

Sarah Glassmeyer (1:06:39)
chip availability and then just the fact that all the financial, like how much money OpenAI is burning still on a daily basis. Yeah, and then the fact that, know, in previous, you know, when pets.com went under, Thomson Reuters did not even notice. If OpenAI goes under, what’s that going to be for the Harveys of the world? Or, you know, the fact that there’s so many legal tech companies and like, they’ll be monolinguistic, but they still are using, you know, OpenAI and Claude and various ones that it’s like,

What’s going to happen there? I also do worry talking about how we were like earlier talking about the ROI You know one of the neat things of the past three years has been the ability to spin up a legal tech company like the barriers have been loaded Lower greatly like you really didn’t you just needed an API connection for a lot of these companies You could just make a legal tech company And it’s getting to the point like there is you know, I have 3,000 listings currently on the legal tech hub that

next year there’s not going to be 3,000 plus teams or there will be like that many like you know think we’re going to start seeing some of these littler companies that never really made a ⁓ grab into the market never really got funding that just kind of ran out of

their runway and they’re probably gonna fold or get absorbed in some way. So I think we’re gonna see kind of like that mini bubble, then, but yeah, like if the big one goes, I don’t know, it’s gonna be like the volcano under Yellowstone. think we’ll have more things to worry about than legal tech going bad. It’ll be a bad thing.

Greg Lambert (1:08:05)
⁓ All right, I’m going to jump in and make my prediction now So I was listening to something coming into work this morning and it talked about how reliant companies are on essentially the three or four big AI companies. So we can probably say OpenAI, Google,

Anthropic and X with whatever they’re calling it X AI because they have their own infrastructure and companies like META are using open source Chinese AI systems to try and fill in the gaps that they have with their own systems because it just hasn’t worked.

I my big, big hairy audacious projection, my b-hap, is the meta, I think, is in more trouble than people think it is. And so it’s going to be interesting to see if they fail on producing or keeping up with the AI siblings that are out there, cousins, I guess.

I think they’re in trouble. And, and so, you know, it’s, and if a big company like Meta can be in trouble, just think of the other ones that are trying to develop their, their own AI systems or to develop something that’s not tied to those, those four biggies. they may be in trouble as, as well. So Marlene.

Marlene Gebauer (1:09:32)
So I think next year you’re going to start seeing the impacts as, ⁓ and this is something that I listened to on that podcast moonshots about the economy of scarcity and the economy of abundance. And while I am a strong believer that there will be new jobs and new things for people, think, next year and possibly the, you the year after we’re going to see a lot of disruption there as, you know, some of these, these new technology tools,

you know, become better, start doing some of the basic work that people are doing and, you know, people have to find themselves sort of a new spot before, you know, we sort of moved to this, you know, everybody gets their allocated, you know, stipend per month or something. So I think you’re going to see some of that and hopefully you’re going to see organizations preparing their people for that.

Greg Lambert (1:10:26)
All well, Sarah Glassmeyer, Niki Black, I want to thank both of you very much for coming on for year two of our year in review. Thank you very much.

Niki Black (1:10:33)
Thank

Sarah Glassmeyer (1:10:36)
Thank you for having me.

Niki Black (1:10:36)
Thank you. A

lot of fun.

Marlene Gebauer (1:10:40)
And thanks to all of you, listeners, for taking the time to listen to the Geek in Review podcast. If you enjoyed the show, please share it with a colleague. We’d love to hear from you on LinkedIn, TikTok, and Blue Sky.

Greg Lambert (1:10:50)
And Sarah, where can people find out more about you?

Sarah Glassmeyer (1:10:55)
Probably the easiest way is to go to sarahglassemeyer.com, my blog I’ve had since 2007. And they are linked to all the social media places that I flit around on. But yeah, think LinkedIn or Blue Sky are probably my most active right now.

Greg Lambert (1:11:09)
All right, and we’ll be expecting that blog post at the beginning of January. ⁓ And Niki, how about you? Where’s the best place people to find you?

Marlene Gebauer (1:11:12)
Hahaha

Niki Black (1:11:18)
Definitely at this point, LinkedIn’s probably the easiest.

Marlene Gebauer (1:11:23)
And as always, the music you hear is from Jerry David DeSica. Thank you very much, Jerry.

Greg Lambert (1:11:27)
All right, bye everyone. Happy New Year.

Marlene Gebauer (1:11:29)
Happy New Year!