Latest / Legal Late Night / Is AI a Bubble? The Future of Legal Research & Hallucination Checkers (feat. Steve Tover)

Transcript

Is AI a Bubble? The Future of Legal Research & Hallucination Checkers (feat. Steve Tover) · Legal Late Night

• Jared Correia (00:00): Hello everybody. We've got show the promises to be at least mildly interesting for your listening and watching enjoyment. I'm your host Jared Correia. I'm the CEO of Red Cave Law Firm Consulting. For my monologue. I'm introducing an entirely new segment called Artist Profiles. In the interview, it's Steve Tover of Lex Sphere and any law. We're talking about the state of legal research for attorneys in the counter program. I asked Steve to choose between pairs of other famous Steves in a game called Too Many Steves. Now let's get to our first ever artist profile on Legal Late Night. Jared Correia (00:51): As I was looking at my complete list of perfect albums for my eyes only currently, I was thinking there are a lot of singers and bands I really love who never made a perfect album. In a lot of cases, I would rank some of those artists far above those who achieved perfect album status. So I've decided to build out some artist profiles as a way to spotlight more of my favorite musicians who never created what I would consider a perfect album, which is obviously a pretty high bar. I've also found that lots of folks say to me, that musician you talked about the other day, I'd never really heard of them before, or they know very little bit about them, especially some of the older artists that I like. And so if I can introduce people to new music, new artists, the Legends of The Lost and Found as it were, I'd like to think that that's something of a public service. (01:45): So today I'm bringing forward a singer songwriter that I think relatively few people are aware of these days, who was nevertheless one of the most popular and successful musicians of the 1970s. It's Harry Chapin. Harry Chapin is one of my favorite musical artists of all time. He's unique in that as a popular musician, he wrote really long songs. His second biggest hit ever was almost seven minutes long. Many of those songs are exceptionally bittersweet, varying into black comedy. These are not the kind of songs people were generally listening to on the AM radio in the 1970s, and yet of the 14 singles that Chapin released during his career, all of them charted in at least some capacity. He was one of the top selling artists in the 1970s and was making $2 million every year, not adjusted for inflation up until 1981. Although he did release some traditional pop songs, Chapin specialized in what he calls story songs, lengthy tracks, which told a complete tale. (02:50): I guess he was sort of like the Geoffrey Chaucer of 1970s pop music telling stories and fleshing out characters more completely than you could do in a truncated three minute radio release. Though you think under that description he'd be playing clubs anonymously because who wants to listen to all that shit? TLDR really? Then Harry Chapin may be one of the more surprising success stories in music history. Chapin was born into a talented family with many musicians, including his brothers with whom he performed regularly, both before and after he became a solo artist. His nieces formed a band, the Chapin Sisters, who are still performing today, and his daughter used to be a singer as well. His grandfather was a famous painter who illustrated Robert Frost's early poetry books and Harry Chapin himself was a multihyphenate before committing to becoming a musician Full-time. Chapin was a documentary filmmaker and his legendary champions, a movie about famous boxers with previously unreleased footage was actually nominated for an Academy award. (03:56): Of course, Chapin's debut and output as a musician was also quite extensive. He released an album each year for pretty much the entirety of the 1970s, and he was also a part of your live act having released the greatest hits live record in the mid 1970s, and many of his other live recordings are now available in compilation formats. Chapin's Discography features some unbelievably ambitious recordings, lengthy operatic, dark deep. He deployed a baritone with significant range in his band, which is a crazy look for a pop record. I want to hold your hand. This is not now at this point you may be convincing some frustration with me. This is potentially similar to the four minute mark of a Harry Chapin song with another six minutes to go when you're like, get along with it. Okay, let me get closer to the point and at least half of the point is the songs. (04:58): Now, if you're like, I've never heard of this dude before in my life, let me tell you at least one Harry Chapin song and probably two Chapin's most famous song cast in the cradle about a father who ignores his son and then receives his comeuppance is part of the American songbook. It reached number one on the US charts. If you're a dad, this is perhaps the most frightening song ever written and it will absolutely scare you straight. This song's hook is basically steered into my memory. You know the cast in the Cradle and the Silver Spoon, little Boy Blue and the Man in the Moon when you coming home, dad? I don't know when, but we'll get together then. Son. God damn my soul is wr. You may also know Taxi a Lesser Hit and a lengthier song in the more typical Chapin. Vain Taxi runs a solid at 6 44 and the conce of the song is of a taxi driver in San Francisco who picks up an old girlfriend in his hack as the driver muses on the question of happiness. (06:00): He knows that neither party is really good in what they wanted. The ex-girlfriend who wanted to be an actress is now acting happy in a big old house in an unhappy marriage, while the driver slash narrator who wanted to be a pilot is still flying definitely, but with the help of some illicit drugs, so this is maybe not the cab you want to get into necessarily, but Taxi had been a hit on Boston Area Radio before being released as a single and so Electro Records Chapman's label had a sense that this would be a commercially successful venture, which is probably why his debut album called Heads and Tails features all kinds of taxi iconography on the cover. Now, in terms of his other hit records, there are certainly some good ones in there as well, including Could you put your light on Please? Which is a plaintiff fairly straight up ballad. (06:54): That's the lead track on Heads and Tails. His first album Sunday Morning Sunshine is a peppy love song from his second album, sniper and other Love songs. Yes, that's the real title that was also released in 1972. A Better Place to be is a girth of your tune from the same album, which is an exploration of the intensity of loneliness. This song reminds me of my dad, not because he was super lonely, but since the Night Watchman character works at a tool and die factory just like my dad did. My dad also was a person who introduced me to Harry Chapin via his record album collection, which I still have. WOLD is a little on the nose, but it's the Tale of a Morning Disc jockey chasing jobs around the country. In a series of conversations with his ex-wife, he learns that she is not in fact interested in getting back together and that she has moved on as he continues to move around after it left her to pursue various gigs looking to make himself whole. (07:59): I think you can probably guess where this one ends up. It was inspired by an actual Boston Radio personality who would call his ex-wife on the year. I want to Learn a love song is the true story of how Harry Met Sandy, his wife, who hired him to teach her guitar and eventually ended up leaving her husband for him, yikes as a bad beat. Interestingly, country singer Donna Fargo released what is effectively a cover of this song in 1977, but from the point of view of the Wife Flowers Are Red is a song about how teachers try to get their creative students to conform sort of a perfect song for Harry Chapin. You could say Sequel is one of his last hits. It's a unique song in that it is actually a sequel to Taxi, one of his most popular songs because so many people are asking what happened to the protagonist of the song. (08:52): Chapin wrote the epilogue. Now I love Harry Chapin, but this song is a fucking travesty that never should have been released. Plus it has a happy ending. We're in the Chapin Universe. I'd like to say this was a cash grab, but I legitimately think Harry thought he could turn this into something he couldn't. You can skip this one. Now, one thing about having lots of songs bordering on 10 minutes in length and writing about unconventional topics is that Harry Chapin has a significant catalog of album songs that never charted, many of which are exceptional. As Chapin alluded to himself in concerts, he had trouble getting his songs on the radio. Some of my favorite Harry Chapin uncut gems then are Greyhound. This is a top five song for me, which is a dark two-hander about commuting on the ti or bus line, alternating between moping, passivity and hell bent anger. (09:47): This song is about the destination, not the journey. Any old kind of day is the best song you'll ever hear about a random day in your life where nothing happens. Dogtown is a song about Gloucester, Massachusetts where I used to live about the wailing days when semen would embark on years long perilous journeys and leave their wives isolated for long periods of time if they even ever returned. Just your typical pop song. It's also almost eight minutes long mail order. Annie is yes a song about a man who purchases a mail order bride. Mr. Tanner follows an Ohio laundromat owner who tries to make it big in New York City. He rents out Carnegie Hall. He sings and he's savagely panned by critics six. String Orchestra is a brilliant comedic riff about a no talent guitarist who sounds a heck of a lot better with a band backing him. (10:45): Chapin actually plays the guitar intentionally out of tune on the solo parts. The Rock is a variation of the Boy who Cried. Wolf Dirt gets Under the fingernails is about a mechanic as it gets a glow up for his wife only to find out that she's secretly taking up the messy hobby of painting. In her spare time, the Mayor of Candor lied. Yeah, I know not a lot of subtlety in the title. There is a political epic with a couple of gnarly twists Corey's coming tracks. An old man who works in a rail yard and tells of a daughter no one thinks it's real until she shows up. Dancing Boy is a beautiful song about his young son who learns that he can draw a crowd if he dances while his dad plays guitar. I left off two songs. There are two songs that I believe represent Peak Chapin, neither of which were hits. (11:34): The first is 30,000 pounds of Bananas, which is a tragic comic dirge about a truck driver who's killed when his brakes give out on a steep hill outside of Scranton, Pennsylvania. That's the tragic part, which results in a tsunami of mashed bananas being spread all over town. That's the comic part. The studio version of this song is excellent it's own right, but if you want the full Harry Chapin experience, listen to the version of the song that appears on 1970 six's greatest stories live in which he relays aspects of the making of the song and even performs the two original endings. Before he finally settled on the last published version, he talks about how he stole the guitar riff from the song from Chet Atkins. He discusses this hilarious call and answer refrain with his brothers in a band where every time he suggests a way to end the song, they return with his distinct. (12:29): Harry, it sucks. I think the endings are all good though. He discusses how he tried out a country ending where he suggests maybe something about motherhood this time because the song already had a truck Well played. Harry's in complete control here, flourishing all his powers, interacting effectively with a band, the Crowd and Los Angeles Memorial Choir, which also appears on the side. This is just a great tune. And then last but not least, Harry Chapin's best song is The Epic Sniper, which is the sort of title track off of his second 1972 record, sniper and other love songs. This song is pretty staggering in terms of its breadth and depth. The protagonist of the song is a real person, Charles Whitman. On August 1st, 1966, Whitman killed his mother and his wife and then climbed to the top of the tower at the University of Texas at Austin's main campus, firing on people for over an hour and a half, killing 11 and entering 31 more like Chapin himself. (13:36): This incident is largely forgotten, but at the time it occurred, it was the deadliest mass shooting perpetrated by a lone gunman in the United States history. Chapin's angle here is to try to get inside the head of the shooter. Why would anyone do something like this? What was the cause of it all? In the song, the Shooter intensely prepares for his rampage and engages in it while Chapin inter intersperses others' perspectives about him, where they seek to disin themselves from his actions. The lyrics supposed that the shooter was ignored by his mother as a child and is seeking answers. His conversations being held at the end of a rifle point with his victims. The character feels as if he needs to find a more experiential life and by killing his victims, he proves his existence. It's the gunman's perverse triumph, the high point of his life when he feels the most real as he slaughters instant people until he is finally killed himself In real life, Whitman's father was emotionally and physically abusive to both him and his mother. (14:39): This is a deeply fucked up existential song about the roots of violence in America. Definitely listen to the studio version of this track, but not surprisingly, there is an excellent live version. This song from Chapin's 1975, appearance on PBS Soundstage. Harry Chapin though left a legacy far greater than his music. He was a tireless fighter for eradicating hunger around the world. Chapin saw hunger and poverty in a country as prosperous as the United States, in particular. As an insult to America, not only were more than half of Chapin's concerts benefit performances, it's estimated that he gave away a third of the total of his paid concert proceeds as well to charitable causes. In 19 70 70, he organized a benefit concert in Detroit with other headliners and seventies singer songwriter stalwarts, James Taylor, Gordon Lightfoot, and John Denver. That same year he was appointed to the Presidential Commission on World Hunger by Jimmy Carter and he was the only member who showed up to every meeting. (15:43): His efforts inspired other prominent anti-hunger events like Live Aid USA for Africa and Hands Across America. He was a co-founder of World Hunger Year. There remains the Harry Chapin Foundation, which is provided over by his wife. In 1987, Harry was awarded the Congressional gold medal for his charitable contributions. It is not an exaggeration to say that Harry Chapin was one of America's greatest philanthropists. Now they say that only the Good Die young and Harry Chapin died on July 16th, 1981 when he was only 38 years old. He was involved in a car accident on Long Island Expressway when he was rammed from behind by a semi-truck after he was changing lanes erratically and slowing his speed. No one knows exactly what happened. It could have been that Chapin was having car trouble at the time, although he may have gone into cardiac arrest during the accident. Unsurprisingly, Harry Chapin was on his way to a benefit concert at the time. Next up is Steve Tover of Lex Sphere in any law. We talked about the modern state of legal research, how law students research now and whether AI is a bubble. So let's pop that bubble right about now. (17:04): Well, I've effectively run out of things to say which is awkward because this is a podcast, so I'm just going to take a little nap. I'm trusting you all to wake me up in about an hour. Okay, you think you can do that? No, I'm just fucking with you. While I would dearly love to take a nap, now is certainly not the time because we've got legal research to talk about everybody and who among us is not entertained and intrigued about legal research. If that's you, which I know it is an attorney, have good, great news. Our guest today is Steve Tover, who's the CEO at Lex Sphere and any law. Steve, welcome to the show. How are you? Steve Tover (17:43): Hey Jared. Thanks for having me. Doing great. Jared Correia (17:46): So Steve, we've known each other for a little while. I see your stuff on Link LinkedIn, I don't even know if I'm fully aware of what Lex Sphere is. So would you mind educating me and everybody else out there about what this even means and does? Steve Tover (18:07): Sure, happy to. As you know, we started originally with any law, which was an alternative legal research platform to some of the big boys, Westlaw and Lexi Nexus. To be specific, (18:22): That is a B2C play for either consumers or solos to be able to do legal research on an individual basis. And the idea was to offer something that would be, and originally it was free and then offer something that would be on a low subscription basis. Over time, we received a lot of interest in enterprise type models for data and for hosted legal research services. And in the last six months we've pivoted to that direction specifically so that we have something that we can work with enterprise partners to be able to offer them a standalone hosted service or an API for companies that want to be able to offer their law firm clients a legal research service as one of the add-ons. And then of course for data for companies, especially now in the last couple of years, the AI companies have gone so far in their development and their reach and now their really idea of having good data has become something that there's a big need for. We're sort of filling that breach and that's where L Sphere was born for those two purposes. Jared Correia (19:39): So it sounds like some of this was driven by a desire from partners, potential clients of yours, and some of this was driven by advancements in ai. Would that be fair to say or? Steve Tover (19:51): Yeah, I think so. I mean there's been a couple of reasons why this has happened. So much of the legal research space has been consolidated in the last five years. When we first started the company about seven years ago, we really were one of about eight or nine companies in the legal research space. Since that time, there have been several acquisitions. There have been a couple of companies that have closed and specifically very recently, one of the major acquisitions was when VLE acquired Fast Case that had previously merged with casemaker and they were sort of in the level that we're dealing with now. Cleo acquired the whole group. So Fast Case was one of the main providers of data servicing a lot of the companies out there that were providing some type of legal research service. And then along the way, all the AI just kind of a lot of investment in that space. (20:50): And again, it goes back to the massive need for the reliable data and we're really one of the only sources left with that. And we've had a lot of approaches in the last year about kind of filling that gap because the players that were providing that data previously are not really doing that anymore. So we're in that stage now where in the last few weeks we launched L Sphere to fill that gap and we have several opportunities talking to companies that are interested either in the data to either train their systems or to be able to use a platform to provide some type of integrated service on an existing platform that's for law firms or lawyers and visuals that support them. Jared Correia (21:39): Gotcha. That all makes sense. So you're hitting on a lot of stuff I want to talk about, so I want to expand on some of what you just said, but before we get there, can you talk a little bit about any law, as you know, there's been a lot of changes in the legal resource space, but it sounds like it's still valuable to have a product out there that can kind of be useful for consumers and also the smaller firm lawyers. Steve Tover (22:02): Yeah, I mean when we originally started any law we thought it would really be solos and smalls that were coming to us and when I say smalls like under 10, three or four partners in a firm that were litigators that required something that might be too costly going to sort of the big boys in the space. What we found was that a lot of individuals, especially when it was a free product, we were getting a lot of litigants, people that were representing themselves in court looking up things online, get a lot of people just looking up their name online and they find it in Google and then Google points them to us and other places that have their name or cases listed as AI has become bigger and bigger. We even integrated AI onto the platform. So for example, we added in addition to standard legal research service, which would include case lookup citation linking. (23:00): So to be able to track citations in cases to be able to do analysis, understanding how deep a case goes and tracking it back and forth to recently adding AI summaries and AI holdings so that people can more quickly identify the cases they're looking for. But what happened in the last year or so is so much has moved to how do you use AI to make legal research faster and more efficient. And co-counsel, when Case Text had developed Kara and co-counsel, they really pushed the envelope on what they thought lawyers would use and they really are the ones that broke the mold on that with the acquisition of VL by C, Cleo VL has Vincent, which is their own AI component. Companies like any law or L sphere, we're not developing any ai, we're using existing ai and so we'll use an API right now we're using an API by open ai and we'll use that to do queries with prompts that we've prepared that will enable the types of things that I've described, like summaries and holdings. (24:11): This next phase of where we are, we are actually implementing other AI on top of our data to allow for analytics and pattern recognition, how judges rule, how courts rule types of decisions and jurisdictions, attorneys track records, things like that. We're also building out an AI tator because we already have certain functionality that allows for that and currently the best Tator Shepherds and if you want to get to that level, then the next level is going to have to be probably utilizing something with ai. And I think there's going to be alternatives to that coming out, not just by any law or L sphere, but by a lot of the companies that are out there because it enables so many more things really that you can do with the data. Jared Correia (25:02): That's interesting to me. So you've got these AI features and you're kind of servicing some lawyers and some consumers. I'm interested to know do consumers use AI legal research features? They have access to it, Steve Tover (25:15): So we're moving away from that because they're not using the same level that a lawyer would or a paralegal would. The actual research features like proximity operators or linked citations, they're not trained in using that where the summaries come in handy, but we're finding that in the last year or since we've moved to the subscription model that really the majority of people are really more professionals by far. What's interesting though is if you look at the whole access to justice market, there's a big need for some solution that would help the consumer market. We know there's like 30 million people that represent themselves pro se every year in the United States according to statistics, and there are a lot of nonprofit organizations working to help those people. So there's a need for it and there probably will be if either us or someone else will come along with a service that would not require deep training and AI is a great way to do that with anticipated questions or target topics, things like that, that would help them probably more than the standard legal research platform. Jared Correia (26:36): Got you. That's fair. All right. Now the other thing you brought up, which I think is intriguing as well is so I about this part of it, but yeah, so for those who dunno, CASEMAKER and Fast Case, they were providing legal research services through bar associations in the United States and Fast Case basically overtook Casemaker and then Fast Case was acquired by V Lex and then VL was acquired by cle. It's like one bigger fish eating the smaller fish over and over again. Steve Tover (27:04): Yeah. Jared Correia (27:05): So Steve, I know you're obviously biased in this space, you run a research company, but what do you think that does to the legal research market in both the United States and Europe? Because v Lex is a big player in Europe obviously. What are the knock-on effects of all those mergers? Steve Tover (27:26): So that's a really interesting question. I'm not sure I'm the best person to answer, but I think that in some respects competition helps drive prices down. Supposedly it breeds quality. Obviously LexiNexis and Westlaw are considered the gold standard. They have a reputation. Westlaw is the one issuing most of the citations anyways for the primary citations, the way the courts work. So there's a lot of trust in West and Lexus is certainly with the top law firms. I'm not sure what's going to be and how Cleo's planning to use the vle Fast case Casemaker case law itself. But Cleo's a very big player in the practice management space. They already have a good share of the market (28:22): And so they obviously see something there that will benefit their user base by having some type of integrated platform. I'm not sure what's going to happen sort of at the lower end of the market, again with the solos and the smalls because a lot of concentration is happening up at the higher end of the market as far as the bigger firms who can afford to pay more. I think we're hoping that what it will lead to is an opportunity to maybe grab some market share on the legal research side for the smaller firms and the solos that need access to this type of thing. I don't know what's happening also with the bar associations. I just don't know what's happening with Fast Case and Casemaker who had all those relationships with them. I don't know if Cleo I heard, but I'm not sure it's true. (29:14): I heard Cleo's not going to be continuing those, so there might be an opportunity there, but there'll be other players that come up. There are some AI companies that are turning into legal research companies with data and their own approach to legal research, not classic legal research the way that our platform was originally designed or some of the original platforms. But even if you look at something like a co-counsel, co-counsel is now being offered across the board not just for legal research but on all of Thomson Reuters products and services. So this consolidation, it's not just legal research companies, it's consolidation of data with technology and along the way there's other things that people benefit from. So you have the legal research companies like us, but now you have publishers. There's a lot of publishers out there, right? Bloomberg, Walters, Klu, they have really good data and a lot of the AI companies are gathering this data to be able to offer something comprehensive. So I think in the long run it'll probably lead to higher quality products. Hopefully it'll bring the price down and maybe it'll leave a window open for us to grab a little bit more market share for where we are Jared Correia (30:26): Clearly was trying something interesting in merging like business management and research together, which has never really been tried at this level. And then like you mentioned with the Bar association stuff, as of right now, my understanding is that those relationships are going to stay in place, but who knows where that moves to the future. Okay, so one thing that I thought was interesting that you mentioned too is so you were talking about how some AI companies are becoming research companies. There are some research companies that are intentionally becoming research companies that are built in ai, and do you think there is potential to grab market share as an AI first research provider versus Lexus and West, which seemed to have weathered every storm to this point, including the cloud? Steve Tover (31:23): Well first of all, if you look at Harvey, right, Harvey made that deal with LexiNexis. So they've developed, if I'm not mistaken, I think they developed an API for them and they're using the Lexus is using the Harvey API to run their data so that they're now providing AI case text was acquired by Thomson Reuters. So these companies that were AI first have already become integrated into those platforms. So I think that's how the big boys have answered that question. The Clio acquisition is opening a whole new direction. We've talked to a lot of companies that are service providers at the enterprise level, so practice management, eDiscovery, and there's a real interest in they say, look, we already have a client base. We'll add another thing to the client base. And this brings the question of separating or combining the idea of the business of law with the practice of law, which is a phrase I heard a lot recently. I was at a conference about a week and a half ago and that came up several times. The idea of the business of law is the practice management, the billing, time management, all of those things, case management, but the practice of law, legal research falls into obviously the practice of law and that's where we're seeing the consolidation go next. And that's going to be interesting to see. Jared Correia (32:52): That's a good point. So Steve, in terms of the AI stuff in legal research, obviously the legal research providers were in on AI fairly early. What do you think is still left to be done there? What features do you think come out, where are we headed with this stuff? Steve Tover (33:14): I mean, I could see from what we're doing, I think that the thing that's going to probably have the greatest value is using natural language (33:23): And having semantic search that it's very intuitive. I don't think it's there yet. I think I was at this conference and the person who gave the keynote spoke and he said, the stage that AI is at today is the worst AI we're going to see. The AI is getting better, the infrastructure is getting better, the things that it can do we haven't even dreamed of, right? It's going to get there. When it gets to the point where the semantic understanding of what somebody's trying to do when they're using the ai, it is going to be amazing. It's going to be very effective to help move things along. And that's I think where that's probably going to be the biggest jump so that somebody can talk about a topic or an issue to the AI and immediately get either cases that they need if we're talking about legal research or even summaries that would help them quickly understand if this is something that might have value and eventually getting to the point of saying, these are the legal issues that you should be aware of in these cases and this is why this case might be more effective. (34:36): And then combining it all with things that have been done from a pattern standpoint. If you're going to a specific judge that has a track record. And I think that as that becomes more accurate and we get away from hallucinations as we know them today and we get to the point where you can really rely on the AI to really deliver good stuff, I think that's going to be a very big leap in the legal research space for sure. But in general, in the practice law, Jared Correia (35:04): Yeah, there's some really cool stuff on the horizon for sure. Well, yeah. Let me ask you about this because you just brought it up, the hallucination thing. I think that's one of the reasons why lawyers have not necessarily aggressively been adopting AI legal research in particular. As somebody, as a research company, are you just looking to get better every day in terms of eliminating hallucinations? Is there a certain percentage you're looking for? What do you think it would take for lawyers to reliably use AI legal research without worrying about hallucinations? Or do they always have to worry about it? Steve Tover (35:46): So I think you have to compare the AI as a tool that's going to help you whatever professionally, just like you would use an associate, I don't think anybody would use an associate to do their legal research and not check their work. Certainly not a first year associate. (36:05): And I think that's how you have to look at AI at the current stage that it's at. It can't be trusted by itself. So that's number one. The lawyers at the end of the day, they're the professionals. They're responsible for the work that their associates put out and they're responsible for checking their work in Word, even if they have a spell checker. AI is the same thing that it's effective, it's very good at processing information. And as the LMS get better, the processing will get much better. It's already much better than it was just two years ago. From our personal standpoint, we have certain things that we do to reduce hallucinations. Number one, we only use our own case law because we have a massive library of case law. We don't go outside of our library, so we're not generating new concepts or new citations or new cases. We're only using our ai, but that doesn't mean that there can't be hallucinations, so somebody still has to check it. What we have done is, and other companies are developing this as well, we've developed a tool which is basically a hallucination checker. Jared Correia (37:10): And Steve Tover (37:10): What it is, it's a citation validation functionality, and you can upload, if you use AI to generate a brief, you can take that brief and upload it to any law, or you can take that brief and pull out the citations and upload those to any law application that does this, and it'll check those citations against all of our cases in our library, and it'll then identify if we have 'em. And if we don't, we will point them out to you that you should check these further. Because that's really the biggest question is identifying the anomalies and say, wait a second, this one doesn't look real. Let's look deeper. There's other places you can look for that type of thing. So I think that the H nations are real, but at the end of the day, it's like anything else, any other tool, and they're going to be important to identify, but at the end of the day, the lawyers are the ones who have to do the work and just be careful of what they're putting out in the documentation that they're creating. Jared Correia (38:09): Let's bring this back to the consumer side for a second. Now, I know that you have a product, especially since you've adopted the subscription model that mostly lawyers use, but in terms of lawyer work going forward, is AI legal research for consumers going to be a problem? I hear this from lawyers all the time. Why would I hire clients telling them why would I hire a lawyer? I'm just going to go look at my legal question on chat GPT, and the answer seems pretty good. Steve Tover (38:43): Okay, and then what if you're litigating and you're representing yourself at some point in time? I mean, unless you're really, really smart, at some point in time you're going to need somebody to help you. And we're very careful and other companies are as well to say, look, if this is generated by ai, make sure to consult an attorney because no matter how good the self-represented litigant is, they have no training. In addition to just finding a case that might be helpful for you or finding a law or a legal issue or an insight, somebody still has to write the brief. If you're presenting to the court, somebody still has to understand the court procedures. (39:25): Where are you going to get the statutes and regs from? You're going to go ahead and understand them by yourself, then that's fine. If you can do that. Most people can't. And a lot of people that are self-represented actually don't even have necessarily good language skills. There might be people that are here from another country, they might be somebody that might not have a high school education. So I think that is a pretty big leap to make. I don't think the AI is there today to help people in anything, automated documentation, legal research, any of that stuff. I think you really need help, unless you're going to do something, have a very specific website that does one thing, do not pay where you want to fight a traffic ticket or you want to fight your gas bill or something like that. Very much designed for consumers who don't have to know anything about the law, and it does a very specific function for them. I think if they want to get complex in the way that we're talking here, the best thing they can do is get educated on the issue and then maybe use the AI or use some tools to identify lawyers that are expert in that area of law, in that jurisdiction as well, and then connect with them correctly. And I think that's what's going to happen more and more is connecting people to the right people using better intake like Gideon, that type of thing, and using the data from that to match people to each other. Jared Correia (40:55): Good news, everybody. Steve, not worried about you losing work to ai. All right, Steve, I got two more questions for you. Sure. The first is ai a bubble, which you've been hearing a lot about lately. And if it is, doesn't matter. What are your thoughts on that? Steve Tover (41:10): So there's a difference between the technology of AI and its usefulness and the investment in ai. So I don't think that the technology is a bubble. I think the technology is real and it's going to be a jump, just like the cotton gin and just computers and just like the internet, this is a big jump. It's going to affect everything, and it already has. It's not even close. I don't know if the valuations are going to continue to explode the way they are. (41:46): I was reading last week about three specific investments that I know about in the AI space. There were like 50 x, 60 x and 80 x times revenues, and that's pretty big. Usually, maybe it's seven to 10, maybe 20 if it's real high expectation. So there's a lot of expectation that there's going to be a big, big jump, another jump in valuations. But you have to keep in mind that in the last probably five years, three in particular, but the last five years there's been, I don't know, I think over 50 maybe close to a hundred AI companies just in legal tech that have started that at some level use AI or developing ai. What's going to happen is there's going to be massive consolidation. So there's not going to be 50 or 80 companies. There's going to be three or four, and those are the ones that people are betting on. That's the ones that are getting that kind of valuation. So I think there will be some companies from my own unfortunate experience going through the first bubble in 2000 and 2001, it was my first startup. It's going to happen. There are going to be companies that are going to close. We've already seen it with Robin ai, (42:57): Which was a very potentially successful company earlier this year. Everybody was excited about them. But you have Harvey and Lara and GCAI and philanthropic, I think, and you've got a lot of them. I don't see there being 50 players in another two, three years. There might be three or four. They might be focused on specific areas, but I think that you have to separate out the actual technology from investor mentality. Jared Correia (43:24): That's fair. Alright, last question for you. If I'm a law student and I'm coming out and I'm going to start practicing, what do I need to know about the future of legal research? Steve Tover (43:35): Well, the first thing you should do is probably go learn how to be a plumber or an electrician. You'll make a lot more money, but you can cut that if you want, but the Jared Correia (43:43): We'll leave that in. Steve Tover (43:46): Okay. Because that's where the future's going to be. I think Jared Correia (43:50): Plumbing and electrical work. All right, everybody listen up. Steve Tover (43:56): But look, learning how to just use AI is fantastic, right? Whether it's legal research or anything, learning to be comfortable with it, like learning how to use a computer, if you recall when you got your first computer, and I remember there was a learning curve there. And I think it's the same thing with this because the tools are going to be only as good as the people at their comfort level at using them. And again, we're just at the beginning. There's going to be so much more functionality coming out and so much access that when we're looking at legal research specifically as we see it today, I don't think it's going to look the same in five, 10 years. I think it's going to be a lot more interactive and there'll be a lot more prompts to help people move along. So for a law student, I think just becoming familiar with these tools and not being afraid to use them, being the early adopters, being the one that brings in that technology into your law firm as a young lawyer, I think that's going to be important. Just like when people were bringing Westlaw and Lexus to their law firm, what they learned in law school, that's what they were using. That I think is going to be the saying thing here. So that familiarity, I think will help them become more effective than the ones that are not going to be using it. Jared Correia (45:15): Great. Steve, that was a really helpful discussion. We covered a lot of topics. You got a little bit more time to around for one more segment? Steve Tover (45:24): Sure. Jared Correia (45:25): All right. We'll be right back with Steve Tover from Any Law. (45:36): All right everybody, welcome back. Here we are again at the Counter program. It's a podcast within a podcast. This is a conversational space where we can address usually unrelated topics that I want to explore at a greater depth with my guests. Expect no rhyme and very little reason. Steve, I'd like to play a game with you today that I'm launching for the first time called Too Many Steves, which is loosely based on the Dr. Seuss story. Too many Daves in which Mrs. McCabe names all of her sons, Dave, which allows the good doctor to muse on alternative names for each boy. Have you ever read that one? It's one of my favorite Dr. Seuss books. Steve Tover (46:14): I don't remember that one. No, Jared Correia (46:16): Really all I'm going to have to send you a copy, but until such time, this is where you come in. I've noticed that there are a lot of Steves including you. So I have paired together different Steves and I've got six pairs of Steves from relatively similar Steves all the way up to very different Steves. So I'm going to read out two Steves to you, and you just have to tell me who's your favorite Steve in the group. Steve Tover (46:50): Sounds good. Jared Correia (46:51): Are you ready? Steve Tover (46:53): I am. Jared Correia (46:54): Alright, good. It's an all Steve party. Alright, let's stick with technology first. Here are two Steves that people know Steve Jobs and Steve Wosniak chat. GPT helpfully allowed me to come up with some quick summaries. I think most people know who Steve Jobs is, but he's the Apple Co-founder. Head of Pixar, revolutionized personal technology as chat. GPT tells it through visionary design, transformative products, and relentless innovation. And then we have Steve Wozniak, who's a different kind of dude. He also worked for Apple as his co-founder. He was an engineer. He is celebrated for his inventive hardware, brilliance, playful creativity, and philanthropic spirit. This is a tough choice between two Apple co-founders, but Steve Tover, do you have a favorite, Steve favorite Apple, Steve Jobs or Wozniak? Let the Ball Royal out begin. Steve Tover (47:52): No, they're about even for me. Jared Correia (47:54): Even. Steve Tover (47:54): They're about even for me. But I think I would take Wozniak just because he's the lesser known one. Jared Correia (48:00): Okay. Steve Tover (48:00): Deserves some love. Jared Correia (48:01): Alright, alright. I'm sure the Wozniak family appreciates that. Very good. Thank you for choosing. I thought you were sending this game off the rails already by not selecting. All right. Are you ready for the Steve Pear number two? Steve Tover (48:17): Go for Jared Correia (48:18): It. Somewhat related here as well. So we've got Steve Martin, who my daughter is currently obsessed with because she's been watching only murders in the building. He's a famous comedian, actor, writer, musician, known for his absurdist humor, and he's actually a great banjo player as well. And then we have Stevie Nicks Fleetwood Max singer-songwriter. She's into mysticism, she's got poetic lyrics and she's also had a solo career as well. So we've got Steve Martin and Stevie Nicks, which is your favorite, Steve? Steve Tover (48:55): I'm actually a big fan of both, and that's closer to my age group as well. But Steve Martin, hands down, I used to know all his routines by heart from the Let's Get Small album to Saturday Night Live appearances. It's not even close. Jared Correia (49:12): Yeah, it's tough to be Steve Martin for sure. All right, we are moving on to P three in the Steve Bowl. We've got Steve Irwin, the Crocodile Hunter, Australian Wildlife expert, conservationist. And then we have, now we're getting into some different Steves here, a very different Steve, Steve McQueen, actor cultural icon, star, bullet. I think Steve McQueen's been dead for a while and he's still very relevant from a cultural standpoint. So you got two really different Steves here. Do you prefer Erwin or McQueen? Steve Tover (49:54): So I'm actually a big Steve McQueen fan, and I think the thing that sold me on him was the Magnificent seven. I think he was the king of pool, as they say. And that movie just sold him. He was young, he was 30 years old when he did that movie. I think that Steve McQueen wins that one Jared Correia (50:13): Tough beat for Steve Irwin, but I also feel like it's really tough to be cooler than Steve McQueen. Well, I'm glad we're hitting all these cultural touchstones for you. I'm excited. All right. Now we've another very different pair of Steves the Steve Pair. Number four, we've got Steve Nash, Canadian basketball, legend two time N-B-A-M-V-P Elite Playmaker, MBA coach versus in the Bracket, Stevie Wonder, singer songwriter from Motown and beyond, multi-instrumentalist, groundbreaking albums, social activists. We got Steve Nash, Stevie Wonder, which is your preferred. Steve, if you had to choose Steve Tover (51:03): Both greats really greats? I would choose Steve Nash only because I really watched him play a lot when I was younger, when he was in the NBA and I really loved him, even though I was a Laker fan, we used to see him a lot. He was on The Suns. So great player, great player. Jared Correia (51:21): I didn't know you were a Laker fan. Had I known that I would've invited you on the show. I was literally a Celtic game last night. Steve Tover (51:29): I'm also a bird fan though, just so you know, Jared Correia (51:31): You've redeemed yourself. Steve Tover (51:34): Okay. Jared Correia (51:35): What's your favorite Steve Nash era? Like the sons era? Steve Nash. Is that one He was at best. Yeah, Steve Tover (51:41): Definitely the sons. I think so. I mean, watching him play, I mean there were a few guys like him if you really wanted go down that route, but there were some pure point guards. There was him, there was Stockton. You had a couple of guys on the east that were amazing. Of course, my favorite was Magic Johnson, but he wasn't a classic pure point guard. Jared Correia (52:04): But I think Steve Tover (52:05): Steve Nash really was, and he was amazing. Jared Correia (52:07): It's tougher to find pure point guard. A lot of the NBA point guards now are score first guards. Steve Tover (52:12): Yeah. Yeah, for sure. Jared Correia (52:13): Okay, we're getting a little obscure here. We got two more Steve's. Steve Tover (52:17): Okay, Jared Correia (52:18): Two more. Steve Pierce, Steve Pier, number five. Steve Carell comedian actor was in the office. Television show was in the movie, the 40-year-old version. And then we've also got Steve Aoki, who's a dj, producer and entrepreneur, famed for his high energy performances and blockbuster collaborations. Who we like here? Steve Carell. Steve Tover (52:46): I think you're going to go with Steve Carell. Jared Correia (52:47): Okay. Are you a fan of the Office? Steve Tover (52:51): Not really. I just like him as an actor. I think he's very funny and he makes me laugh. Jared Correia (52:57): Do you have a favorite Steve Carrell movie? Steve Tover (52:59): Maybe Dinner with Schmucks I think was very funny. Jared Correia (53:02): I've never seen that. Steve Tover (53:03): It's really funny. Jared Correia (53:05): Worth watching. Steve Tover (53:05): He actually isn't the main, he's in it, but Paul Rudd is amazing in it actually. But they're both good. Jared Correia (53:12): Oh yeah, Paul Rudd is good too. Alright, we're cruising along here. We've got one more Steve Pier, and as I said, we're going a little obscure here. Obscure Steve. Number one is Steve Reich. I dunno if you've heard of him, but he's a minimalist composer whose rhythmic innovation phasing techniques and influential works transformed contemporary classical music and inspired generations of modern composers like Radiohead. And then we have Steve Harvey, comedian, author, television host, celebrated for his quick wit, A host of Family Feud I think as well. Big mustache, loud clothing. So we have Steve Reich and Steve Harvey. Who's your best? Steve among this group Inc. Steve Tover (54:03): I'm going to go with Steve Harvey on that one. I like him Jared Correia (54:07): And Steve Tover (54:07): He's got a lot of charm and he's a funny guy. Jared Correia (54:10): Good stuff. Now the last thing I'll say is we know who the best Steve of all is, and that was Steve Tover from Any Law Steve Tover (54:19): And thank you. Jared Correia (54:21): Who was kind enough. I just teed this all up so I can say that who was kind enough to appear on the show today. Steve, I really appreciate you spending some time with us. Steve Tover (54:30): Well, thanks for having me, and I think this is a great format and I appreciate you allowing me to be here. Jared Correia (54:37): Oh, Steve Tover (54:37): Thank you. Really nice. Jared Correia (54:37): My pleasure. Thanks for our guest, Steve Tover of Lex Sphere and Any law. To learn more about Steve and Lex Sphere and any law, visit both lex sphere.com, L-E-X-S-P-H-E-R-E.com, and AnyLaw.com as any law.com. Now, because I'll always be a nineties kid who still enjoys a cargo pant, but whose true passion is burning CDs for anyone who would listen. I'm now just doing the modern version of that, which is creating Spotify playlist for every podcast episode that I record where the songs are tangentially related to an episode topic. This week's playlist won't leave you hungry. It's the collection of some of my favorite Harry Chapin songs and it's sponsored by Spam. No wait, it's not, but I wish it was. Fry Spam is a straight up culinary delight. Y'all join us next time when I slip into something a little more comfortable.