The Truth About AI Receptionists (A Law Firm Owner's Reality Check)

Welcome back to The Legal Intake Experts podcast! For more than a decade, Answering Legal has helped growing law firms ensure they never miss a chance to connect with new leads. Now, we’re pulling back the curtain to share our best strategies for strengthening your intake process and turning more callers into clients.

In Episode 13, Nick Werker and Tony Prieto tackle one of the biggest questions facing law firms today: should AI play a role in your intake process? They explore the rapid rise of AI tools, where they can be useful, and where they fall short when it comes to handling real client conversations. Along the way, they break down the differences between automation and human interaction, the risks of relying too heavily on AI, and why empathy and accountability still make all the difference on that first call.

Check out the episode below. You can also enjoy it on YouTube, Spotify and Apple Podcasts.

The Legal Intake Experts is part of the Answering Legal podcast network. Interested in learning more about Answering Legal? Click here to learn more about 400 minute free trial!

Check out the previous episode of The Legal Intake Experts here!

This podcast is produced and edited by Joe Galotti. You can reach Joe via email at [email protected].

Episode Transcript:

We are back. Once again on The Legal Intake Experts podcast. Since I'm not going to have a varied intro, I'm gonna vary my gesticulation around the intro.

Cadence.

Yeah. Yes. Uh presented by Answering Legal. I am Nick Werker joined as always by my co-host Tony Prieto. Tony, how are you today?

All good. It's uh summer already in March and uh I know we're going to be diving into AI today. Um, so you know, I figured I'd start with a very uncontroversial, uh, state saying that, uh, summer starts in March in Miami.

Uh, I am going to derail us as I often do, and I will tell everyone a funny story about myself. I have a lease that goes back in a month. And I also am a big planner. And they had 0.9% financing at Subaru. And so I bought a Subaru. Now, the lease that I have uh goes back in a month, and I wanted to return the lease early just to get it out of my driveway, but there is so much snow and there's still so much snow on the ground at this very moment in time.

And this is why I'm telling you this story is because Tony is lying to us all. There's still snow in March. It's not summer. Uh there's still so much snow that I when I was cleaning off this car lost the key in the snow somewhere. And I definitely either shoveled it into a giant four foot high pile or shot it out of the snowblower. Um either way, it is uh nowhere to be found and I won't return the car because they're going to charge me like $400 or $500 for the key that I lost. Um so I can't return the car and it's just sitting there. Um even though I make my last payment in a few days. So no, it's summer. it.

You know, the world is a is a place of of vast differences. Uh and it is 80° outside. Let me tell you, if it had snowed any time in Miami in the last 50 years, it would be melted and you would find your key or you wouldn't have lost it in the first place.

It would never have fallen in the snow fallen the sewer drain, I suppose. Uh I really hope that it doesn't melt and go flowing into the sewer drain on my on my block. Anyway, what were we talking about? You can you…

Yeah. Um, AI.

Okay.

Uh, specifically AI, of course, for running a law firm. Uh, and everybody has their opinion on it. Nick, how do you use AI in your day-to-day basis?

I don't know that I want to tell anybody that.

All right.

Here's what I here's what I will say about uh how I use AI. I have several custom projects inside of Claude that I use to keep track of things and where things are. And I don't really use them that much anymore because once I used those projects to get everything all set up, I went back to using my brain for things.

Uh, and I have to tell you that I really was so enamored with ChatGPT a year,maybe like a year and a half ago probably. And today, this is true. I used it today. This is the CrossFit Open. Dave Castro, director of the CrossFit Open. He wants to put out a hint on what the workout is going to be. So, he puts out this hint for the first workout last week. He says, "It's so obvious.” Nobody gets it. And then he reveals uh what the hint was. And he was like, "I was so nervous because if anybody screenshotted this thing and uploaded it to AI, it would show you that it was a statue of the person performing what was going to be a wall ball throw, a medicine ball throw to the wall." And nobody got it.

And so today he posted another hint for the workout that's going to happen in two days on Thursday. And I screenshotted it. Now, if anybody wants to help me with this, that would be great because I I would like to figure it out and I'm not going to be able to sleep for two days. It is um like seeds of something blowing in the wind over what I can only imagine as a California landscape because Dave lives on a ranch in uh Aromus, California.

And so I was like, "Oh, I'm gonna do what he said that nobody did last time and take a screenshot and upload it to ChatGPT." And ChatGPT gave me the most vapid, vacuous answer and explanation of what this scene was and how it would uh pertain to CrossFit that I was like, I cannot believe that I ever used this to like organize my thoughts or my brain.

And so that's my listen, you wanted a strong opinion, that's my strong opinion is ChatGPT is useless to me.

I um I already dated this recording by saying that it's summer in March. Uh but you asking for um someone to help you solve a problem that is due in two days despite the fact that this episode won't be coming out for several weeks is very funny.

Um I have a I have a similar story. So uh I got a call from my dad this weekend. He's trying to make like a parmesan and crusted mahi mahi and he needs .17 cups of breadcrumbs. Uh and

Why?

Because he… okay he so his actual question was what is .17 cups in ounces. And he went to holy smokes Google and he googled that and Google told him there's no there's no unique measurement but a dry cup of ingredients is usually 4.5 ounces. Complete nonsense because obviously cup of sand is going to weigh very differently from a cup of I don't know bacon fat.

Um and so uh he called me and said what's 17 of 4.5 and I said dad that is something that AI can help you with but it clearly cannot explain to you the difference between volume and weight.

That was a very convoluted way to describe your relationship with your dad for which I have much more…

And my relationship with AI. But so I want to ask should AI be part of the intake process at all?

Oh, I'm so excited for this. I didn't know. Okay. Okay. Okay. Okay. Okay. Wait. Time out. I'm going. This is I'm gone. This… you should not have lit this fuse. 

I'm the the guy shooting the gun at the NASCAR race.

This is crazy. So, were we talking to Gyi and Conrad about how things have come like all the way around? Um and then you and I were talking about this. Okay. Here's what I think. We were, that's why we were talking about this Don Draper Mad Men. We were like, oh, back in the day when phones first existed, uh, there was no voicemail. There was no automated anything. You would call and if the person missed the call, it would go to your answering service and like a receptionist would take a message and so on and so forth. Right.

Everyone had a personal answering service. I I brought this up because I was reading Rosemary's Baby.

That's right.

And everybody in that book has a personal answering service.

And so then comes the invention of automated answering, right? Like this voicemail, this uh what would you call it? Answering machine, right? With the little cassette tape. Holy cow.

Yeah.

I just picture uh who leaves the message and the Seinfeld episode. Is it George? And then they have to break into the girlfriend's apartment.

Yeah.

Yeah. Yeah. Yeah. That's George.

It's always George.

It’s always George. Um so that became a thing, right? And then that was like so acceptable culturally for so long. And then it turned out that you could really ignore that and people didn't want to interact with your voicemail anymore, especially in business, right? And getting a quicker response, especially in the age of information, became much more important. And so now what we're seeing uh at scale is the acceptability to use AI answering for your phone. And it's going full force.

People are so interested in this especially because the cost is I would say uh percentage wise lower but if you looked at it in impact terms it's really not that much lower but whatever I could I could go down the rabbit hole on that.

So that's like the new fad I just can see that like we went from okay we want person communication then we went automated then we want person now we want AI phone answering and the same thing is happening where it's like what's new is old again or what's old is new again in digital marketing.

And so I think this whole thing of I was talking to Jared Correia about this too on his podcast is he's like what do you really think about AI answering I was like AI answering is so hot right now that people are going to flock to it and they're going to use it and it's going to be so popular it's going to do I don't know wonders for your firm and then something is going to happen and they're going to be like oh but having a real person on the phone with empathy, like they're going to compare the the two things back and forth and they're going to say, you know, I'm going back to this and I'm going back to that and we don't like AI because like I said, it's it's vapid and vacuous. Um cuz it can get an answer technically correct, but is it helpful? I don't know.

Today, the Supreme Court declined to take up a uh a case uh or rather I suppose it uh in 2018 a man created artwork with a personally designed uh large language model um which was like right on the cusp of all this starting and um he tried to copyright it court said no you can't copyright AI stuff and then kept appealing, kept appealing and then the Supreme Court just denied they upheld the the ruling of the lower court which is that you cannot copyright AI generated art or text.

That is sort of the question there is that it's is that ahead of its time or is that behind the time. Right? Are we moving away like like you said Nick are we moving away from humans perhaps for now and toward AI or are we moving um or are we moving back toward it already? Right?

Things move so fast now where like uh you know there's this there's this idea that everything happens in 30 years cycles. This like pop culture idea if you listen to rock stars now their music sounds a lot like music from the '90s. Uh and I think the '90s had the best music ever. So I'm obviously in heaven now.

But um things move much faster in the technology space and um I know that for example like Sam Altman has said that you know they're looking for ways to monetize this technology they're putting ads they're try they're testing ads in chat GPT and like outside of enterprise solutions like phone answering, I am not a genius, but I can't see a way to monetize these technologies for the personal consumer you know and that makes their sort of prognosis on their longevity pretty difficult.

I also think there's an ethical argument to be made in this regard and and this is actually happening politically right now uh like the exodus of uh Anthropic or quit ChatGPT or whatever people are using model-wise. But is it ethical for us to continue to outsource all of what I will say like labor to AI, right? And is it wise in the long term?

And here's what I'll say about this as like an economics major in college, right? Is over time, what is this principle in what I believe is macroeconomic theory is that wages are sticky. So like inflation happen involuntarily. The government prints money. The costs go up because people want to make more money, but wages remain sticky because organizations uh want to increase their profit and one easy way for them to increase their profit is to slowly increase wages. Um and there's ways to get around that, right? You switch jobs and you do this and you do that.

But I think ethically this creates a huge problem because say you are what I will uh very arrogantly say is uneducated about marketing and you think you can replace all of your marketing with AI and that you're just going to go to ChatGPT and say give me a marketing plan and five blog posts and I'm going to post it and generate me a website uh using base 44 and and and all these tools right. And you can do that. And you can save, let's just say your marketing costs you $100,000 a year and you pocket that $100,000.

Ethically, was it cool to fire the human being who needed that paycheck to do it just so that you could advance the $100,000 and have uh again what I will call uh completely hollow empty marketing knowing that or maybe you don't know or having to reckon with the fact that you're going to cause yourself some damage both to the economy and and to your own business.

So, uh, I know that's a crazy philosophical question, but it's already becoming like companies are laying people off because of AI, and they're just using it as an excuse. And I I'm going to get political. We may have to cut this, right? But if you use a buzzword in like uh I don't know, like media, I'll use inflation again as an example. If you say that there's inflation, companies will just start to talk to to charge more for things that haven't been affected by inflation. It's just an excuse for them to charge more.

And so, are we allowing companies to cut labor and and and uh and do huge layoffs because they've been negligent in other areas and they get to blame it on like, oh, our employees should just be doing more and producing more because of AI. I very clearly have my own uh views on capitalism, but I just uh I feel it would be negligent for me to hold on to that information.

Well, without even getting into sort of Blade Runner are are, you know, what is a human questions? Um, we can get into the question of like, do people like talking to robots? Um, I might be an exception here, but I feel like I can identify something that's been generated by chat GPT just by reading it or looking at it. And you know that colors my perception of of the business because I know how inexpensive that was.

Uh and if I'm interacting with a business that doesn't even want to pay a call it freelancer to design their logo. Um like how much do you care about the business and therefore how much are you going to care about me? You know but even beyond that we can talk specifically about AI voice and phone solutions. I think right now people don't mind talking to AI because it's not that different from an IVR. It's not very different from a voicemail. Now, of course, we've banked the drum over and over again that IVR and voicemail don't work, but the AI receptionist might work better than that.

But the question is, how long are they going to be okay with that before they get annoyed with it or tired of it? How much better will those voice will those voice programs have to get in order to stay ahead of that curve?

Some of them are surprisingly really conversational. I think this is the problem that I have with people whose current um platform they use to further this um I don't know technology if if if it's what you want to call it cuz there's clearly people who are in bed with this technology or or getting like an affiliate fee for for referring it.

And I see these public facing personalities say some things like, "Hey lawyers, um, AI voice is here right now. It's so good and it's $20 a month and if you're not using this, you're going to get left behind and your competitors are using this and they're capturing more blah blah blah blah blah blah blah blah blah."

But I don't know how long that will remain acceptable for. I don't know how long it will remain legal. And I've interacted, obviously I'm in this sort of industry. I've interacted with enough of them to know that they're not responsible for anything.

And so the the upkeep of them, this is this is a really great example is uh Robert Williams shared something like this the other day where the person was like, I got rid of my CRM and I coded my own CRM and I made it and I got rid of my receptionist because I have an AI receptionist and I coded that receptionist and now that's $0.

But when it breaks, it costs me $12,000 a month in API calls to to use it as opposed to the I don't know, same salary that I was paying the other person. Hopefully, you're I don't know. That would be nice.

Um and and like the board is really upset with me, but like I told them this is all just futuristic and it's broken, but I'm the only person who can fix it. Blah blah blah blah blah blah blah.

How do you hold this technology accountable? How are you using it? How are you maintaining it? Whose job is it to maintain it? You know what I mean? I just don't understand the use of that at scale, especially with law firms because the competition is so high. The costs of case acquisitions are so high and just it's so it's so sexy right now, dude. Like, we'll sell you this AI receptionist and you pay us $300 for the year and we'll set it up and it'll be so great. But I just don't see it as this maneuverable thing that you could be scrappy with your business about. It's just it's a technology that you bought and it and it can serve one task. I just don't see it scaling.

Yeah. I mean, speaking of scale, I mean, I we work in marketing and and even though I understand it, I do get a little bit annoyed whenever a form submission comes in on a website and it's just a bunch of gobbledygook, meaning someone slam the keys on their computer in order to see the other side of the form.

And like, I get it. Absolutely. I feel bad doing that myself. So, I haven't done this, but I'm willing to bet money that if you call the number on your local Morgan and Morgan billboard, you're going to speak to a real person. You know what I mean?

That's a good point.

And obviously, they have vastly greater resources than you do. Um, and that allows them to not have to cut corners on quality. But sometimes if you want to, you know, be in the big leagues, you got to play like you're in the big leagues.

Yeah. It also comes down to I don't think, and here's here's really where I think it affects law firms that I'm speaking to particularly. The best example I have of this is that huge companies have their own customer service policies.

So, one thing that I've dealt with recently, I hate experiential explanations of things, but I'm going to do it anyway, is I use ButcherBox because my wife is pregnant and neither of us really want to go shopping for that much stuff. And ButcherBox has a really great selection of stuff and in my experience, it's decent quality when it shows up correctly.

And they're, I don't know, like consistency is so poor. So, I constantly find myself having to reach out to them and be like, "Hey, this thing came ripped open. This thing is missing." and they don't have a live chat. They don't have anybody that you can call, but they do have a process by which you can make a claim and get this thing um refunded or resent or or whatever. They have policies and I'm so cool with that.

Yeah.

They don't do live chat. They don't do phone calls. They do email. Right. If your firm is equipped to do things in a certain way and you want to do things a certain way, deploying something that's just so static as AI doesn't give you this ability to like maneuver and provide experiences.

So, I'll give you a good example is I don't think that people are looking to talk to a live chat agent on a website for your law firm as the um number one way to get in contact with you. In fact, there are a number of studies that say phone calls are still the number one way people want to get in contact with you, right? And if you spend time creating and curating this AI to talk to people on this chatbot or this phone call thing and it's wrong, there's like how do you retrain it?

I just don't understand how you continually maintain that thing and change it all the time to do that. Whereas if you have an intake person or a receptionist or an answering service and you just like talk to a human be, you know what I mean? Like how do you communicate with it and how does it know what to keep doing and improving? I don't think I don't think it it can just match up to what your process is going to be that matches to your strengths.

To be fair, those are things that I think are very easy to improve upon cuz those are an experience and use case issue, right? It's not an issue with the core technology of like let's say we have um an AI chat solution and we know how effective it is at getting leads to sign up.

And regardless of how you feel about the whether or not that effectiveness is going to improve over time which obviously a lot of people in the legal tech space are very uh bullish on AI. It's going to get better. This is the worst it's ever going to be, etc. Um even if you aren't if you're willing to accept you know that it's 60% effective um but it's you know 80% cheaper or whatever it is um there are places where the technology can get better without the technology improving right there's use there's a user input areas where like let's say your your tool gets a UX where you just get to do it yourself you get to to tell it what to do and and you can actually change it.

That's the space where I see the current tools as they exist being able to get a lot better. The problem is that those are the places that are being in a lot of cases ignored by uh people putting together these solutions.

But we've been like we've been very negative on AI in general and AI voice. Um however we here at Answer illegal do have an AI powered solution in our arsenal which is our um our AI powered intake chat. And one thing that I think is very cool about that is that we because we are a support organization um it’s very… the problem that Nick was just describing which is if this screws up how do you change it? Uh the answer is you call us and we change it for you. And that's the difference, right?

Uh imagine how much better your experience with butcher box for example would be is if you could just call a human being and say, "Hey, this is a problem. I need you to solve it." Uh that level of support is what we offer on our on our AI solution.

So it really depends on the quality of what you're offering. So I will go back to your example of this uh chatbot that we offer because I work very closely with the person who's responsible for the chatbot and all the time we deploy these chat bots and we're constantly figuring out ways to make the bot restricted in what it can answer because it's so dangerous if it could give uh like legal advice or anything like that. And also it has to have a primary objective because it just it needs guidelines and everything. It's the same way as you would like train an employee.

And I also know how many leads we deliver using this chatbot to law firms. They're it's totally anonymous to me, right? And we deliver a considerable amount of more leads that people are happy with using this live agent answering service. So I personally think and maybe here's where I'm going is you have to understand the audience that you are engaging with.

I think personally for me I deal with all law offers and if I were to write a ton of content about or using AI um or outsource all of my sales to AI like my intake, right? Um, I think even though lawyers say they're so excited about AI and they want to use it, it's so problematic that I know, I shouldn't say I know, I assume that the results would be poor even though pundits and talking heads and lawyers and influencers go out there and say how great AI really is.

I don't see it. I think it's a hot topic that you use to sell your product. I really do believe this. Anybody who's out there pushing AI in this way, saying that it's going to revolutionize is selling you something. Well, that's that's the thing. AI itself, right? The term AI is a marketing term made up by Open AI, by Microsoft. Uh the actual term for it, almost all of these things are large language models. You ask them a question, they go back to their massive database and try to construct an answer based on what they know to satisfy your query.

Um the the term AI is almost as much a marketing term as like replacing the word tissue with Kleenex, you know what I mean? Um, and I mean even in Spanish, uh, you know, here in Miami, at the very least, if someone asks you for a tissue cuz they need to blow their nose or whatever, they ask you for a Kleenex, not for a tissue, you know.

Um, and so, uh, AI is a lot like that in that people who are very much pro AI are are, yes, often trying to sell you something. They're also trying to sell you in a full body scan. Um, in medicine, there's a trend recently of people getting just full body scans. Um, and having a doctor look at these scans and tell them everything that could possibly be wrong with them.

And the reality is that AI is not equipped to do that. It's more like a scalpel. And you can build or buy a tool that serves a very specific purpose. Analytic purpose. An intake purpose with like a chatbot, something like that. Um, but anyone who's trying to tell you to uh anyone who's trying to tell you to turn your whole business, your everything over to an AI model or several of them is trying to sell you a full body scan.

And if you've ever seen the brilliant television show House, uh, Gregory House hates full body scans.

I've never had a full body scan. I haven't even gotten a scan of my shoulder. I just go to my physio and he tells me that it's fine.

I get so frustrated talking about this topic because I really at a at a core level I never and to to quote Dr. House people what does he always say are constantly lying.

Yeah. People always lie.

And I just I don't believe your AI solution is infallible. I don't believe that you have a thousand x’d somebody's uh revenue with uh AI. I don't believe that you used Facebook ads to make a hundred million dollars. I I just don't I don't care.

Um I know you said it. I can rent a yacht in Miami and make it look like I'm a freaking hundred millionaire. I could go right now and rent a Lamborghini and you and I could go I could take a picture and be like, I did this one hack and now I have a Lamborghini and here's you should pay me $1,000 a month and I could I just don't believe it. You know what I mean?

It's something that requires oversight. And so I think just deploying it to solve problems and and expecting it to just work in perpetuity and firing people to to accomplish that is I I I'm just I'm out on it. I'm out. And for those reasons, I'm out.

Yeah. I I I don't know. I personally don't use a lot of AI. Um because I like to think um and people who outsource their thinking to AI will find themselves just doing less thinking. And your brain is a muscle and you know, uh progressive overload is the only way to grow a muscle.

So, um, but I do think that there are very specific uses for this technology, and one of the reasons why nobody's really come up with any of them is that everybody's trying to get it to solve every problem because um, not to get too too controversial with it, Silicon Valley has not provided a meaningful change to people's lives since they invented this thing, you know, uh, and that was uh 21 years ago and they've just sort of been coasting off of that ever since.

So they're trying to find something that will be the next trillion dollar uh you know venture capital investment and it is right now it is AI but they one of the problems with like you know the way Uber built it its business model was disrupt you just operate at a loss using VC money to disrupt the local cab economy until you break them completely and you're the only option and then you start charging prices that will make you a profit.

That is the that is the MO and I the only way that works is if you're offering a product that people on the ground use not it doesn't work that way for enterprise solutions. And AI has many many enterprise applications um but they're trying to sell it as something that will revolutionize your life um and I just don't I don't see it.

But, you know if Matthew Mcconaughey is behind it, I don't know.

Another… I am so pale.

Salesforce AI.

I need to go outside. Uh I will always grin at private equity losing. Sorry.

This has been a real downer episode.

Yeah, I'm just frustrated. I I don't like it.

You know what it is. You woke up today upset. You're the in person who woke up today upset.

I did. I should have 100%. I spilled coffee all over my computer. Well, I spit coffee all over my computer screen. Um my dog and I are in a fight. We're fighting. Uh I don't just It's one of those days where like uh I don't know. Everything's uncomfortable.

You're in a fight with your dog. He's right there next to you whining for more attention.

You know what's funny is I don't know that he is. I don't think he wants to be here today. And I think that's why we're fighting. I think he's like, I wanted to be at home and you didn't let me stay home today. We should have both taken off today. That's what we should have done.

Yeah. Well, you know, if you wake up, you know, wrong side of the bed, sometimes that's what PTO is for.

Anyway, uh we talked a lot about a lot of things, but one of the things we did talk about today was Answering Legal. Uh, and I want to let you know that uh, you can sign up for our human answering service by going to answeringlegal.com and, uh, there's a link in the bio to get started with a 400 minute free trial.

Thank you for being a positive force on this episode, Tony. Um, and I smiled when you said human. I recognize my bias, but I also think it's really cool that I have principles enough to believe in what I do every day and uh and fight for the humans.

Humans versus robots. I don't like you robots.

You’re just a regular John Connor.

That's right. Anyway, be sure to check us out um on the next episode of Legal Intake Experts. All episodes of the show can be found on Apple Podcast, Spotify, and the Answering Legal YouTube channel.

We the two human beings who host this podcast you next time.

See you next time everyone.