Back to Basics: Understanding AI Tools from the Ground Up

Episode Notes

Transcript

Announcer:

Web 2.0 innovation collaboration software, metadata got the world turning as fast as it can hear how technology can help legally speaking with two of the top legal technology experts, authors and lawyers, Dennis Kennedy and Tom Mighell. Welcome to the Kennedy Mighell report here on the Legal Talk Network

Dennis Kennedy:

And welcome to episode 392 at the Kennedy Mighell Report. I’m Dennis Kennedy in n Arbor.

Announcer:

And I’m Tom Mighell in Dallas.

Dennis Kennedy:

In our last episode, Nikki Shaver, co-founder of the Legal Tech Hub, joined us as part of our fresh Voices on legal Tech series. Great insights from Nikki in the episode, be sure to give it a listen. In this episode, we have another very special guest in our Fresh Voices series, in Fresh Voices. We want to showcase different and compelling perspectives on legal tech and much more. Tom, what’s all on our agenda for this episode?

Tom Mighell:

Well, Dennis, in this edition of the Kennedy Mighell report, we are thrilled to continue our fresh voices on Legal Tech interview series with Uwais Iqbal, founder of some Plco, a London-based AI consultancy. He’s a knowledgeable and insightful contributor in the legal AI space, and we are excited to talk to him today. We want our Fresh Voices series to not only introduce you to terrific leaders in the legal tech space, but also provide you with their perspective on the things you need to be paying attention to right now. And as usual, we’ll finish up with our parting shots, that one tip website or observation that you can use the second that this podcast is over. But first up, we are so pleased to welcome Uwe Iqbal to our Fresh Voices series. Uwais, welcome to the Kennedy Mall Report.

Uwais Iqbal:

Thanks. Thanks for having me, Tom. Dennis, it’s a real pleasure to be here.

Tom Mighell:

Before we get started, can you tell our audience a little bit about yourself, just a little background, what’s your role, what our audience should know? We’re going to get a lot more into your background, but what are the basics that the audience needs to know?

Uwais Iqbal:

I think the first important thing to know is I’m not a lawyer, so I’ve never been to law school. I never practiced at a lawyer. I’ve never actually worked in a legal capacity. My background’s as an AI engineer. So I’ve spent close to the past decade building and delivering AI systems in the legal sector. So I worked at a number of legal tech startup startups, if anyone’s been around for a while. You might remember Eigen Technologies in the 20 17 20 18 days of the extraction hype cycle. I also spent time at the innovation lab in Thomson Reuters building Cool POC and RD projects using legal content. And then I also spent time at Thought River who were doing contract review and contract negotiation using ai. And about two and a half years ago, I set up my own company Simplex, thinking about how to do AI consulting in the legal sector. So my background is as a AI practitioner who’s interested in helping the legal industry actually deliver value with AI and step past the hype,

Dennis Kennedy:

First of all, it’s awesome for us to have you as a guest on a, I usually say it’s not always easy to talk with lawyers about tech. And sometimes I get frustrated how difficult it still is to explain technology old and new. And I think it’s great that you reminded people that AI in various forms has been around for a while. So about the technology and its benefits to those in the legal profession, especially as you’ve probably found if you’re not a lawyer, but I think you do a great job at it, and I love seeing your posts on LinkedIn and other things that you do. Would you talk about your approach to communicating with lawyers and others in the legal profession about technology, especially AI and what you found works well for you?

Uwais Iqbal:

Yeah, so it is a very difficult challenge taking something as complex as AI and trying to explain it and add to that the complexity of trying to explain things to lawyers who aren’t always the most open about admitting they don’t know things. So there is that challenge. I think what works really well is simplifying things down as much as possible. I think one of the reasons I got into building plco and what we’re doing around AI education was it emerged organically. We saw that there was a lot of hype in the market and a lot of people having inflated expectations about ai. So typically the majority of people have only been exposed to AI through either headline, sensationalist media articles or the marketing material from vendors. And both of those are skewed to deliver a rosy picture of ai, which isn’t necessarily the truth. And I think that’s not a problem where it’s not a fault of anyone’s own responsibility.

I think it’s just natural that way, the way that things happened, the way things have happened. So I think what’s interesting is that if we go back to fundamentals and go back to basics where we think about how do we define our terms, what do we mean when we speak about ai, what do we mean when we speak about gen ai? What do we mean when we speak about a large language model? And being very specific in terms of creating a shared vocabulary and understanding of that language. I think I’ve been doing AI education now for almost two and a half, three years with the work we’ve been doing at symplex. And we’ve done some amazing things in terms of educating lawyers. We’ve worked with some of the largest law firms, we’ve done things across the globe and educated, I think to date around 3000 lawyers.

So I’ve got some muscle around what works and what doesn’t work. What typically doesn’t work is getting right into the technicalities of complex jargon and specifics about the technology and sometimes people posture and they think it’s cool and impressive to speak about the number of parameters in a language model or how complex or amazing rag based architecture is, or why fine tuning or reinforcement based learning is the future of ai. And all that’s well and good if you are technical and you want to be a bit geeky, but for the large majority of lawyers in the industry, they’re not interested really in that stuff. They’re primarily interested in understanding the technology, getting their head around a lot of the jargon, which exists a lot of the buzzwords, a lot of the technical terms, and also thinking more practically around how can they make best use of the technology for their day-to-day work.

And what we’ve tried to do in our educational philosophy in the educational work we do is ground that in being very simple. So it’s in the name iCal, how do we go from things which are very complex to things which are very simple, but do it in a way where it’s educational and it’s fun and it’s entertaining so that people don’t feel it’s something dry or it’s something which is a drag. And I think the big challenge, even now 24 months after chat, GBT and everyone in the legal industry has lost their minds to ai, I think there’s still a big gap in terms of how do we get everybody in the entire legal industry to a baseline understanding of AI where everyone is at least using the terminology in the right way and knows what they’re speaking about with the buzzwords. The biggest challenge we face is that we’ve spoken with chief innovation officers of some of the largest law firms in the world, and they’ve confessed to us that they don’t really know the difference between ai, machine learning and natural language processing, and they’re responsible for making decisions about purchasing software and that can run into the millions of dollars.

So it is very worrying. So for us, it’s thinking about how do we slow down and how do we think about AI in legal as a long-term journey? It’s going to be 10, 20, 30, 40, 50 years before we actually get to a space where people are doing this right? And it’s widespread and we’ve got that level of adoption and we’ve cracked it to think about we’re very early in that cycle of technological evolution in the industry. How do we get the basics and make sure everyone in the industry has that shared foundation so that when it comes to actually adopting the tech, we’re doing it in a much more effective and productive way. So that’s our bias to it. And as a practitioner, that’s my biggest bias. It’s how do I get everyone educated as quickly as possible so that we can move past a lot of this crosstalk with buzzwords and actually get to doing and building core solutions which would move the needle with the legal industry.

Tom Mighell:

So this is usually the time where I ask a question about technology competence and the lack thereof that we see that lawyers have in the area. But I’m going to switch the question up just a little bit with the general notion that legal has a reputation for slow tech uptake that they just have. And so I’m wondering in your experience, what you’re seeing with the law firms and the lawyers that you’re working with, what’s the barrier to that? Is it primarily the education or is it something like culture or data quality in their firm or ROI or is it something that I’m not even thinking about?

Uwais Iqbal:

You mean what’s the biggest challenge to adoption exactly,

Tom Mighell:

To the uptake, to the adoption of ai?

Uwais Iqbal:

So one of the biggest challenges is education. What has happened is that there’s been a frenzy of law firms buying shiny AI tool because it’s the new trendy thing to do, and we’re seeing the same behavioral patterns we saw with the extraction hype cycle. Everyone’s going out buying a shiny gen AI tool like they did five, six years ago. Everyone went out to buy a shiny extraction tool. But I think the lessons still haven’t been learned around how difficult technology adoption is within legal. And that add to that, the complexity of ai add to that the complexity of an entirely new technology that people haven’t really been exposed to and don’t really understand and don’t really know how to make use of and get the best out and then queue the prompt engineering stuff that follows from that. So I think that one of the biggest challenges is adoption, but one of the biggest challenges around adoption is education.

And one way to think about this is that lawyers you can think of as Michelin star chefs, right? They’ve gone through a rigorous training maybe 10, 20, 30 years of their life. They’ve gone through law school, they’ve gone through culinary school, they’ve earned their stripes, they’ve done their training years, they’ve done their qualifications, they’ve moved up the ladder to get to their positions. So they’ve got maybe 10, 15, 20 years of invested training and education of how to do things in a certain way. And then they’re Michelin star chefs, right? They’re trained to try to get the best quality and the highest level of precision in the work they do. That’s where they have that reputation of being the best at what they do. Now, if you have a Michelin star chef and he’s been cooking on a fire based would kind of think his whole life, and then you introduce a microwave to him, it’s not going to be an easy transition for him to go from cooking the way he’s used to cooking for 10, 15, 20 years of his life on an open fire to now starting to use a microwave and try and shift his entire workflow, his entire experience, his entire way of professional kind of work product into this new era of working in microwaves.

So I think that’s the frame, at least in our minds of how we think about educating lawyers around AI and legal. And the biggest challenge is that a lot of the work lawyers are doing is high quality, high fidelity, high precision work. So in the UK we have things like jacket potatoes. If you’re making a jacket, potato can get away with making that in a microwave. That’s something you can pop in a potato into a microwave, run it for maybe a half an hour, 20 minutes and you’ll get a baked potato out, that’s fine. But if you’re making a souffle and you’re a Michelin star chef and you have at your disposal what is a microwave, and you try to use a microwave to make a souffle, you’re going to end up with a wonky souffle. So I think part of the challenge is that people are being pressured to adopt AI and use ai, but we don’t really have a good way of conceptualizing what types of legal work is useful for which types of toolings, right?

So we like to paint a spectrum. You have microwaves, you have pizza ovens, and you have personal chefs. And this is what we’re doing at Plica where a lot of the tools which have been brought into firm sit in that microwave category and everyone’s trying to cook everything in the microwave, even if it’s a souffle, and they’re ending up with one key results and they’re questioning things and they’re questioning the entire enterprise of AI as a result. And then you have personal chefs, you have pizza ovens, which are more direct and specific. So if you want to make a pizza, you go and make a pizza and the pizza oven and it does something very precise. So there are those types of tools on the market. And then we’re seeing an emergence of use cases or opportunities of applying AI in legal where it’s far too complex for off the shelf tools where you need some level of kind of custom bespoke tailored solution.

That’s where you have a personal shift, that’s where you have AI practitioners, that’s where we work as simplex. We work on the side of helping firms build those customs types of solutions. So I think part of the challenge with adoption of AI in legal is where we’re too eager and we’re too impatient to try and see the results with ai. We haven’t really understood in my mind, the technology enough to actually say particularly how to be very focused and structured in our strategy from a pragmatic standpoint, but also we haven’t got the right conceptualization of the technology that it’s something which it has levels of fidelity and the work lawyers are doing has levels of fidelity. And in order to make the most of the technology, there has to be a match each level of fidelity for those outcomes to be delivered. And I think part of the challenge is that we leaders and decision makers who are making these purchasing decisions don’t have that level of insight into ai.

And what we’re seeing now is the first symptoms of people having buyer remorse or people having this friction with tools being brought into firms, very, very expensive tools being brought into firms, but the uptake’s not there, and they’re starting to question why their uptake isn’t there. So we’re starting to see the first emergence of the problems and the symptoms, but the root cause comes back to a lack of education. It comes back to a lack of understanding. It comes back to a lack of actually understanding the tech and how to deliver it into an industry like legal, which there isn’t really a playbook. It’s something everybody’s figuring out as we go through it, but I think there’s a lot more we can do to get educated about AI to make sure we can do things and avoid mistakes.

Dennis Kennedy:

Yeah, I love that analogy because I’ll be talking with lawyers and they’ll tell me what they’re thinking about using AI for, and I just scratch my head. It’s like, why would you even be thinking about doing that? It’s like the wrong tool. It’s like, are you using this pizza oven for something that has nothing to do with pizza? Right? And so I think there is something, I like what you said and the notion of fidelity to the work that you actually do. And I sometimes like to think of it as jobs to be done is one way that I think about it. But I think that analogy is really great. Another question I had like to ask is that I felt in let’s say about 10 years ago that the large London firms, and maybe that means the UK firms were way ahead of the rest of the world on ai, and I’ve sort of assumed that’s been the case, but most people will tell me I’m wrong on that. Now that the US firms have jumped ahead, I’m not sure I buy that. But I, I’m curious if you, not to say who’s ahead, but do you see differences in how firms outside the US approach AI adoption as compared to what you might know about US firms?

Uwais Iqbal:

To my knowledge, I haven’t seen a significant difference in firms inside the us, outside the us, inside the uk, outside the uk. I think what I’ve seen is caricatures of different types of profiles or different types of strategies firms are taking. There’s been the profile of firms who’ve gone out making lots of noise and posturing that they’re doing something very fancy with ai. But then behind the scenes, the traction isn’t really there when you speak to people on the ground. There’s another caricature of firms who are sitting in the wings waiting for other people to do things, and then just following the playbooks. Other firms have, I’ve seen caricatures of other firms who’ve taken a completely different approach where they’ve gone completely build and they’re hiring and they’re building internal muscle to do this, but then they’ve got a much longer time horizon. So I think there are maybe four or five prototypical kind of profiles of what firms are doing from a strategy standpoint.

But I think what’s interesting is what you mentioned or our jobs to be done. So we have a framework at simplexa, what we call the legal AI actions. And what this is, is to try and take the conversation away from applications. So the way we like to think about the world of AI is that we have the first layer, which is the technology layer. That’s what you have with gen ai, large language models, and now agents in that technology layer. Then the second layer is the application layer. Those are applications which make use of the fundamental technology. So you have church bt, you have copilot, you have insert any AI tool into that list. And then you have use. So what’s your intended use of the application or what’s your job to be done? So do you want to summarize a document? Do you want to create talking points for a webinar or a podcast about AI and legal?

And at that use layer, what we’ve done is we’ve really done a lot of work to get close to lawyers and what they do to try and articulate what are the types of activities lawyers are already doing on a day-to-day basis where you can now explore AI being involved or AI being employed in some way, shape or form. And then we’ve listed these out in terms of actions of what lawyers actually do, rather than speaking about the technology. And we’ve ended up with things like extracting labeling, finding information, summarizing information, organizing information, drafting information, interrogating for information, translating, transcribing. So at the moment, there’s eight or nine of these actions and the conversation becomes so much more productive because if you’re able to speak at that level of use or at the level of jobs to be done, then you’re much more precise about the actual problem you’re trying to solve.

And then you can move down to the application layer, figure out if you need to build or buy and then leverage the underlying technology. And I think firms are starting to think about that. And I think that the larger firms, what’s happened in the industry is that the larger firms have got excited and because they have the budget and the capacity, they’ve gone out to buy expensive tools and they’re able to make that mistake. So they’re able to throw spaghetti at the wool and see what sticks. So they’ve gone this technology up approach. So they’ve gone brought in the technology, they’ve distributed it across the firm in this decentralized manner, and they want to see what comes up from that in terms of use cases and use. But they’re struggling to get the adoption because it’s something so general, it doesn’t really fit into any one’s specific workflow or any one specific practice.

There’s the mismatch, but then what we’re seeing is that mid-market firms, because they have restricted budgets, they have to be much more strategic about where they spend their money. They don’t have the ability and they can’t afford to make those mistakes. So what we’re seeing is that they’re taking a much more targeted approach with very, very specific use cases. So let’s start, we have maybe four or five practices within our firm. Let’s focus on the real estate practice. Let’s find one or two key high value use cases where AI can be applied in the practice of real estate. And let’s see those use cases to fruition specifically, rather than trying to bring in a tool which will try and solve everything for everyone, for every use case within the entire firm. So what we’re seeing are these two opposing approaches firms are taking. One is a technology up approach.

Let’s throw spaghetti at the wall, see what’s six. It’s very expensive, but it’s gone as attention because firms can posture, they can put out press release, they can do the whole innovation theater around it. And then mid-market firms, firms are starting to be more strategic where they’re taking this more tailored approach. And what we were doing at plco, and this is something, it’s a combination of maybe two and a half, three years worth of work in the industry around AI and legal. What we’re doing is we’re creating a methodology around that. How do we create a methodology which is focused on specific targeted use cases and structuring a step-by-step engagement in a way where we can take people from the education phase into exploring targeted use cases into building tailored solutions so that we can get people to adoption with ai. So I published a blog post, a newsletter article about this today, is that if we actually want to optimize for adoption of all of the other variables we can optimize for around AI or around AI and legal, if we really think about that, then what we really need to do is get really specific about solving for specific use cases because it’s much more targeted, it’s much more niche, and we know what success looks like because we know what we’re solving for.

And I think that tension is going to exist within the legal industry, that there’s going to be people who are trying to boil the ocean with ai. And then there’s going to be people who are going to be very, very specific and very, very measured about how they approach it. And in my view, time will tell, but in my view, there’s very, very structured and very, very pragmatic approach. We will outlast the more larger technology driven approach because because it’s closer to home and it’s closer to the actual problem. And we’ll see what happens over the next few years and how that plays out.

Tom Mighell:

Alright, we’ve got a lot more to talk about, but we need to take a break for a quick word from our sponsors and then we will be back with Al at Simco.

Dennis Kennedy:

And we are back with Al at Simco. We find in our Fresh Voices series that we love to hear about our guest career paths and our audience does as well. Would you talk about your own career path and what kinds of things you’ve done to get into your current role and focus?

Uwais Iqbal:

It’s probably very different to a traditional career path in legal. So I never went to law school at university. I studied theoretical physics, so I’m very technical, I’m very abstract, I enjoy that kind of thing. It gets me out of bed in the morning. And then after university, the first job I landed was at a legal tech company, Eigen Technologies. And I joined the company when they were seven or eight people. So it was very early before they had got funding and before they were working with some of the larger banks. But that was an exciting endeavor into the startup world and working with the team. Once they got funding, it was scaling very rapidly, it was quite fun and exciting, and we were at the cutting edge of thinking about how AI could be used to extract information from contracts. And then I moved over to the innovation lab at Thomson Reuters, which was a different type of beast.

So it was a larger corporate, it was much slower. And in the innovation lab we had the opportunity to really experiment and explore AI with legal content. So back in 2018, we were doing head note summarization for case law, pre transformer models, and then the innovation lab was very exciting, but Thomson Reuters is a large corporate, so things struggled to get out of the innovation lab. And I wanted to do something which is a bit closer to and closer to users and closer to actually driving value. So I came back to the startup board and I came to Thought River, and they were looking at how AI can be applied to if you have, for example, a playbook standard and a drafted clause, can you detect deviations and can you surface risk and do cool things around that for contract review negotiation? And something very interesting happened across these three different companies.

As an AI engineer, as an AI practitioner, what I found was that I was coding literally the same algorithms across each of these different companies. And each of these companies had a massive stack of marketing on top saying, look at us. We’re doing something very unique and special with ai. But as a practitioner underneath, I was writing the same code. And then this just struck me as something very fascinating and something very odd that why is it the case that from a technology standpoint under the hood, it’s the same algorithms, it’s the same code, albeit with different data sets focusing on different challenges within the legal industry. But then there’s this masquerading phase we have on top that vendors play all about the

Tom Mighell:

Marketing.

Uwais Iqbal:

It’s all about the marketing. There’s this masquerading phase. Vendors play on top to try and position themselves as doing something unique. Of course, I mean every company has to demonstrate why they’re unique within the market so that they can have some traction and they can survive. But I think that the challenge here is that it comes back to this education piece, which is why I press on about it a lot is that specifically with ai, there’s an information asymmetry which is being exploited. So in the technology world, the technologists and the vendors that they understand ai and then on the client side or on the law firm side, they don’t have the information about ai. They don’t have an understanding about the technology. I mean that’s changing, but it’s slowly, there’s this massive information asymmetry between the awareness of the buyers and what the vendors are putting out.

And now when decision makers are making purchasing decisions, they’re making purchasing decisions, not aware that the information asymmetry exists. And the challenge becomes then how do you, what’s going to happen is if this information asymmetry exists, people are going to buy the technology, they’re going to be disappointed and they’re going to end up dismissing AI as an entire enterprise because it didn’t fulfill on the promises and so on and so forth. So we’re starting to see that. The interesting thing for me is what happens if we exist in a market where that information asymmetry no longer exists? What happens in a market where everyone has access to the same level of information and education about the technology and there isn’t this masquerading or there isn’t this kind of false flag marketing happening or this whitewashing of AI so that people can actually see where the value is and we can get to the promised land of the productivity plateau much, much faster than where we actually are.

So my vested interest in all of this from an education standpoint is how do we educate people as quickly as we can so we can skip through the phases of the hype cycle and get to doing more productive things faster? And my career journey has been interesting because I’ve been on the side of actually building these solutions on the vendor side. And now with the work I’ve been doing with Symplex, I’ve been working closely and seeing the other side of that story in terms of the challenges people are facing inside law firms, the pressures they’re under and the types of decisions they have to make and the types of things they have to deliver internally within their organizations. And I think that that perspective is being born into a lot of the work we’re doing at Simco to try and think about is there a way to create a bridge between those two worlds? Is there a way to put aside all of the hype, bring the technology in a way where it’s clear and it’s transparent, but also educate law firms, educate leaders, educate decision makers, so they’re part of that process and that journey and we can get to those productive solutions much, much faster.

Tom Mighell:

So let’s talk about the notion of buy versus build. I would wager, and kind of what you’re saying so far is that it sounds to me from what you’re saying that most firms, certainly the mid-market, but also I’m assuming large firms too are preferring to buy rather than build. What are you seeing as I guess common mistakes that law firms are making when they’re thinking about this issue? Are there any mistakes in their thinking that they could be doing better at to come to that conclusion in a better way?

Uwais Iqbal:

I should call this out. So I’m biased because I run a consultancy, so I sit on the build side. A lot of our business is on the both side. So that’s a bias I have in the nature of the business we do. But the mistakes people are making on the buy side is that they’re not thinking about where the technology’s going to be used before purchasing the technology. So almost use cases is a post-purchase decision, which is the wrong way around. You need to start. So this is the pyramid we have. So the base is the technology. Then the second layer is the application, and the third layer is the use. If you start with the technology and bring in an application and you haven’t got a desired use for it or you haven’t got your jobs to be done defined or you haven’t got your workflow processes mapped or your data organized in terms of how you want to use it for a particular practice, then what will happen is that the tool will come into the organization and the lawyers themselves, the fee earners, they’re so busy doing the fee earning, they have no capacity in Headspace to actually figure out how this tool can be used in their work.

So someone has to come in and do that work. And typically firms who have the innovation capacity and who have legal technologists, they sit in that space and they try and bridge that gap. So you have practice innovation attorneys, they exist in that space, but for the majority of firms, they don’t have people who can be dedicated to actually thinking about joining the two world. So the challenges on the buy side is that people are buying technology without really thinking about how it’s going to be used or what the right use cases are. The build question becomes very interesting because what’s happened now is that as CHATT has emerged and gen AI and large language models have taken off, the cloud providers have made available the same large language models through cloud enterprise services that are powering some of the largest and greatest AI applications like GPT and copilot.

So if someone has that technical capability or they have that technical expertise, for example, within their own cloud environment, they can deploy the same large language models, which is powering some of the shiny legal tech tools that are on the market. You can get access to the core technology without having to go through the process of buying an application. And then what that is doing is it is shifting the dynamic that where previously when it came to purchasing software, people almost always, they primarily chose to buy, but now firms have that ability that the technology has become somewhere spread and it’s become so accessible, firms can start building before they buy. And we’ve seen some firms do this. So some firms are building in order to address the information asymmetry so they can understand their capabilities of what they can get out of the box of the cloud providers, what the cloud providers are giving them before going out to market to buy, because then they have a better picture of why should I pay so much money for this vendor when I can stand something up in my own cloud and it’ll cost me a fraction of the price.

So that’s part of the thinking around building. So another aspect to the build versus buy equation is that when it comes to adoption, a lot of the legal tech tools which are on the market because they are VC backed, that creates a dynamic where the vendors themselves want to try and chase exponential gains in revenue so that they can feed back to investors and investors can get the return that they need. And what that does is it creates an uplift where the vendor products, they are so generalist and they’re so feature laid in where they try and appeal to so much of a market, so much of the market, so much of the total addressable market where they become general tools that are effectively microwaves and they don’t solve anyone’s specific challenge because they want to try and address the market and they want to sell as many seats as possible.

So that tension exists within the funding realm of legal tech. And what happens then is that these tools become microwave when legal practitioners are wanting to cook so flas and they’re wanting to create high fidelity outcomes. So that’s where you have the microwaves, you have the pizza ovens, and you have the personal chefs. And now whether you build or buy has to come back to your use case. And this is why use cases are really important. If your use case is that you want to make a Chay potato, then you should go and buy a microwave. So you can cook a Chay potato if you want to make a pizza or you want to, if you’re in a patent space and if you need a patent drafting help with patent drafting or you’re in real estate and you need a specific tool to help you generate real estate reports, then you need a pizza oven.

You need a specific AI application for your practice, which does that specific work product that you’re interested in. If you are, for example, in a banking practice and you’ve got this specific checklist that you have to do as part of your fam process to deliver outcomes to your customers, then you’re not going to be able to do that in a microwave because it’s very specific to the practice you have. There isn’t a pizza oven on the market because nobody’s got to that space of building very specific tooling for a banking practice. But what you’re going to have to do is think about how do you build in that space because nothing on the market can actually solve your problem. So you can’t use off the shelf tools to solve that problem. And then that’s what we’re seeing is that firms have bought in tools for very generalist tools, these microwaves, there are some pizza ovens which exist.

But then at the other end of the spectrum is that when you get these really practice specific use cases, you need someone who has the capability to come in and to actually build AI to that spectrum of what the lawyers actually need within that practice. And then this is part of the work we’re doing at Simco, we’re thinking about how can we help firms for those very niche and specific use cases where they can’t buy things off the market or things which exist on the market aren’t enough and they need something to go further, how do we help firms build in that space and build those bespoke solutions?

Tom Mighell:

Quick follow up to this. If law firms are building, are you finding that, or maybe do you have a recommendation of do we work with a foundation large language model? Do we work with the legal AI tool? Does it depend if we’re doing a real estate versus an IP versus something specific? What’s your recommendation if we’re deciding to build?

Uwais Iqbal:

It depends on the use case. So in my view, I want to solve one use case at a time. And I think the challenge is that because we’re so eager to solve everyone’s problems, we don’t actually solve anyone’s problems in the process. So if you go use case by use case, typically what we’ve seen in the bespoke builds we’ve done for customers is that the foundational models from the cloud service providers are typically enough to support what lawyers want to do. The biggest challenge becomes how do you create the right workflow, the right user interface or the right type of UX interaction with the lawyers so that they can come in, that they can use the tool, they don’t need a lot of technical headspace, they don’t need a lot of training, they can simply come in, the workflow is laid out, they make a couple of clicks and they can get what they need and they can be on their way. So a lot of the challenge and the bottleneck is on the UX side, how do you build those workflows and that UI design so that lawyers trust the tool? And you’re not asking lawyers to go through endless hours of training on prompt engineering to actually use something specific, but they have a UI which is designed specifically for that use case for the outcomes, and then they can use it and they can be on their way.

Tom Mighell:

Alright, we still got a few more questions for Al at plco, but we need to take another quick break for a short word from our sponsors.

Dennis Kennedy:

And now let’s get back to the Kennedy Mighell report. I’m Dennis Kennedy.

Tom Mighell:

And I’m Tom Mighell, and we are joined by our special guest ESAL at Plco. We’ve got time for just a few more questions. Dennis, those questions are yours.

Dennis Kennedy:

So two questions really. So first, I mean I teach a lot of students about AI and I’m really impressed with what they’re doing and their perspective and their insights into what’s coming, but they’re also looking for different types of career paths than just traditional practice. So how would you encourage today’s law students and new lawyers and I guess others in legal tech to find career paths in legal, tech, legal operations, and non-traditional careers in law? So that’s the first question. The second one is we always like to have our guests do a little bit of the research work for us and say, who are the fresh voices in legal tech you would like to signal out and maybe see as part of our series?

Uwais Iqbal:

To the first question in terms of what young lawyers should do in the industry in terms of thinking about career paths is I think they should stay as close to the technology as possible in whatever they’re doing. If they can find a way, if they go into practicing law, if they work for a legal tech company, if they work for an in-house team in an operations capacity, I think wherever they see themselves, what they should do is think about how they can become the translator between the existing traditional world of lawyers and practice and what lawyers have been doing for decades and years on end. And the new technology which is emerging. And I think that the trouble is that we don’t have enough of those translators who can sit between the two worlds and who can connect the two worlds. And if someone is really interested in how they can position their career to make the most out of what’s coming with AI and legal, we are just at the start of this, this is just the beginning, is to be in that space, become someone, become a lawyer who can speak to lawyers and understand the world of legal, but also become someone who can speak to AI engineers can speak to data scientists, can speak to, speak to software developers, can speak to product managers and actually translate across those two world.

Can you speak to the technical world of law? And then can you also translate that into the technical world of AI or engineering or product? So I think if people are able to think about how they do that for themselves wherever they are, whether they’re within a law firm, whether they’re in a legal tech company, whether they’re in-house or whether they they’re somewhere else, I think that can be a very exciting space to be in because those people will be very, very valuable in terms of enabling the future of technology within the legal industry.

Dennis Kennedy:

And then who might you suggest as potential guests for us? Who are the people whose opinions and perspectives most interest to you these days?

Uwais Iqbal:

Sure, that’s a good question. The trouble you see is that I’m a skeptic, but I would recommend a handful of people. I think Catherine Bamford is very interesting. She’s doing a lot in the document automation space to actually help people move away from ai, but actually think about it from a data standpoint, from document automation. Also, I would say thinking about speaking into more practitioners in this space, so more people who have profiles like myself that people who have profiles as AI engineers, data scientists, product developers who are in this space of thinking about how do we build this? I can give some recommendations. So I have a friend, Harry Green, he’s involved in building, he’s a lawyer who’s transitioned into a role as an AI engineer and he’s helping build a legal tech product. And yeah, I think those two people will be good recommendations.

Dennis Kennedy:

Yeah, we had Catherine on before, who’d love to have her back again. I think she actually mentioned you was the first. I think several people have mentioned you, but I think she was one of them as I recall.

Tom Mighell:

Well, this has been great, but it’s time to go. We want to thank Uwe Sigal at Simco for being our guest on the podcast. Ues, can you let our audience know where they can learn more about you or get in touch with you?

Uwais Iqbal:

Sure. So I’m very, very active on LinkedIn. I post about AI in legal and I’m quite contrarian in terms of how I think through things from a practitioner standpoint. So feel free to drop me a like or feel free to follow me on LinkedIn and feel free to drop me a message and then happy to speak and always talk shop about AI and legal.

Dennis Kennedy:

So thank you so much, Uwe, you are fantastic guests, great information and advice. I think you have some really helpful analogies and your perspectives are great. And as usual, I find so many topics to discuss and so little time. So we’ll have to get back to you at some point I think. But now it’s time for our parting shots, that one tip website or observation you can use the second this podcast ends waste, take it away.

Uwais Iqbal:

Great. So this is a bit of a shameless plug, but we’re speaking a lot about AI education. So how do you get educated as someone who wants to learn about AI in the legal industry? We’ve done a lot of educating lawyers. So to date, we’ve trained over 3000 lawyers across the globe and we’ve done things for firms like Link later and Bird and Bird. And what we’ve done is we’ve tried to condense all of that knowledge about how to train lawyers about AI into a free course that you can get via email. So it’s a seven day email course, you sign up on the simplexa website. So if you go to the simplexa website, there’s a resources tab, and under that resources tab, there’s a link to an email course called the Legal AI Basics and the Legal AI basics. It is a seven day short email course where every day you’ll get an email to your inbox teaching you about ai. And across the seven days, we cover what we think are the basics that everyone in the legal industry needs to know about ai. So feel free to sign up, it’s free sign up, go through the course, let me know if you have any feedback, share it with your colleagues, share it with your teammates, and share it with anyone else who you might think be interested in learning about ai.

Tom Mighell:

Awesome. My parting shot is fairly simple. Based from the last podcast or maybe a couple podcasts ago, I mentioned how I had gotten some rather jarring notifications about people trying to change my passwords. So I found this article, it’s really helpful when you get an email that says that a password request has been made. There’s a little bit of a panic moment there, and there are some things you should be doing after you receive that password reset email that you didn’t request. So I put a link in the show notes, there’s some things that you should at least pay attention to, make sure that it’s not going to be successful,

Dennis Kennedy:

Dennis. And as I do AI experiments, I am really interested in the idea of reusable prompts. And so there’s a number of ways to do that, but I really like using the Text Expander app for that. So what I’m able to do is to take a prompt, use it as a template, have a little quick keyboard shortcut for that, that I type in, and then the prompt is there and whatever AI tool I happen to be using it at the moment, I can use it in any of them, is a great way to do two things. One is to reuse the prompts, but then also to get me to think about prompts in the sense of reusability.

Tom Mighell:

Alright, so that wraps it up for this edition of the Kennedy Mighell report. Thanks for joining us on the podcast. You can find show notes for this episode on the Legal Taught Networks page for our show. You can find all of our previous podcasts along with transcripts on the Legal Taught Network website. If you’d like to get in touch with us or suggest a topic for an upcoming episode, reach out to us on LinkedIn or remember, we always love your questions, so you can leave us a voicemail at 7 2 0 4 4 1 6 8 2 0. So until the next podcast, I’m Tom Mighell.

Dennis Kennedy:

And I’m Dennis Kennedy and you’ve been listening to the Kennedy Mighell report, a podcast on legal technology with an internet focus. We wanted to remind you to share the podcast with a friend or two that helps us out. And as always, a big thank you to the Legal Talk Network team for producing and distributing its podcast. We’ll see you next time for another episode of the Kennedy Mighell Report on the Legal Talk Network.

Announcer:

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

Continue Reading