sam-dixon-the-supercreativity-podcast-with-james-taylor_scp341-sam-dixon-full-video.txt
James Taylor (00:08)
Sam Dixon is the chief innovation officer of law firm Womble Bond Dickinson in the United Kingdom. He is also a practicing lawyer in the firm’s restructuring team. His innovation journey started in online retail in the early 2000s and has led him to law via a brief detour through the world of DJing. And if you don’t know Womble Bond Dickinson, I’m gonna put my hand up here. They are a client of mine. I’ve spoken to them for a number of times before. I think they have about a thousand.

lawyers in the US and the UK and they cover lots of different areas of business as well. And we’re going to be taking a deep dive into the work that Sam and his team do around innovation, specifically in the legal profession. So Sam, welcome to the SuperCreativity Podcast.

Sam Dixon (00:50)
Good afternoon, James. Thank you for having me.

James Taylor (00:52)
So share with us just now what’s going on in your world, what currently has your focus at the moment?

Sam Dixon (00:57)
Well, I think it would be very difficult to give an answer other than generative AI, to be honest with you. It’s stormed onto the scene. We’re working our way through the hype cycle, of course. And it’s keeping us all very, very busy, both in terms of what we can do right now with it, but also what the future trajectory is going to be and where we might end up.

James Taylor (01:18)
And take, how did you get into this role? I mentioned that you’re obviously a practicing partner, you’re a solicitor, you do DJing as well. Tell us how this journey into the current role happened.

Sam Dixon (01:29)
Well, I’ve always had a bit of an interest in doing things differently and innovation and in the use of tech. Going back to, you mentioned the online retail at the start of my journey. And essentially for me, that was working in an outdoor equipment retailer in a shop and then ended up taking over their mail order business, turning that into an online retail business in the relatively early days of online retail and doing things like using AdWords and a computer.

a different way to how anyone has used them before. So we were targeting essentially a product comparison approach that no one else was using at the time. And fast forward in, sorry, go on James.

James Taylor (02:07)
I think, yeah, I was gonna say the move into the law. So you didn’t come initially from the legal profession then. You kind of, you were starting in e -commerce and retail first, and then you kind of, how did you find your way into the law?

Sam Dixon (02:22)
Well, in some respects, in a relatively conventional route, in the bit of online retail, and then I was doing a law degree. Now, being honest with you, whilst I was doing a law degree, I was doing quite a lot of DJing and event promotion, and I didn’t apply for a training contract two years ahead of time in the way that people normally do, because frankly, I thought I was going to be traveling the world DJing. And as I went into my final year, I had a number of offers to work in various parts of the world as a DJ.

But in my final year, we did real legal work. And I ended up dealing with a multiple conspiracy to murder case, so some gangland stuff. And it was fascinating. And at the same time, the firm I’m currently with, which at the time was called Dickinson Dees, prior to a number of mergers, they approached the university I was at and said, do you have anyone who might be interested in starting a training contract in a few months’ time? So I went in and had a chat. And one of the icebreaker questions was, do you know?

tell us something interesting about your week last week. So I explained that I’d DJed for an artist called Chesney Hawkes. It turned out the head of graduate recruitment was a massive Chesney Hawkes fan, and the rest, as they say, is history.

James Taylor (03:37)
So you went into, obviously, you started initially in the kind of legal side around criminal law. But my understanding is that one Bond Dickinson is known really for commercial law, and it’s that kind of world. And then you specifically, you focused on the restructuring side of things. So how did you get into that particular part of commercial law?

Sam Dixon (03:58)
Absolutely, well as you say it’s a commercial law firm so the criminal side of things was just part of my university experience. As soon as I joined what’s now Umubon Dickinson it was commercial law from the start and I started off as a trainee like all lawyers do and I rotated around a few different seats and being honest I didn’t know an awful lot about restructuring. I didn’t know what to expect but as I was going into my final seat it was in the height of the recession.

And there weren’t necessarily the opportunities in other areas that I would have perhaps liked to have done. And so someone gave me the chance to go and work in the restructuring team. And it turns out it was a really good fit for me. So I qualified into restructuring and insolvency and spent a number of years learning the ropes and really enjoying the restructuring sector, helping to try and save businesses and that kind of thing. But over time, I saw there were a number of things that we were doing in restructuring, which

I think it’s probably fair to describe as a bit boring at times. James form -filling, doing the same precedence again and again that he didn’t necessarily feel was the best way of doing it. There must have been a better way. And that got me to start looking at document automation. Originally using things which with hindsight look like quite archaic technology now and borderline kind of coding. Whereas,

things have evolved and it’s got easier and easier to do. But that first time that I saw a suite of documents that are automated, where in the past it would have taken hours to produce it, to then be able to just go out to answer a few questions and it’s going to tell me what documents I need. And it’s going to produce them all for me with all the associated paperwork. This is just amazing. And it was that that really was the trigger point for me gradually over time becoming more and more involved in our innovation.

efforts across the firm.

James Taylor (05:53)
I know with a lot of law firms I’ve worked with in the past, they initially got very excited about RPA, robotic process automation, taking some of those agreements, automating things to a certain extent. We’ve now moved from just doing that into obviously artificial intelligence and machine learning as well. When you’ve seen these technologies start to be applied in the work of law firms, where do you typically see it? Is it in document?

Is it being document management side? Is it looking at risk in relation to contracts? Is it helping lawyers draft agreements or something else?

Sam Dixon (06:31)
So I think it depends on the particular technology. But if you’re talking about that document automation side of things, that’s the production of documents very clearly. When you look at machine learning though, that’s usually around something like a document review. So that might be a due diligence exercise. It might be someone wanting to understand their range of contracts, whether it complies with their policies in a particular area. They might be trying to understand the lease portfolio and wanting to extract various bits of information across that portfolio because,

they’ve inherited it from someone else, it perhaps doesn’t have the level of detail and structure that they need. And that original transition machine learning tools worked to a degree to deliver that task. And there has been some success on that front. I think where it gets really interesting is that generative AI is able to do a lot of that work without the need for lots and lots of examples.

and the proactive work to train the particular machine learning tool to look for particular points of interest.

James Taylor (07:41)
Now your work obviously around innovation, a lot of law firms struggle around innovation because it’s traditionally the legal industry is one that’s focused on billable hours, your six minute increments or however long it is. And so everyone obviously partners associates very focused on those billable hours and to be able to take a step back and think actually, is there a better way of doing this or a smarter way, a more productive way of doing it? It’s often very difficult when people just focus on those billable hours. How did you…

to change the mindset within a firm, especially with those colleagues of yours who maybe said, we’ve been doing this for years, why do we have to change it?

Sam Dixon (08:20)
I think you chose a great word at the start of that, James, which is traditional. And law is a very traditional sector. People have a very traditional mindset. What we’ve tried to do to get people to think a bit more differently and to try out some of this new stuff is, first of all, just to get it on their radar. Because if they don’t know about it, then they haven’t even got the option of trying to do things differently.

So we do things like our International Innovation Week, where we work with our colleagues in the US and where you’ve spoken before, to just tell people stories about innovation, give them that inspiration. It’s about making them want to get involved. And are you going to get absolutely everyone to take a more innovative approach? No, you’re not. With the best will in the world, that’s never going to happen. But you don’t know exactly who is receptive to it and who isn’t.

So the key for me is creating that culture of innovation where it is part of the conversation and where people are enabled to suggest ideas and implement ideas. So, sorry James.

James Taylor (09:30)
No, I was just going to say, so take us through like an example of that. Let’s say you’re a partner in the restructuring part of the business and you’ve been doing something for a while and it works, it’s fine. But you’re thinking, listen, maybe I want to try using some of these AI tools and thinking about things in a more innovative way. What does that process look like? Is it lead often led by the partner? Is it led from the innovation lab within it? Is it led by an associate? Is it project managers?

How does it start to come together?

Sam Dixon (10:03)
So we take a two tier approach to innovation, James. And what I’m talking about there is our starting point and the first tier of that two tier approach is what we call our self -service toolkit. So we’ve been focusing on trying to sort of leverage that experience across the whole business. So rather than having a team of people centrally who are the ones who innovate, the others.

We try to stay away from that and empower everyone to do differently what they know best, what they’re dealing with day in, day out. So the self -service toolkit is a series of pieces of software with training materials made available in a way which minimizes the need for central approval, be that cost approval or whatever else so that people can pick things up.

take ideas forward with the appropriate approvals from within their own teams so that ideas aren’t being duplicated and so that time isn’t being wasted when it should be applied on something else. But really empowering everyone to be able to take that forward. We do recognize though that some of what people want to do might be a bit more ambitious. And it might require a little bit of coding skill. It might require some admin.

access rights from an IT perspective, which with the best rule in the world, we’re not going to expect our lawyers to have. In my view, I don’t think lawyers need to be coders. I mean, query in the modern world of low and no code, how much coders need to be coders? But lawyers for me certainly don’t need to be. And therefore, we have the second tier, which is the complex automations. And that’s where you’re looking at potentially an end -to -end process.

with a lot of heavy automation in there and probably integrations between multiple different systems.

James Taylor (12:01)
So let’s say on that then, let’s say you’re having all folks from different parts of the business say, we’ve got this problem, we think we could use, we would like to use AI to help us on this problem. Someone else sees an opportunity. How do you then kind of sort and sift all that, decide actually this is where we’re going to focus on? Is there strategic kind of pillars that you’re focused around? Will you only do projects where you think it could benefit across the whole firm rather than a particular department in the firm? How do you decide?

Sam Dixon (12:30)
Well, I think it depends on what support is needed, James. I mean, the idea of the self -service toolkit is we don’t need to decide. The individual teams, for example, restructuring can say, right, well, we’ve got a sale agreement here, which we think will benefit from being automated. There’s a self -service tool available to let us do that. And we, as a restructuring team, can assess whether or not we want to use our own team resource in order to progress that.

So that’s one side of it, obviously on the complex automation side. There is a need for prioritization. And that’s done on a business case basis, as you probably expect. And so we listen to what teams want. And sometimes teams don’t necessarily know what they want. This is as much about a conversation as it is about anything else. And I’m currently touring all of our UK offices just talking to people.

floorwalking, having conversations around, right, well, what are your pain points? What are you trying to achieve? What is the most boring thing you’re doing, the most frustrating thing you’re doing? And where have you got to work around? Because that’s always a great example of an innovation opportunity is where someone’s currently trying to work around existing systems and processes. So we have those conversations. We then review the business cases and we work out where the best ROI is, frankly. And sometimes,

That’s where there is really well structured workflows that are high volume and it makes a lot of sense to put a lot of structure around that. But other times there might be things which could benefit a very broad part of the firm. And actually someone in one particular team has come up with an idea which is of much, much broader application. And then it’s not around that really well structured work necessarily. It might be.

around a particular aspect of a task or just something that everyone’s doing. Like for example, reviewing agreements in Word. There are aspects of that. One of the tools that we’ve put in the self -service toolkit recently, which has been really, really popular, is, it sounds like the simplest thing in the world, but essentially it’s a sidebar in Word that scans your Word document and it allows you, rather than…

jumping around the document back to definition tables and clause 12 .5 that’s referenced down here, it allows you to just double click on the work, the defined term, or double click on the relevant clause, and it pops up alongside you. So you don’t have that constant context switching and loss of focus. And it turns out that is something that is applicable to a really large part of the work that we do. And it’s saved an awful lot of time and to be frank, has probably improved consistency as well because…

even the best lawyers in the world get tired and sometimes miss things.

James Taylor (15:27)
So how do you decide, and this is the thing I’m starting to see with some firms where they’re saying they’ve been using a lot of off the shelf products from different providers. Some that focus very much on the legal industry, others that are just more broad like the chat, GPTs of this world, for example. And then there’s those firms that say, actually, we are going to employ our own data scientists. We’re going to build our own models. We’re going to do that work as well, which is obviously more expensive and sometimes is not in the…

the usual field that a law firms would do. Have you made that, have you started coming up against that? Have you started thinking about that decision about, okay, when are we going to have to maybe think about starting to build our own things that will have basically be WombleBone Dickinson creative products?

Sam Dixon (16:12)
The buy v build argument is something that’s been going on in the legal sector for a long, long time. And I’m largely still of the mindset that firms that specialize in developing software, be that for legal tech or otherwise, are probably going to win and probably going to produce something better than we can. It’s the old adage of keeping the main thing the main thing, isn’t it? What we do is provide legal services. We don’t develop software.

Where we really have the advantage is in using that software, in integrating different pieces of software together using APIs where that might not have been done previously. And that’s where the edge is. However, I do think there’s a bit of a blurring of the line now between buy and build, because how you’re using some of these tools increasingly starts to look like building rather than buying. So.

I mean, you reference people building their own models. Personally, I think the idea of someone building their own large language model as a law firm is bonkers. The scale that is required and the broad capability that existing models have, it just doesn’t make sense to me for someone to build their own model. But you can go and use Microsoft Azure Studio, OpenAI Studio, to develop your own.

GenAI powered chatbots. And that’s something that we have done. Now is that buy, is that build? We’ve created the system messages, we’ve played around with the p -values, the k -values, the temperature, we’ve put the data set into it. I don’t know whether it’s buy or build in that situation where we bought in a tool and then we’ve used it to build a GenAI chatbot. And frankly, I guess it doesn’t really matter which side of the line that falls in.

And we’ve started relatively simple and there’s an ambition to build out from there. So what we’ve built so far, we’re calling iWomble. And iWomble currently looks over our 70 something policies and procedures and answers plain English questions about them. In line with our AI policy, people still need to verify the answers. They can’t completely rely on what’s provided, but we’ve found, passed over a relatively small data set, what we’ve

built, bought, is really quite accurate. And people are finding it far, far easier to get to the right answers, especially when some questions are answered across multiple policies. And it brings them all together, gives them an answer like if I asked you the question and you knew all of our policies, the kind of answer that you would give together with citations so that people can go and check them. And then that saved a lot of time. And due course, we’ll build that out on a modular basis to look at all sorts of other areas of the business as well.

James Taylor (19:12)
If I was a client, one question is going to be coming up in my head just now. Okay, when I get my bill and I have partner X, 20 hours, a junior associate X amount of hours, paralegal X amount of hours, where does the AI, where does AI Womble AI, AI Womble live in that? Are you going to bill that as a separate kind of product service within it, or are you just seeing this as something that just helps augment the…

the folks that actually already work in your firm.

Sam Dixon (19:45)
I think it’s still quite fluid at the moment, to be honest with you, James. I think the whole sector is trying to work out how generative AI is going to impact pricing. For me, currently, it’s a question of looking at things on a case -by -case basis. We are still, I believe, at a relatively early stage in gen AI being applied to law. So there are a number of major vendors who are building gen AI solutions onto existing products. And the roadmap for those is sort of

during the course of this year. So a lot of the legal specific benefit is still to be realized. When you look at something like iWamble, for example, those questions around policies and procedures, clients are benefiting there because some of those policy questions will be around how a particular process works. And we’re making sure that people get to the right answer quickly. Now, some of that probably wasn’t time that was.

was ever charged to clients in the first place because it’s our own internal policies. Some of it, depending on the nature, might well be. So there’s some savings and efficiencies there which would just be passed on. The more general potential for this around document review and things like that will be discussed on a case -by -case basis. So it might be that clients are given the option, for example. It might be, look, we’ve got this new tool.

We can do it this way, and we can agree a fixed price of x. Or if you prefer, we can do it a different way. And it’ll be charged based on time, but those rates might reflect, for example, the overhead cost of the AI. So there’s different ways of looking at it. Whether we’ll get to a place where there’s a technology charge.

that sits on files alongside time, for example. That might be one route that the industry goes. But I know that when other industries have tried that, there’s been some pushback there. So I think that could be challenging. One thing that could well happen, James, and as I say, this is all really fluid at the moment, it might well be that as with other overheads, essentially, chargeout rates are adjusted.

to reflect the fact that work is being done more efficiently, but there’s a big investment cost in these tools in the first place. So it could be that it’s offset that way, but time will tell.

James Taylor (22:08)
Yeah, I was I was thinking we had one of our other guests we’ve had on the the season is Sir David, Professor Sir David Allman, who was the formerly head of GCHQ. And so what with all the intelligence services, one stat he shared was of the public of publicly indexed information is out there, I when you’re on Google or Yahoo, wherever, that is only 0 .3 % of information exists, because most information exists on intranets or

dark web or places that the general public do not have access to as well. And one thing I was wondering there was, were some of these larger clients that you might have where you say, listen, we can maybe do a, almost a quick pro -crop because AI, we need data to train it on. And if there’s some way that we can obviously, it needs to data security and all the things and anonymize and all the things you would normally have, but there’s almost a bit of a competitive advantage there because some of your clients who have large data sets,

large amounts of information going back many, many years, that’s amazing information to train an AI on.

Sam Dixon (23:14)
It is, and I’m sure over time we will see more and more licensing style deals to get access to content that isn’t indexed, like the Financial Times announced last week or the week before. I think for a law firm, it’s really, really challenging because you’ve got that duty of confidentiality, you’ve got all the information security requirements, you’ve got all the data protection requirements, and…

Again, I guess it goes back to keeping the main thing, the main thing. To what extent do we want to be distracting from what we do best by having conversations with clients around potentially acquiring their data for a different purpose to how we normally use it, which is essentially what we will be talking about. And I don’t know.

Call me traditional James, call me someone who’s been trained as a lawyer, but it just makes me inherently nervous and I wonder if actually the way that that will progress is for the really big players to approach some of the big multinationals and do licensing deals with them in the same way as they have done with the FT rather than trying to partner with law firms who leverage it.

James Taylor (24:28)
Yeah, it’s almost a little bit like the within the UK with the NHS data that’s been anonymized. So and just different companies that have access to that so that they can run models and do, you know, use AI assay models and things to kind of test but individuals information is not necessarily known by that AI. I’m interested, you know, with your your other side of view, which is the DJ, which we’ve come to just now. I mean, the world of music is being

changed so fast just now because of AI, especially generative AI. I was playing with something for an event I’m doing in London later this week, a tool called Suno AI. I don’t know whether you’ve been playing with it at all where you can just write in some ideas for some lyrics, give it the kind of theme that you, the kind of style of music you want to do, and it will do you a track. And it sends this real sounding audio, real sounding vocals, real sounding everything. And it’s great. It’s good for getting a first idea. And some of them are quite funny as well.

Are there any things that you’ve seen within the music space and your background as a DJ where you’re kind of thinking, actually, I’m starting to see this now within the legal profession, perhaps in terms of how the job is changing or how we’re going to have to take a certain approach or a certain position in terms of how we work with AI?

Sam Dixon (25:47)
Well, I think there are going to be parallels across lots and lots of different sectors. And I think if you look at how DJing has evolved over time, I started in the CDJ era. So just after vinyl, the early days of CDs being the medium of choice, simply because you could carry more of them. But you were still actually mixing. You were still exercising a degree of skill over time.

And I suspect there’s lots of modern DJs who would be screaming the, their radios right now, but over time, technology and earlier forms of AI have increasingly been able to say, right, well, this is this many beats per minute. That’s that many beats per minute. and what we can do is marry that up and we can line up. So when you, when you drop the one track over the other, actually you’re slightly off in your timing and therefore I’m going to adjust it for you and make it play. So what that’s done over time is level the playing field a bit.

and allow people to, more people to access that particular profession and to perform at a standard that they weren’t previously able to perform. And I suspect that generative AI might have a similar levelling approach for certain bits of legal work going forwards. And the key will then be, well, what skills are the ones that really add the value?

when you get to that place. And for me, when you look at what’s going to be left as AI advances, and who knows when we’ll get to the end game on that. But it’s going to be things like the social intelligence and that complex problem solving in a complicated set of circumstances. And that’s where I think lawyers need to be really focused on developing those skills and making sure they remain relevant. And the other thing is you’ve got to accept

that the world will change and you can’t fight against that. It’s better to be part of that change and to embrace it and to accept that the skills that you needed before are different to the skills that are needed now. I mean, handwriting is a lot less relevant in the role of the modern lawyer than it used to be, James. I think many lawyers would argue with that. So it’s just another step in the evolution.

James Taylor (28:13)
guess also, you got a front row seat of this in terms of all those young trainees are coming in young and older trainees, you know, people are going to university later in life. But some of the younger trainees who have come up being that first generation who have been using chat GPT for in their exams, for example, in different ways, they’re now coming into your firm saying, Hey, I was able to use this when I was at university studying for my law degree, why can I not have access to these same tools in a firm? So you’re kind of you’re coming.

You’re going to be coming up against that as well. Quickfire questions as we start to finish up here as well. Is there a quote or a line or statement that you kind of live by that you often kind of is kind of your, is your kind of compass in life in some way?

Sam Dixon (29:06)
I’m not sure that there’s any one statement to be honest with you James.

I mean, I was once mocked roundly and I suspect I’m about to be again, but I will do it because it’s the answer to the question, I guess. I was once stopped and interviewed by someone from Radio Lancashire and asked exactly the same question, essentially. And as a 16 -year -old, I said, keep it smooth. That was my response. And I guess as cringe -worthy as that was and as much abuse as I received from…

James Taylor (29:34)
Keep it smooth. Keep it smooth.

Sam Dixon (29:43)
friends at the time for saying that. I think there is a certain truth in just staying calm and objectively assessing problems in order to properly work out what an innovative solution to that problem is, which is my attempt at a neat segue into your next question.

James Taylor (30:04)
And then, so on that, we’ve been talking about just having a sense of perspective on things and assessing in that way as well. Are there any tools that you use or apps that you use you find are very useful in the kind of work that you do? We’ve mentioned obviously the Chat GPT and different Microsoft tools and things. Are there any that you particularly think, I don’t think I can live without this now?

Sam Dixon (30:31)
Well, I think it’s probably more use case for one of the tools we’ve touched on rather than anything else. And that is generative AI for me has become a really useful sounding board. It’s essentially a coach for me. So I might draft a new idea, a new element of our strategy, and then say to it, right, act as our managing partner and ask.

what questions would you ask about this documentation? Critique this for me. And I use that on an iterative basis to try and anticipate some of the things that people might ask and some of the challenge that might be put back against a particular idea.

James Taylor (31:12)
I use a very similar thing I call virtual masterminds where I’ll put in an idea for a project or a business plan that we have, and then I’ll give it six people that I respect who have very different perspectives on business and things. And I’ll say, okay, find all the flaws on this. And it’s fascinating because you get to see from different perspectives and then we kind of do the human bit, which we kind of triangulate all of that and say, okay, that’s fine. Now that I know how Elon would approach it or how Warren Buffett would approach it.

now I need to think like, okay, how would Sam Dixon approach it? What would we do? What about books? Are there any books that you can recommend to our listeners, maybe around AI or the future in some way?

Sam Dixon (31:56)
Well, Superintelligence by Nick Bostrom, I’m sure you’re familiar with it, James, is a fascinating read about AI and about, well, I would guess the potential doomsday scenarios that we might all face. But in approaching those doomsday scenarios, you do get a reasonably good feel for the progression of AI over time and a good grounding in what AI is.

James Taylor (32:24)
And if people want to connect with you to learn more about Womble Bond Dickinson, more about the firm, what’s the best place for them to go and do that?

Sam Dixon (32:31)
Well, you can find me on LinkedIn or on X, I am at Innovation in Law.

James Taylor (32:38)
Sam Dixon, Chief Innovation Officer of Law firm 1 with Womble Bond Dickinson Thank you so much for being a guest on the SuperCreativity Podcast.