The Business-First IP Playbook: David Hyams on Mapping Value and the Limits of Generic AI

In this episode of IP Innovators, host Steve Brachmann sits down with David Hyams, Co-Founder and Chief Business Development Officer of Longship Legal, to explore what happens when a patent attorney stops asking, "Is this patentable?" and starts asking "What is this business actually worth?" Drawing on a career that spans big law in Boston, in-house roles at Bose and AOL, and a CleanTech start-up, David brings an unusually holistic lens to IP: one that treats patents, trade secrets, contracts, and copyright not as separate disciplines, but as a matrix of legal tools in service of a company's real intangible assets.

David walks through the founding philosophy behind Longship Legal: that too many companies are spending serious money getting patents granted without ever connecting those documents to actual business value. He explains why the first conversation with any client has to start well before IP, with questions about how the business makes money, who its customers are, and what truly differentiates it in the market. 

On AI, he's clear-eyed: it's a powerful accelerator of business processes, but it cannot replace the strategic human judgment an IP department needs. He also draws a sharp distinction between general LLM tools and proprietary legal AI built on restricted data sets, and why that difference matters enormously in practice.

👨‍💼 About the Guest: David Hyams is Co-Founder and Chief Business Development Officer of Longship Legal, a digitally native, AI-forward IP firm he co-founded in 2023. With a career spanning big law, in-house roles at Bose Corporation and AOL, and IP boutiques, David brings a uniquely holistic perspective to intellectual property, one built on understanding the business before the patent. 

🔧 Sponsored by DeepIP — your patent intelligence platform, from idea to enforcement.

🛎️ Subscribe to hear more real stories, evolving practice, and the future of IP law.

Transcript

Welcome And Deep IP Sponsor

Steve Brachmann 0:00

Hello, I'm Steve Brachmann, and welcome to IP Innovators, where we profile the careers of leading patent professionals and explore how AI and patent technology reshape intellectual property practice. This podcast is brought to you by Deep IP, the go-to AI patent platform for demanding law firms and in-house IP teams. Fully integrated into your workflow, Deep IP is setting the standard for how patents are drafted, reviewed, and managed from idea to enforcement.

Steve Brachmann 0:32

Today on IP Innovators, I am joined by David Hyams, co-founder of Longship Legal. David, thank you for taking the time to speak with us today.

David Hyams 0:41

Hey Steve, thanks for having me on.

Steve Brachmann 0:43

Absolutely.

David’s Path Into IP Law

Steve Brachmann 0:44

So we always start with a little look back at a retrospective. How did you get into patent law?

David Hyams 0:51

Yes, so let's just jump right into it. Um, I, you know, I've always been interested in technology, even when I was a kid. Uh, I started out in undergrad on a pre-med track. So I always knew I was heading into the sciences in some way. Um I got into law practice, though, after realizing as an undergraduate that I had a lot of different interests. Um, law is something that uh came later for me when I finally did go to law school some years after undergraduate. Um, I uh found myself uh interested in and attracted to uh areas like intellectual property. There was always sort of a technology theme underlying even my interest in law school. And uh it colored the way I uh pursued law practice afterwards. So um I started out uh right out of law school at a big white shoe law firm in Boston. Um and uh even there, I couldn't quite decide what I wanted to do. I actually ended up splitting my time between the corporate M&A department and uh intellectual property. And uh I spent maybe three or four years at big law, uh moved on to an in-house legal position at Bose Corporation just outside of Boston, um, a very well-known technology company in the you know acoustics and audio space. Um, when we relocated to Washington, D.C. in 2005, I took a job in-house at AOL, which was all technology all the time, and spent a great deal of time learning about how the internet works. Um, I left that practice after, say, two or three years and joined a cleantech startup in the DC area. Again, all technology all the time. After the cleantech company, I joined um IP boutiques and uh focused intentionally at that point on uh patent preparation prosecution, which is sort of the one skill set that I hadn't really focused on up until that point. So I'd done a tremendous amount of transactions work, I'd done a tremendous amount of um intellectual property transactions, served as a subject matter expert on IP, helped write policies at AOL around open source software use, did a ton of licensing agreements. Um, and I used that opportunity to focus then on uh patent prep and prosecution. So uh sort of rounding out the holistic skill set. Um technology was always a theme, and that's kind of how I found myself practicing IP law.

Steve Brachmann 3:29

So early on in your practice, did you find that the technologies you were prosecuting kind of matched what your background was in engineering? Or uh did you find yourself spreading out into many different directions?

David Hyams 3:44

Right. That's a fair question, almost a trick question, actually, because uh formally, you know, I'm not an engineer by education, right? I have a broad general science background. I started out pre-med. Um and in fact, my earliest interests were uh, you know, biotech. I did uh a couple of summer research fellowships working for the American Cancer Society, doing pharmacology bench research. Um, and it was really when I joined um the large law firm and then especially at Bose, uh, where I was getting a lot of sort of on-the-job training on you know the technologies that we were using there. The obviously the mindset is very much the same. And um the same sort of scientific approach was the same. And it helped that I've always been very, very interested in this stuff. So, you know, learning about acoustics, learning about electronics at Bose, um, learning about um how the internet works at AOL, learning at subsequent jobs about um computer and telecom networks and devices that connect to them and ways that information is transported across those networks, um, the way those networks uh connect to endpoint devices and so on and so forth. Um, it's so I guess you could say there was a lot of sort of on-the-job training, a lot of on-the-job exposure, um, which I suspect a lot of attorneys in this area can relate to, right? Especially if we're doing cutting-edge work or working with cutting edge technologies. By definition, it's something that you didn't learn about in school, right? So you have to bring your mindset, your framework, your ability to learn to these uh new technologies, which is you know one of the reasons that it remains fresh and fun.

Steve Brachmann 5:27

Right. It's a constant science education. And talking to attorneys, it really seems like people kind of grow along with the areas that they're prosecuting in. And you you have this really nice diverse background in different technologies. Uh, how did that background uh get you into the place where you were looking to start Longship Legal? And what's the work that you do there with that firm?

Founding Longship Legal For Value

David Hyams 5:52

Right. So I'm fond of saying that no experience goes unpunished. Um the evolution to Longship Legal has been very interesting and is by far the most fun and the most fulfilling thing I've done in my legal career so far. Um, Longship was born from uh a series of conversations with my now business partner, Tim Johnson, where we started asking, we were at the same law firm uh doing patent prep and prosecution, and uh we started asking ourselves questions like so, what is the value of what we're providing to our clients, really? We were getting patents for clients. They were spending a tremendous amount of money and getting uh patents granted from the patent office, US and foreign patent offices. But we couldn't really uh connect the dots between the amount of effort that they were putting in, that we were putting in, uh, the documents that were getting generated and granted, uh, and the actual bottom line value to the companies that we were serving. So, in part, Longship uh was an attempt to answer that question. How can we understand what the real value to businesses of intellectual property are? And how can we start providing that real value to our clientele?

Steve Brachmann 7:16

So, on a day-to-day, what does that look like when you're uh working with a company to try to extract that value?

Turning Business Goals Into IP Strategy

David Hyams 7:24

So one of the things we had to do was take a step back and look at um, well, first we started looking at intellectual property holistically, but then we quickly realized that uh intellectual property is really a set of legal tools, and that's kind of all they are, to capture the value of intangible assets that technology companies are creating all the time. Right. Um, and I think part of what enabled us to do this was uh my background, which was more holistic than you know, your typical patent attorney, um, having spent so much time looking at transactions and um the underlying technology as uh as a product council, essentially helping AOL, Bose product teams build those products, um market them, sales, taking them to market, that sort of thing. Um we're able to take a step back and look at separate out the concept of intangible assets, which technology companies are constantly creating, and the legal tools, forms of intellectual property, not just patents, that can be used to capture the value of those intangible assets. And uh as you know, many people know, that's actually kind of a matrix. There are intangible assets that can be um afforded legal protection in a number of different ways with using different tools depending on uh what it is you're looking at. So, for example, a software process may have patentable aspects, it may have copyrightable aspects, it may have aspects that uh is beneficial to the business to hold back as a trade secret. Um, same thing goes for um, you know, the ornamental aspects of products, um, you know, physical devices, uh, consumer electronics, um, all different aspects. So that is the that's the framework and the mindset that we bring to helping our clients. We start with now, in order to really answer the question, what is the business benefit uh of intellectual property? We found we had to step back even further. So to answer your question, now when we engage with clients or even when we talk with prospective clients, we have to begin that conversation um way earlier than that. We have to start with so tell me about your business. What is your product or service? How do you deliver that to your customers? Uh, how does your business make money? All of these questions are geared toward one, understanding the business fundamentals, trying to understand where they are in the business journey. But second of all, it enables the second part of the conversation, which is what are the intangible assets that are generated by the business? And more to the point, what are the key intangible assets, just not necessarily all of them, that are aligned with the business goals that essentially enable the business to make money, that provide them with the opportunity to win in the marketplace. Yeah. And identifying what those key intangible assets are, now we can take the next step and look at what, if any, legal tools, forms of intellectual property may make sense for that business. And we build a strategic recommendation for the business uh from that holistic framework.

Steve Brachmann 10:51

Great. Generally speaking, what kind of issues do the companies you work with have in identifying their key uh intangible assets?

David Hyams 11:02

Well, that's a really interesting question, right? Because you would think, uh, I think a lot of attorneys might think that, all right, I'm engaging with this business. Uh they they're the business experts. And in fact, that's what we as lawyers have constantly been taught, even from uh law school on, right? We're taught to carve out and not worry so much about or even actively ignore business considerations, right? Business questions. Those are often, you know, we're taught that. Like I certainly know I was, those are business considerations. Uh that's that's up to the business people to make those decisions. Your job is to make legal recommendations. Um, the only problem with that mindset is that law doesn't exist in the abstract. It if we're if we're doing business law, it has to serve the needs of the business. And uh the services and the recommendations that we provide have to be aligned. So uh the assumption that, and one of the things I found is that the uh underlying assumption that the business um should know what it's doing, always does know what it's doing, may or may not be totally true. Right. And so part of our role as uh, you know, uh trusted guides and experts in this area is to unearth some of those things. So to your question, um, you know, how do we help uh it the question kind of answers itself. We help these companies by helping them think through what are your key intangible assets.

Steve Brachmann 12:35

So you've had this career uh spanning different technologies, working in many different sectors. Uh, we like to just discuss the evolution of different technologies, and and what have you seen in your career that you've worked with, uh maybe a sector of technology that's really advanced, uh just in a way that's impressed you, you wouldn't have been sitting 25 years ago thinking, wow, this area is gonna go there.

When Miracles Become Mundane Tech

David Hyams 13:01

Well, I ultimately we're gonna head in the direction of AI because that's in the top of mind. But right before we get there, um I I have to go back to 2007 and uh the the news article that I saw that uh that said that um Apple was launching a phone. And um the more I read, the more I saw that uh that Apple was gonna use iOS as the operating system for their phone. And if I were really smart, I would have bought as much Apple stock as I could have afforded at that moment because I had realized this is a brilliant idea. They were gonna put uh an Apple, essentially computer running iOS in the pockets of as many people as would buy these phones. Now, of course, I I'm not that smart and I didn't do that.

Steve Brachmann 13:53

Many of us are n't, so don't kick yourself.

David Hyams 13:55

So here I am, right? But um even then, though, I never would have imagined, as I think a lot of us, that uh these these devices would become so ubiquitous and such an incredibly important part of our lives. Uh, I'm entertained by the pictures that surface every once in a while of um here's all the stuff that you used to take with you on vacation, and then here's your phone, right? So it's replaced cameras and diaries and maps and travel guides and all these other things that we now uh now are just sort of abstract information that we access on our phones.

Steve Brachmann 14:31

Right. And it makes perfect sense that we'd be heading here uh form factors of computers getting much, much smaller since 1960s, Moore's Law holding sway and all. Um but yeah, no, uh I I remember having my first uh smartphone. I was always very late to technological waves, so I waited about two years. And even then you're still gobsmacked by what are all these things uh that you can accomplish with this. The old there's an app for that, uh which has gone out of uh practice as an advertising slogan, but still holds holds our imagination in a big way.

David Hyams 15:12

Well, what's interesting to me about this is obviously not just the advance of technology and uh you know, the like think about the Artemis mission that just uh flew around the moon and uh just the uh staggeringly different amounts of computing power that went with Artemis as opposed to went with Apollo, right? Um, but that also it strikes me that uh it's says more about people that the miraculous becomes mundane so incredibly rapidly. Um, you know, knowing a little bit about how telecom networks work, um, I'm uh also, I guess, surprised, but also not surprised at sort of how grumpy people get when these miracle devices that we carry around in our pockets uh fail 1% or less of the time.

Steve Brachmann 15:58

We get very entitled to our technology working the way that we want it to.

David Hyams 16:02

Yeah, it's interesting.

Building An AI-Forward Law Practice

Steve Brachmann 16:04

And this being a good time to look at the advent of new technologies uh and how those are currently developing. Uh, do you do much work with Longship Legal in uh AI implementation or working with any sort of AI training at all?

David Hyams 16:18

Yeah, we work with a number of AI companies uh as clients, and we also make use, heavy use of AI tools within our practice as well. Um, you know, Longship, we formed it in 2023, and um we uh the the firm is entirely built in the cloud. So we're a digitally native um and AI forward uh operation, and uh that's that's proven to be really advantageous for us in a lot of different ways. I found that the cloud infrastructure supports my uh business development efforts and enables me to travel like seamlessly uh just about anywhere and access all the information of the firm. I mean, that's sort of a really down in the weeds example, but um, it's been um a lot of fun and it's been very, very useful.

Steve Brachmann 17:05

What kind of artificial intelligence technologies are you interacting with? Because we say AI, uh, but it's such a broad definition. Um What kind of AI solutions have you worked with implementing?

David Hyams 17:19

Yeah, well, that's a good point you raised. Um, you know, AI as a term has become such a broad blanket term that it almost doesn't mean anything anymore. It's a little bit like the word technology at this point. So...

Steve Brachmann 17:31

A bit.

David Hyams 17:32

Just as an extreme example, right? Ubiquitously, we all use spell check, right? It is an automatic process running in the background. And um, you know, we it it's completely unremarkable now. But in some, you know, the broadest view, that is actually a form of intellectual property. That is a form of artificial intelligence.

Steve Brachmann 17:51

Right.

David Hyams 17:52

Um the even before LLMs came out, large language models, um, I was doing a lot of work in uh machine vision and um computer or robotic mapping and um uh driver assistance uh systems that used um you know machine vision and decision making and path planning, um, and uh I guess uh systems that also enable drones to navigate, sense and navigate in three-dimensional space. Um all of these sort of fall under the purview of AI. So when we're talking about AI today, I think we're really talking about the sort of narrower area of large language models and how they interact with large sets of data and provide um a much more sort of natural language user interface.

Steve Brachmann 18:47

Sorry, go ahead.

David Hyams 18:48

I was gonna say we we work with uh you know a number of clients that are leveraging um large language models and different aspects of AI, sort of data analytics, and uh you know, just use leveraging the uh a computer's natural ability to uh chew through vast amounts of data and do pattern matching and data analytics um in a number of different ways. And um we also use uh AI tools, generative and others in our practice.

Steve Brachmann 19:20

And you work with you said AI companies, and uh part of your work is identifying key intangible assets. AI is interesting. You software more generally, you have patentability issues. Um so I wonder, do you ever have to talk about, and we don't have to go deep into section 101 here, but uh in identifying intangible assets with these AI companies, uh what kind of issues do you have given the context of you sometimes have uh uncertain patentability in those areas?

Separating Intangible Assets From Patentability

David Hyams 19:59

Right. So I mean the the 101 issues, 35 USC 101 issues you just alluded to, that would sort of affect the strategic recommendations we would make at the end of our analysis, um, where uh we were talking about specific kinds or specific forms of intellectual property that may or may not be relevant to the business. Um those considerations wouldn't affect the business or the technical analysis of identifying key and tangible assets. And I think this is one of the reasons that we separate those out, right? The so let's pick a, you know, let's a hypothetical company that is using some kind of large language model and it accesses some large data set. Um, let's call it a proprietary data set, right? Something that the company has assembled, they're providing a service uh that uses their proprietary data set. Uh the system also, again, in this hypothetical includes a you know LLM front end and perhaps some reasoning engine that uh uses some kind of structured reasoning to analyze the data set. Um we could look at all of that and we would sort of run through the analysis of our well, are are there is there anything in here that um before we get to the you know what would be considered patentable, what would we want to hold back as trade secrets, what you know, just copyright relevant here? Um we would have to go through the process of analyzing, first of all, what what is the business doing? Oh, they're providing a service. Okay, they're providing a data analytics service, and their clients interact with the system in the following ways. And these are the inputs and these are the outputs. Um, where are those data sources coming from? Are they scraped somewhere? Are they um contractual in some way, right? Or are they buying data or uh accessing it through some kind of licensing deal? Um, because contractual. Those contractual relationships, the access to the information, the ways that they use, they, the company, use contracts to provide their services and to govern how the services can be used, to govern how contractors and employees build the services. All of those are also potential intangible assets of the company, right? One of the weird and fun things about intellectual property is the way that it cuts across traditional silos within a business, right? So you can have intellectual property implications in contracts, in employment agreements, certainly in licensing agreements, customer agreements. How do you want to protect the intangible assets that you've built within the company via contract? Of course, there are sort of your traditional IP considerations of patent, trade secret, copyright, trademark for branding, that sort of thing. Yeah. But uh so I hope that this by explaining the framework a little bit more, it's giving you a better sense of like why we break apart uh looking at the identifying the intangible assets in a company versus down the road, then making a strategic or decision about what forms of legal tools you might use to capture the value of the intangible assets.

Steve Brachmann 23:20

Right. And that is the way most legal teams work in identifying what do we have here as an IP asset. There are aspects of this that are a little out of order compared to how we might look at this. A typical patent attorney might sniff a patentability issue and just shut it down right there. Whereas you're looking through and trying to find the intangible assets and the way that lawyers look at risk sometimes and how that can kind of uh end the conversation a little early.

David Hyams 23:50

Yeah, I think our approach is a little bit more holistic and a little bit orthogonal to the traditional approach. I mean, certainly we used to be those patent guys, right? And the the only thing that I did in my previous job was look at patentability. And um, it's a narrower lens. Sometimes that's what a client is looking for, but uh, we work with earlier stage technology companies. So they are uh the the clients that we've worked with are fairly appreciative of the holistic approach. Um, we're looking at uh ultimate value to the enterprise rather than acquiring uh, you know, patent documents.

Steve Brachmann 24:28

So engaging in artificial intelligence in a few different ways. You're actively using some tools, you're actively working with some companies. Uh I imagine you have a unique perspective on the sector.

Agentic AI Feels Familiar

Steve Brachmann 24:41

Uh, what do you see right now? We talk a lot about agentic AI. Uh we've gone from generative a bit to agentic and what can AI do as an agent for you. What are you seeing in the current wave of AI that might be interesting that maybe the mainstream hasn't picked up on yet?

David Hyams 24:58

Yeah, that's an interesting question. I mean, certainly the buzzword of the day is agentic.

Steve Brachmann 25:05

Y eah.

David Hyams 25:05

And, you know, looking at AI agents and what what can they do for us and how revolutionary they're going to be. Um I, you know, I have enough gray hair which you can't see on this podcast. Uh to remember back well, we were talking about agents back in 1999 and 2000 and 2001. Right. Even then, we were talking about software agents. We were talking about AI agents, although that is probably in retrospect a form of artificial intelligence talking about those agents. Um, it was the idea of enable, obviously, it's the idea of enabling uh a piece of software to go off and uh semi-autonomously or autonomously do something for you, the user. Um, we're talking about AI agents doing something very similar to that, right? Um and perhaps the uh you know the agent loop might be a little bit different, maybe, than what we were talking about back in '99, 2000, 2001 timeframe. But the underlying concept is still very, very similar. And I, you know, like I think these are fantastic tools. I enjoy using them and we use them very effectively in our own practice. Obviously, our clients are building businesses based around them. In a lot of ways, they are evolutionary rather than revolutionary because they are built on the under the same underlying computing platform, the same underlying sort of compute logic that we've been using for a long time in in other areas, right you know, which isn't to say that they're not useful. I really enjoy them um and I think they're very cool and very useful. Um, but u h ...

Steve Brachmann 26:41

Everything old is new again sometimes.

David Hyams 26:45

Sometimes.

Steve Brachmann 26:46

Yeah, yeah.

AI Accelerates Work Not Judgment

Steve Brachmann 26:48

Society talks, people talk about AI uh in terms of what is it going to replace? And you understand that as a conversation that people have. But in truth, AI really augments uh and we we just find different ways of doing the job that we had done before. Where do you see AI augmenting and kind of changing what you do on the daily uh from a productivity standpoint?

David Hyams 27:17

Yeah, this is a really good point. Um you know, go back to November, December of 2023, and uh in particular folks in the legal uh the legal profession were frankly freaking out, right? Oh goodness, uh AI is coming for our jobs, we're not gonna have jobs, AI is gonna automate all these things and take away our jobs. And um look, uh first of all, no. Second of all, if your if your job was built entirely around typing, then yes, you're in trouble.

Steve Brachmann 27:51

Yeah.

David Hyams 27:52

Right. But there are things that AI can do really, really well and things that AI cannot do at all, right? And um, you know, so I'm I'm fond of well, let me let me get into that. Um in our business, we've found that uh AI systems are accelerators of business processes, right? They are phenomenal at automating certain things. And um we use um AI-powered legal research tools. We use AI drafting tools as well in order to uh reduce what essentially was uh the grunt work of typing certain things.

David Hyams 28:30

So, in that respect, the the real promise of AI to eliminate certain sort of drudgery from our work has actually come true, at least in our practice, right? We're able to generate documents much more quickly than we ever were before. But the real question is, what is the structure of those documents? What is the underlying information and logic of those documents? Or when you're talking about a contract, for example, uh, what's the underlying business deal? If that hasn't been thought through and carefully negotiated by both sides uh to achieve the inputs that go into that call it typing exercise, then uh that's nothing that AI can cure, right? Right. Uh I'm fond of saying the matrix cannot tell you who you are with reference to that movie. But AI accelerates business processes. It enables certain things to be accomplished much more quickly than they ever could before. But so what AI can't do is shape your business. It can't identify your key differentiators, it can't understand why you went into business, what your vision for the business is? Um, what are your competitive advantages, things that you do that your competitors can't or won't? Um, what are the key intangible assets like inventive concepts? And why, which of those 35 things that AI can identify by munging through all of your enterprise data, which of those is the most useful and the most important for capturing your key differentiators and your business advantage? So judgment analysis client relationships. AI can't do any of these things.

Steve Brachmann 30:19

The human touch, yeah. Yeah. On that note, and taking that as a springboard, uh, because we do talk about you know the human aspect and the human touch that still needs to be brought to um any solution where AI is implemented. Uh yet we still do have tedium in our jobs. What kind of tedium do you see on a day-to-day that you wish that there was some sort of AI solution to handle?

David Hyams 30:47

Oh, that's a really great question because you know, part of the reason that we founded Longship was to run away from the tedium, right? Right to sort of chase the fun in business and in doing so help our clients uh do better, more interesting, more valuable things and build value for their businesses. I don't know. As our business grows, we've been applying AI in areas that are susceptible to automation. Um, and I've found them to be very, very productive. I guess I don't have that million-dollar idea. Honestly, though, Steve, if I did, I you know, I should probably jump off the podcast and go start a business.

Steve Brachmann 31:27

Yeah, exactly. Thank you for sharing it with me. No, I'm kidding. Um sometimes the million-dollar idea is recognizing what part of your workflow can be automated through the AI tools that are available. As lawyers, a lot of our work involves earning a client's confidence. And we had spoken before about addressing the confidence gap that exists when applying AI. If you could address that.

David Hyams 31:53

So let me just give you one example.

Trust And The Legal AI Gap

David Hyams 31:56

I was talking with a client uh earlier this week, in fact, and um she was asking me to draft and review uh a kind of contract for her. And she said to me explicitly, you know, I could use generative AI to draft this contract. Um, but then I was talking with my business partners, and there were lots of things that we weren't sure about, or things that we weren't sure that uh we didn't know if there are any landmines or any tripwires in the contract. She was talking about sort of unknown unknowns, right? So but the other thing that she said, which was very interesting to me was Dave, I I trust you. I've been working with you and I trust your judgment. I trust your you know, ability to uh conform, you know, write this contract in a way that makes sense for my business. So in our practice, we've tested the general AI models and tools against proprietary tools. And I've found that um so these proprietary tools I'm talking about, you know, without naming names, they are um available for subscription. Uh, they leverage a form of LLM and they're built on uh very large proprietary data sets, right? So um they are they have available to them in in those tools things that aren't really available on the open internet. And so the general LLM tools aren't trained on them. They don't have access to that information. So, what am I talking about? Things like um case law, practice guides, um uh you know, best practices, uh practice notes, um, opinions, again, laws, regulations. Um and so uh what we are able to do with these tools is not just much faster, even in some cases, real-time legal research, right? But uh we're also able to use drafting tools that leverage uh those vast databases um for products that are much more accurate and um give us much more useful information. So, for example, I uh we've tested all these tools side by side. So I asked a general AI tool to draft a specific kind of agreement and gave it uh, you know, hypothetical parties and parameters I was looking for. And I gave that same information to one of these proprietary tools. And the contract that was generated by the uh general AI tool was it was fine and it sure looked like a contract. And if the parties had signed it, it, you know, hypothetical parties, it would have been enforceable. But the specific the um proprietary tool was able to flag issues for me and say, well, in this jurisdiction, you're gonna have an enforceability issue with this and this, and you need to rewrite it in this way. And here's the case law, and here's why. Um much more accurate and useful uh uh product and much more useful information. Um, the in fact, the the kind of information we absolutely have to have. Now, it's the sort of thing that we would have found previously uh doing our own legal research as we were trained to do back in the day, um, that it's uh much faster to it. So I use general AI tools with great caution.

Steve Brachmann 35:32

Right.

David Hyams 35:33

Because they're only as good as the data sets that they've been trained on. Turns out things that they don't necessarily have access to, the vast amount of information that are hidden behind firewalls in law firms, um, and you know, in all of our own uh years and years of experience, drafting contracts, drafting patent applications, uh interacting with clients, um, our knowledge of what certain fact patterns work or don't work in certain jurisdictions because we've had that experience.

Steve Brachmann 36:05

Yeah.

David Hyams 36:06

Those sorts of things.

Training Data Behind The Firewall

Steve Brachmann 36:07

Yeah, and I'm sitting here thinking to myself, there's there's multiple issues with this from a legal standpoint. Tough to have an AI contract generator when most of the contracts that would be your training data really confidential information hidden behind uh, like you said, the the firewall of a law firm. And so can can you uh so the legal industry has its own uh issues with uh adapting legal um adapting AI tools to legal solutions just from the standpoint of a lot of the training data uh would trigger ethical um issues if that was just used as an input.

David Hyams 36:49

Right. Well, so this is why uh a lot of uh a lot of large law firms have solved that problem by bringing some kind of LLM instantiation inside their firewall and training it on their data, right? So so now they've automated their corpus of knowledge within their firewall, which is great so long as what they're trying to do uh can find a result that resides within that body of knowledge.

Steve Brachmann 37:16

Right.

Jazz Piano And Closing Thoughts

Steve Brachmann 37:17

So, David, we're getting to the end of our time here together, and I always like to spend just a couple of minutes at the end um looking at the person behind the legal career and asking, where do you find some personal zen outside of patent law as interesting as patent law is?

David Hyams 37:36

Personal zen. I like this question. Um honestly, before I even thought about uh medicine and certainly long before I thought about law, um, I have been a jazz piano player. Um, since high school, I've I took it really, really seriously and studied, and that's honestly part of what I studied as an undergraduate. Um, and um I uh to this day uh I am the house piano player for a uh 18-piece jazz band here in the uh jazz ensemble in the DC area. Um, and I play uh as many sort of small group and uh medium group uh performances as I can, and that's you know, public performances and private parties. Um I do the occasional appearance at uh Washington, DC's uh storied jazz venue called Blues Alley, and uh it's it's a lot of fun, and it's another way of interacting with people. Uh, the best part of jazz piano for me is the group interaction. As much as I enjoy playing solo, um, it is uh much more fun to interact with a group of skilled jazz musicians because uh, as you know, one of the uh key aspects of of jazz is improvisation. And yeah, uh for me, the fun of spontaneously creating a you know unique in the moment, uh never to be heard again uh musical creation is uh incredibly fun and satisfying.

Steve Brachmann 39:07

You're very creative outlet, I'm sure. David, thank you so much for taking the time to talk with us today.

David Hyams 39:13

Thanks very much for having me, Steve. I really appreciate it.

Steve Brachmann 39:16

Absolutely. I'm Steve Brachmann for IP Innovators. Happy patenting.