Latest / Sidecar Sync / 28: Intellectual Property & AI Armor with Patent Attorney Robert Plotkin

Transcript

28: Intellectual Property & AI Armor with Patent Attorney Robert Plotkin · Sidecar Sync

• 0:00
  Speaker 1: There are many ways in which AI actually can create increased demand or supply and having this fixed mindset that, oh, if we reduce the amount of labor, that means there's going to be less total work to do, there's all kinds of ways in which the demand is elastic.
• 0:18
  Speaker 1: In a sense, is what I'm saying from an economic point of view.
• 0:24
  Speaker 2: Welcome to Sidecar Sync, your weekly dose of innovation.
• 0:27
  Speaker 2: If you're looking for the latest news, insights and developments in the association world, especially those driven by artificial intelligence, you're in the right place.
• 0:36
  Speaker 2: We cut through the noise to bring you the most relevant updates, with a keen focus on how AI and other emerging technologies are shaping the future.
• 0:44
  Speaker 2: No fluff, just facts and informed discussions.
• 0:47
  Speaker 2: I'm Amit Nagarajan, chairman of Blue Cypress, and I'm your host.
• 0:53
  Speaker 3: Hello everyone and welcome back to another episode of the Sidecar Sync.
• 0:57
  Speaker 3: My name is Mallory Mejiaz.
• 0:58
  Speaker 3: I run Sidecar, the company that brings you this very podcast, and I'm one of your co-hosts, along with Amit Nagarajan.
• 1:05
  Speaker 3: Now today, we have a special interview episode lined up for you with Robert Plotkin.
• 1:11
  Speaker 3: We're talking all about artificial intelligence and IP.
• 1:15
  Speaker 3: For over 25 years, ai patent attorney Robert Plotkin has secured patents for his clients' innovative software, but he doesn't stop there.
• 1:23
  Speaker 3: He then guides them in deploying those patents to withstand the pressures of new competitors and big AI and win historic patent sales and licenses.
• 1:32
  Speaker 3: A lifelong AI aficionado and MIT computer science graduate, robert processes a unique technical background among lawyers.
• 1:40
  Speaker 3: As a patent attorney and co-founder of the boutique patent law firm BlueShift IP, he secures and leverages IP to attract investment, raise funds, generate revenue and secure successful exits on behalf of high-tech clients.
• 1:55
  Speaker 3: Decades ahead of his time, robert first shared his AI patent strategies in 2009's the Genie in the Machine how Computer Automated Inventing is Revolutionizing Law and Business.
• 2:07
  Speaker 3: Now, 15 years later, amidst a global AI frenzy, robert reveals his revolutionary mind process in his new book, ai Armor, which is out now.
• 2:17
  Speaker 3: So we will definitely be linking to that in the show notes.
• 2:20
  Speaker 3: So not only do we have a special edition episode today with Robert, but we have a big announcement Robert is actually going to be one of our keynote speakers at our Digital Now conference this year in DC.
• 2:33
  Speaker 3: The dates for that are October 27th through the 30th 2024 at the Omni Shoreham Hotel in Washington DC.
• 2:41
  Speaker 3: If you enjoy this conversation at all which I think that you will we will be diving deep on this topic at the conference.
• 2:48
  Speaker 3: So I highly encourage you to register.
• 2:50
  Speaker 3: And not only that, but we have a special code for the people who are listening to this podcast right now.
• 2:55
  Speaker 3: If you use the code POD P-O-D to register at digitalnowconferencecom, you will get 10% off the registration price.
• 3:05
  Speaker 3: So what do we talk about in today's episode?
• 3:08
  Speaker 3: Let me give you a little bit of an overview.
• 3:10
  Speaker 3: We talk about IP.
• 3:12
  Speaker 3: What is it and how exactly has it evolved in this AI frenzy that we're currently living in?
• 3:18
  Speaker 3: The great question that we've asked ourselves on this podcast, and you've probably asked yourself too can you copyright AI-generated content?
• 3:26
  Speaker 3: Who owns it?
• 3:28
  Speaker 3: Another question we'll talk about in this podcast how can associations use their assets to bolster a competitive moat?
• 3:36
  Speaker 3: And finally, this is maybe something you haven't thought about, but should associations be thinking about a patent strategy.
• 3:43
  Speaker 3: We will cover all of these topics and more in today's episode.
• 3:47
  Speaker 3: We'll dive right in, but first a quick word from our sponsor.
• 3:50
  Speaker 3: Today's sponsor is Sidecar's AI Learning Hub.
• 3:53
  Speaker 3: The AI Learning Hub is your go-to place to sharpen your AI skills and ensure you're keeping up with the latest in the AI space.
• 4:01
  Speaker 3: When you purchase access to the AI Learning Hub, you get a library of on-demand AI lessons that are regularly updated to reflect what's new and the latest in the AI space.
• 4:11
  Speaker 3: You also get access to live weekly office hours with AI experts and, finally, you get to join a community of fellow AI enthusiasts who are just as excited about learning about this emerging technology as you are.
• 4:24
  Speaker 3: You can purchase 12-month access to the AI Learning Hub for $399.
• 4:29
  Speaker 3: And if you want to get more information on that, you can go to sidecarglobalcom slash hub.
• 4:34
  Speaker 3: Robert, I want to thank you so much for joining us on our podcast today.
• 4:40
  Speaker 3: You know, in previous episodes, amit and I have talked a little bit about IP, but it's not something we've been able to dive deep on, so we're super excited to have you on the podcast.
• 4:49
  Speaker 3: How are you?
• 4:51
  Speaker 1: I'm doing great.
• 4:52
  Speaker 1: Thanks, Mallory, Thanks Amit.
• 4:53
  Speaker 1: It's really great to be here.
• 4:54
  Speaker 1: Thanks for having me.
• 4:56
  Speaker 3: For sure.
• 4:56
  Speaker 3: I'm hoping you can share a little bit with our listeners about your background, what you've done in the past and how you got here.
• 5:04
  Speaker 1: Yeah, I am an intellectual property attorney.
• 5:06
  Speaker 1: I focus primarily on obtaining patent protection for growing tech companies from their founding stages all the way up through growth and exit.
• 5:17
  Speaker 1: My background is in computer science.
• 5:20
  Speaker 1: I started programming as a kid back in the early 80s, loved it, went on to study computer science, then went on to become a lawyer and been obtaining primarily software patents for over 25 years to bolster their company, to increase value, to generate revenue and, particularly, to help them achieve those successful exits because that's what everyone is shooting for Partnerships, mergers, acquisitions, whatever it happens to be.
• 5:55
  Speaker 1: I enjoy using IP as a strategic business asset and maybe we can talk a little bit about that today in the context of associations.
• 6:05
  Speaker 3: Absolutely, and I think we definitely will get into that.
• 6:08
  Speaker 3: But first I think we should go back to the building blocks to kind of set the stage for this conversation.
• 6:14
  Speaker 3: What is IP, kind of by definition and in your own words, and why is it so important?
• 6:21
  Speaker 1: Yeah, intellectual property broadly is a set of legal rights that an individual or an organization or a company owns in their intellectual creations.
• 6:33
  Speaker 1: So that could be a piece of art, it could be a blog post, it could be a website, a video, music.
• 6:42
  Speaker 1: On my side of things, where I deal with patents most of the time, it can be a physical invention a car engine.
• 6:49
  Speaker 1: It could be a drug, a new kind of chemical or what I deal with most of the time, new type of software, new type of algorithm or process implemented in software.
• 6:59
  Speaker 1: Anything that is new and useful that solves a practical problem can be protected by a patent.
• 7:05
  Speaker 1: So, whether it's a copyright in a creative or artistic work or a patent, or a trade secret in something like a formula the formula for Coca-Cola is a very, very famous trade secret Whether it's any of those types of IP, if you as an individual or organization own it, you have a legal ability to stop others from using that IP without your permission.
• 7:33
  Speaker 1: That's why it's usually called intellectual property.
• 7:37
  Speaker 1: It's somewhat analogous to just owning a physical thing like your car, where someone cannot legally drive it without your permission thing like your car, where someone cannot legally drive it without your permission.
• 7:48
  Speaker 1: But IP applies to ideas, information, data, inventions, songs, websites, anything like that, where someone else cannot copy or use, reproduce that work without your permission.
• 8:01
  Speaker 1: So it can be a very, very powerful type of a tool or an asset to have.
• 8:07
  Speaker 1: For that reason, because in the information age and we're probably moving past even the information age now increasing number of organizations have their primary value, their primary competitive advantage, in intellectual property rather than in physical assets where companies used to have their primary competitive advantage that machinery or real estate still important.
• 8:35
  Speaker 1: But increasing percentage of the economy, increasing number of entities, have their primary value in these intangible or intellectual assets.
• 8:47
  Speaker 2: Robert, one follow-up question I have on that opening area of discussion is in the association context there often is a tremendous amount of content and in some cases data, like registries or other data, that's been collected in a sector that an association may represent.
• 9:05
  Speaker 2: Do those elements, do those constitute intellectual property by your definition?
• 9:11
  Speaker 1: Yeah, they would typically be protected by copyright or trade secret or combination of both.
• 9:17
  Speaker 1: The customer lists, any data collected from customers, although it probably would not be copyrightable in most cases by the entity, but it could be maintained as a trade secret and it can be extremely valuable.
• 9:33
  Speaker 1: I mean, we may talk about it more, but as data generally and information and maybe even knowledge, becomes more of a commodity, any entity has to ask what differentiates us, what gives us that extra layer of value that our competitors don't have.
• 9:50
  Speaker 1: And very often it's that private data that can be the basis for real competitive advantage and therefore for valuable intellectual property in the form of trade secrets.
• 10:01
  Speaker 2: I know we'll dig into more of this in the near future, but associations we often discuss with leaders how their brand and their content are, in many respects, latent assets that can be leveraged far more greatly in the era of generative AI and AI more broadly and as we get into the conversation.
• 10:22
  Speaker 2: I just wanted to put that out there because it's the backdrop for some of the thinking that I know many of our listeners will be pondering, which is you know, how does this apply to me?
• 10:30
  Speaker 2: How do I protect my assets, how do I leverage my assets?
• 10:33
  Speaker 2: And in this sector there are some groups that have software to protect or software to leverage.
• 10:38
  Speaker 2: But and I think there's an opportunity for a lot more of that because generative AI makes it so much easier to create novel software, which is a whole other, I think, interesting conversation.
• 10:47
  Speaker 2: But I just wanted to put that out there to contextualize some of this for our typical listener.
• 10:53
  Speaker 1: Well, you mentioned brand and I didn't mention trademarks, which is the other major type of intellectual property, and you're right.
• 11:00
  Speaker 1: For associations, very often that's going to be an important, valuable piece of the picture.
• 11:06
  Speaker 1: A trademark is anything like a product name, a slogan, a logo, often a phrase that's associated with a product or a service that identifies its source.
• 11:21
  Speaker 1: I'll use Coca-Cola again, one of the most famous brands out there.
• 11:25
  Speaker 1: You might say well, there's lots of sodas that taste similar to Coke.
• 11:31
  Speaker 1: Why does it maintain such a huge market share?
• 11:34
  Speaker 1: The brand is a huge part of it.
• 11:37
  Speaker 1: People associate quality and taste and enjoyment and good experiences and all of these things and reliability that they know whenever they buy a can of Coke it's going to taste the same in the way they expect it.
• 11:50
  Speaker 1: They associate that with the brand.
• 11:52
  Speaker 1: So that brand in the law is protected using trademarks and that's a very, very valuable asset for a company to have.
• 12:02
  Speaker 1: Particularly in industries or contexts where there isn't a lot of differentiation based on what inventions you have or technology or maybe even information, it's the brand that could be the real distinguishing feature.
• 12:18
  Speaker 2: And in an era, as you mentioned, where even knowledge is potentially going to be commoditized, even in a highly vertical, narrow domain like a specialty area of law, for example, there is an opportunity for the association with oftentimes a very strong brand in that particular discipline or a particular geography to leverage that brand to do some really interesting things.
• 12:39
  Speaker 2: So we'll get into that more.
• 12:41
  Speaker 1: Yeah, I mean and you know much more about this than me but I can only guess one way in which that's relevant is if you have an association that's been around for a long time, you know there might be a double-edged sword where it may have trouble keeping up with new, innovative upstart organizations, but it's got that old, well-known brand that's trusted and that could be the thing that's really useful in its corner.
• 13:07
  Speaker 2: Well, as an IP attorney, I'm sure the answer to this question is yes, because there's an association for everything, but I'm assuming there's an association for intellectual property attorneys or perhaps even patent attorneys and groups like that, and perhaps exactly what you just mentioned.
• 13:22
  Speaker 2: They may have been around a while and so they have trust and credibility, but perhaps they're not necessarily on the cutting edge in terms of how they deliver or disseminate information or how they connect people.
• 13:32
  Speaker 2: And there's both kind of the risk piece and also the opportunity if a brand like one of those associations were to really aggressively embrace AI and reduce friction and improve engagement in the ways that we talk about a lot here.
• 13:46
  Speaker 2: Yeah, absolutely.
• 13:48
  Speaker 3: Well, before we chat about the AI landscape now, I want to go back a little bit with you.
• 13:53
  Speaker 3: Robert, you seem to be interested in AI.
• 13:55
  Speaker 3: Before it was cool, quote unquote I'll say you wrote a book called Genie in the Machine all the way back in 2009 about AI-generated inventions.
• 14:04
  Speaker 3: I'm curious what inspired you at that time to be interested in AI.
• 14:09
  Speaker 1: Yeah, in fact it goes back even further.
• 14:11
  Speaker 1: I said I was programming when I was a kid.
• 14:13
  Speaker 1: I entered a science fair in high school it must have been 1987 or so with a piece of software I wrote that used a pretty primitive form of machine learning to learn how to play tic-tac-toe, and it was based on a project in the 60s, someone named Mitchie, who did this in a physical computer using matchboxes that were labeled with different tic-tac-toe boards and he had different numbers of beads in them to represent the weights for those of you in the audience who know about machine learning.
• 14:52
  Speaker 1: So he trained a machine learning model in a physical computer.
• 14:57
  Speaker 1: Anyway, I remember learning a lot from that science fair experiment, early hyper machine learning in the 80s.
• 15:04
  Speaker 1: But going to the book the Genie and the Machine, which was about how AI back in the 90s and early 2000s was being used to automate the inventive process and the implications of that for patent law, for example, I address things like can an invention that was invented by or using AI be patented?
• 15:26
  Speaker 1: Would it qualify for patent protection if the invention was designed primarily by a piece of software, for example?
• 15:35
  Speaker 1: And I got motivated to write about that because I was seeing it happen.
• 15:39
  Speaker 1: I was seeing it happen again back in the 90s even, and the early 2000s.
• 15:43
  Speaker 1: I was seeing it happen again back in the 90s, even in the early 2000s.
• 15:47
  Speaker 1: Of course, the technology then was not nearly as widespread or advanced as it is now, both because of all of the orders of magnitude of improvements in hardware and also software.
• 15:55
  Speaker 1: I mean, I wrote the book before deep learning was around.
• 15:59
  Speaker 1: There's been a few really major algorithmic and software advances since that time, but when I wrote the book there were already a few patents that had been granted on what you might call computer-generated inventions.
• 16:12
  Speaker 1: I had an example of a team at NASA that designed an antenna using a type of AI called a genetic algorithm, and that antenna went on a NASA space station.
• 16:23
  Speaker 1: I interviewed an inventor who helped Oral-B create the first version of the cross-action toothbrush using a neural network based AI.
• 16:34
  Speaker 1: So it was real.
• 16:36
  Speaker 1: But I'll say that the book didn't get a whole lot of attention back then because I think people thought this isn't real or they just couldn't connect to it.
• 16:45
  Speaker 1: And then, of course, ever since around 2019, 2020, this has gotten more attention.
• 16:52
  Speaker 1: And then ever since ChatGPT came out the topic has just launched into space, but it's been very edifying and very satisfying for me to see people pay attention to this, and now the US Patent Office.
• 17:06
  Speaker 1: Patent offices around the world have all been grappling with this question of what's now being called AI-assisted inventions.
• 17:14
  Speaker 1: I like that term because there's always humans involved.
• 17:16
  Speaker 1: Can those be patented?
• 17:19
  Speaker 1: What should the rules be?
• 17:20
  Speaker 1: Even, should the AI be named as an inventor or a co-inventor?
• 17:25
  Speaker 1: All of these issues are being addressed by patent offices and courts around the world, and it's only going to become a more prominent topic.
• 17:35
  Speaker 2: I think it's super interesting, for our companies here at our family of businesses are always building new software and leveraging AI more and more, and so you know thinking about that from a software company perspective.
• 17:47
  Speaker 2: I think it's just a fascinating discussion and thought process of what is protectable, what is fundamentally the idea of invention when the machine is able to create such a large percentage of it.
• 17:58
  Speaker 2: I know authors are thinking about that in the context of writing books with copyright protection or the lack thereof in some cases.
• 18:04
  Speaker 2: In the context of writing books with copyright protection or the lack thereof in some cases, when we wrote our AI book for Associations Ascend last year, we heavily used AI, knowing full well that there's a very high likelihood that there's no copyright protection available for that particular title.
• 18:17
  Speaker 2: And in fact, for us we didn't care because we intended to freely distribute products and make it available to everyone in the world, and our goal was to disseminate the information and not to profit from the book itself but, to you know, share that content.
• 18:30
  Speaker 2: So for us it was a perfect use case and we weren't too concerned about it.
• 18:33
  Speaker 2: Plus, I tend to, you know, shoot first and ask questions later in the way I build companies, and it's just kind of the nature of entrepreneurship.
• 18:39
  Speaker 2: But for associations who are thinking actively about this exact topic, you know, for example, do they use AI to help them?
• 18:47
  Speaker 2: Author a journal article.
• 18:49
  Speaker 2: Will they have copyright protection or not?
• 18:52
  Speaker 2: What's the?
• 18:53
  Speaker 2: Is there a framework for thinking about that that you could share?
• 18:56
  Speaker 2: I'm sure that people are, at this point in the pod, anxious to hear a little bit of a decision tree, maybe on what might help them decide what maintains protection and what prevents them from getting protection.
• 19:06
  Speaker 1: Sure, and I am primarily a patent attorney, so it might be a little dangerous for me to veer too far into copyright, but I think the most important guideline for people to keep in mind, even as the law evolves on this question of what is and isn't copyrightable when you use AI to help you do something like write a book, is to ask what was the human contribution to this?
• 19:31
  Speaker 1: I think if something is entirely 100% AI generated, it may not be copyright, protectable by copyright.
• 19:40
  Speaker 1: It may not be copyright protected, protectable by copyright maybe in some places and not others.
• 19:46
  Speaker 1: But the more that you as a human contribute, the more easy it will be for you to obtain copyright protection.
• 19:53
  Speaker 1: And, to be honest, it's probably going to be better for you, from a business or organizational perspective anyway, to be more involved, as I'm sure you saw from writing the book.
• 20:04
  Speaker 1: You said it was heavily you know, relied heavily on the use of AI, but I'm sure that it reflects your personality and your ideas and your expertise and that you brought all that to bear on it as well.
• 20:16
  Speaker 2: Very much so.
• 20:17
  Speaker 2: Our process for the book was we got a group of people together who contributed to it.
• 20:20
  Speaker 2: We built a very detailed outline, largely without AI support, to generate the topics, and then each contributor basically wrote or spoke to a transcription tool to generate raw content.
• 20:34
  Speaker 2: And then we use the AI essentially to distill that content down, in some cases to summarize, in some cases to clean it up and then to stitch it together.
• 20:42
  Speaker 2: The other thing we had the AI do that was really interesting.
• 20:45
  Speaker 2: This was the early days of GPT-4.
• 20:46
  Speaker 2: This was last spring.
• 20:47
  Speaker 2: We fed the AI the nonfiction element and then we asked the AI to develop a narrative for a fictional plot that would go across multiple chapters, the idea being it's something along the lines of a business fable, right, where we were able to share a story about an association leader and what happened to them over time, and it was actually quite remarkable how good the AI was at that creative task, even compared to the other pieces.
• 21:14
  Speaker 2: So all the ideas were definitely ours, but we heavily used AI to create the copy and to refine the copy.
• 21:20
  Speaker 2: We're right now just kicking off the second edition the really meat of that work to update the book.
• 21:27
  Speaker 2: We plan to update it at least once a year and we're actually a little bit overdue for the update.
• 21:31
  Speaker 2: But we'll once again be heavily using AI.
• 21:35
  Speaker 2: But the ideas are certainly coming from us and from our audience of people who provide feedback, so there's a lot of human input.
• 21:42
  Speaker 1: I mean it's interesting.
• 21:43
  Speaker 1: The process you described is exactly the kind of process I both was observing and was suggesting in the genie and the machine in the inventive context, which is that AI, by becoming increasingly able to fill in the details of a goal or an idea or written work or an invention, would enable people to focus on the higher level ideation defining the problem they want to solve, defining the goal they want to achieve and then leaving it to the AI to fill in more of the low level details the equivalent in the inventive context of writing the outline, so to speak, and then letting the AI fill in the details.
• 22:25
  Speaker 1: It's very much analogous to what you do as a manager when you have people who are working for you or with you or when you're collaborating.
• 22:33
  Speaker 1: So in that sense, it's not any different and the process you described is one that I think gets the best of both worlds.
• 22:41
  Speaker 1: It enables the work to reflect, as you said, your ideas, not be bland or generic, to reflect your insights, to provide value to the reader that other people wouldn't have been able to provide, but it saves you time and effort in writing it and maybe, as you said, in the context of coming up with a creative story, comes up with something you couldn't have.
• 23:05
  Speaker 1: I did something similar last year.
• 23:07
  Speaker 1: I was writing an article about whether AI could be named as an inventor on a patent, so I wrote an essay my mind tends to work in sort of essay-like style and I asked ChatGPT to write a story, a fictional story, illustrating that concept, and it did, as you said, a surprisingly good job.
• 23:32
  Speaker 1: It imagined a science fair or an inventor's fair in which there were multiple entries with different inventions, one of which was created by an AI, and the human inventor was afraid that it was going to lose to the AI.
• 23:48
  Speaker 1: But in the end, the judges awarded the human and it had a moral to the story and it was quite good.
• 23:55
  Speaker 1: So I ended up publishing my essay along with the ChatGPT story.
• 24:00
  Speaker 1: This is before I saw your book, using the same sort of general approach, but if you can learn what the AI is good at, you can couple it with what you're good at, just as when you build a team or work together with a team of people right, everyone has their own strengths and weaknesses, and when you build a good team, you take people and you leverage the best from everybody, and the team can be better than every individual individually right.
• 24:26
  Speaker 2: We talk a lot about this, the concept of an explosion of opportunity coming from the availability of more creativity, and we talk about this a lot in our content here on the pod, but also in our courses and in our blog articles and so forth, and a part of what we're referring to is when you open up a modality of expression that is previously unavailable due to a lack of skill.
• 24:48
  Speaker 2: So, in my case, I'm not a fiction writer, I was never good at that, and so that whole idea for the book wouldn't have been possible without the aid of AI other than hiring a person to do it, and then we probably wouldn't have done it because of cost and time and all those things.
• 25:02
  Speaker 2: So, and a similar thing recently is that I don't know if you've had a chance to play with a tool called Sunoai.
• 25:09
  Speaker 2: We talked about it last week's pod, but it's a music generator and I've been using it ever since I discovered that tool.
• 25:15
  Speaker 2: I've been playing with it almost every day because it's so much fun, but I'm sending people songs.
• 25:20
  Speaker 2: My brother-in-law had a 50th birthday last week and I had it sing him a and like French jazzy vocals and stuff like that and throw in some funny stuff from a trip that me and my wife and they did recently and it's just awesome because it's like this creative expression that you didn't previously have available.
• 25:37
  Speaker 2: I have no musical talent whatsoever and first of all, he didn't know it was AI and secondly, he didn't even pick up on like kind of little Easter egg lyrics that were in there until we pointed it out to him and he said man, this is crazy.
• 25:48
  Speaker 2: So I think that's an opportunity for business communication.
• 25:52
  Speaker 2: It's obviously that we come back to in this context of opening up creative doors where previously you know they didn't exist.
• 25:59
  Speaker 2: So that gets to a really exciting form of dialogue really exciting form of dialogue.
• 26:09
  Speaker 1: I wonder, you know because I'll say something else that I use in the legal context and I wonder how it applies with associations, which is that, you know, lawyers are known for speaking in jargon.
• 26:14
  Speaker 1: We have our own language, essentially, and that can create a barrier for communication with clients or even sometimes with other lawyers.
• 26:22
  Speaker 1: I'm in patent law, which has its own terminology, and I do my best.
• 26:26
  Speaker 1: I think I'm pretty good at speaking in plain English to people, but it can be challenging and that's one way in which I've used AI to take something that I write in my own words and translate it to be more understandable to somebody else.
• 26:41
  Speaker 1: And I suspect, when you deal with associations, particularly ones working in a very narrow field or maybe just from being in the nonprofit world, maybe there's a certain lingo or jargon that people use that may make it difficult for them to communicate with either their members or funders or other people.
• 26:57
  Speaker 1: Does that resonate at all?
• 27:01
  Speaker 2: A hundred percent more people to reach them where they are at in a form that they can comprehend and absorb, whether it's a different modality, like taking text and converting it to audio for people who don't read written words well, or it's just preferred for them to listen, or perhaps from one natural language to another, from English to French to Spanish, or from one level of expertise to another.
• 27:23
  Speaker 2: So you might take a medical domain where a doctor is speaking to another doctor.
• 27:27
  Speaker 2: They might speak differently to someone outside of their field and certainly differently at least hopefully so to someone that's a patient or the family member of a patient, and so there's an opportunity around content, where you have an asset, like you know.
• 27:38
  Speaker 2: Take a particular branch of medicine or there's an association for that, and they might have some of the best content on the planet in that particular vertical domain, and usually it's deeply technical because it's related to the people who are the practitioners of that discipline, and so being able to translate that in an accurate way, but still in a different levels right depending on the education level or the context, I think it's a really powerful modality.
• 28:04
  Speaker 2: These are all different modalities of communication in my mind.
• 28:06
  Speaker 2: Context I think it's a really powerful modality.
• 28:07
  Speaker 2: These are all different modalities of communication.
• 28:08
  Speaker 2: In my mind, and that's part of what's so exciting about this is we've unlocked language and unlocked communication effectively more broadly, and so for associations, at their core they're communications machines, so it's super relevant to them in my mind.
• 28:20
  Speaker 3: Going back to copyright of AI generated content, robert, I just want to dig into that a little bit more.
• 28:41
  Speaker 3: I know your expertise is with patents, but just to kind of have this conversation and explore it with you, even just giving an AI model a prompt is significant enough human contribution.
• 28:51
  Speaker 3: So I'm guessing this is a question that the courts will need to answer within the next few years.
• 28:56
  Speaker 3: For sure, but have you seen any of these questions answered, particularly with patents, already?
• 29:03
  Speaker 1: Yeah, you know, so far the question has been answered somewhat differently in the patent context than in the copyright context.
• 29:10
  Speaker 1: I believe that the current answer from the US Copyright Office is that you need to identify, like, which part of the work was AI generated and which part was human generated, which I think is a really unfortunate solution.
• 29:32
  Speaker 1: One, as a practical matter, it may be impossible to separate the two out and secondly, I don't think that's really the right answer from a public policy perspective to say these words were written by an AI.
• 29:42
  Speaker 1: It also doesn't encourage people to make the best uses of AI in writing, which is to focus on what Amit said Come up with your unique ideas.
• 29:53
  Speaker 1: I would say, if we want to get the resulting work out to the public and as widespread a way, as quickly as possible, why should we penalize someone if 90% of the words that came out were quote generated by AI?
• 30:10
  Speaker 1: As long as there was a person who put their unique thoughts and spirit and personality in at the front and, as you said, mallory, perhaps even if it was just or primarily by putting in a prompt, I think that's the wrong way to go and I hope that things change on the copyright side, on the patent side.
• 30:32
  Speaker 1: So far it really hasn't been answered by the patent office so much, or even by the courts.
• 30:40
  Speaker 1: It's really being decided on a case-by-case basis, primarily based on the patent law's requirement of what's called non-obviousness, which is a little tricky to decide.
• 30:52
  Speaker 1: But basically what it means is you submit a patent application for a new mousetrap.
• 30:57
  Speaker 1: You have to demonstrate not only that that's new, that no mousetrap like that has existed before, but that there was some ingenuity required my patent lawyer friends would not like me using that term but that there was some degree of skill required that went beyond the norm to create that mousetrap.
• 33:14
  Speaker 1: Or perhaps that the mousetrap would be if you showed it to a person who a mousetrap engineer, who normally invents mousetraps if you showed it to that person, they would be surprised by how it works.
• 33:27
  Speaker 1: That would be one indication that the invention is what we call non-obvious and therefore patentable.
• 33:33
  Speaker 1: So if you apply this to the context of AI, one example would be if, let's say, you have a new drug with a chemical structure, you might ask could someone have just taken an existing description of the function they wanted the drug to perform and put that into an off-the-shelf AI product and get that chemical structure out without any additional effort.
• 34:02
  Speaker 1: If so, it might be that that chemical structure is considered obvious and not patentable.
• 34:09
  Speaker 1: Said a different way, you generally are going to have to prove to get a patent that you couldn't create the invention you're trying to patent just by putting a sort of obvious or trivial prompt into a system like a chat GPT and get an answer out.
• 34:29
  Speaker 1: That's going to be sorry to interrupt.
• 34:32
  Speaker 2: That's going to be a very interesting test to continue to apply as these AIs become radically smarter over time.
• 34:40
  Speaker 1: It's going to be.
• 34:41
  Speaker 1: Yeah, and I identified this in the book years ago.
• 34:43
  Speaker 1: It's very hard even to apply practically, because how is a patent examiner going to work backwards?
• 34:51
  Speaker 1: Essentially, think about it that again.
• 34:55
  Speaker 1: Let's use the example of a mousetrap.
• 34:57
  Speaker 1: If you submit a patent application for a mousetrap and you say my mousetrap has a spring and a lever and all these different components, really the patent examiner should ask is that the kind of thing that anyone with what we call ordinary skill could have designed with the push of a button using some off-the-shelf AI software?
• 35:18
  Speaker 1: How is the patent examiner really going to do it?
• 35:20
  Speaker 1: Are they going to need to sit at their desk with some AI design software and try putting some descriptions into it that are trivial and see if they get out the invention that you're trying to patent?
• 35:31
  Speaker 1: It's going to be quite hard.
• 35:33
  Speaker 1: It's going to be quite hard, but I'm not sure that there really is a better way to do it.
• 35:37
  Speaker 1: But I think at a high level, the question is for a new invention, is it something that could have just easily been generated by off the shelf AI using a well-known or trivial kind of a prompt or other type of an input?
• 35:55
  Speaker 1: Trivial kind of a prompt or other type of an input.
• 36:00
  Speaker 1: If so, I think not patentable, and you could say that that's going to actually raise the bar for patentability, as AI increasingly helps augment the inventive skill of people.
• 36:11
  Speaker 1: It's going to raise the bar for patentability, but it's going to do something else at the same time, which is it's going to make people more skilled inventors.
• 36:18
  Speaker 1: So it's going to be boosting people's skill, as it's also leveling the playing field, and it's unclear how those two things are going to play out with each other.
• 36:28
  Speaker 1: It's the same in all fields.
• 36:30
  Speaker 1: You know, outside of the patent field, right, whenever I hear people say well, ai is going to make writers obsolete because it's raising the bar and it's enabling almost anyone to become an average writer, yeah, but then skilled writers are going to use it and leverage it to become even more skilled.
• 36:44
  Speaker 1: Both of those things are happening at the same time as each other.
• 36:50
  Speaker 2: And you know we often talk about how the people.
• 36:55
  Speaker 2: You know people are concerned about job loss and at the moment it appears as though what I have to say is true and it might change over time but that people are unlikely to be replaced by AI alone, but they're very likely to be replaced by people who know how to use AI really well.
• 37:10
  Speaker 2: To your point One, this is perhaps a little bit off track, but I'm just curious because I think there is a relation to associations in the work they do.
• 37:17
  Speaker 2: But you know the US Patent and Trademark Office and similar agencies around the world.
• 37:22
  Speaker 2: Do you know, by chance, if they're embracing the use of AI to aid their work and thinking about the mountain of patent and trademark and copyright related issues or whatever else?
• 37:32
  Speaker 2: There are other kind of matters there they have coming through.
• 37:34
  Speaker 2: I know it's historically been a long backlog.
• 37:36
  Speaker 2: Even we do trademark work all the time and it takes months or years to get feedback.
• 37:40
  Speaker 2: Do you think there is an embrace of the technology to help them?
• 37:45
  Speaker 2: Because, to answer the question you asked, I could think, hey, let's take this patent application, take it to ChatGPT and ask if it was not obvious.
• 37:52
  Speaker 1: I know of at least two things, that the two ways in which the US Patent Office has been embracing AI.
• 37:58
  Speaker 1: The first is to classify patent applications.
• 38:02
  Speaker 1: So when a patent application comes in, they have to decide what technology is this for?
• 38:05
  Speaker 1: Is it for mousetrap technology?
• 38:07
  Speaker 1: Is it for antivirus software?
• 38:08
  Speaker 1: Is it a drug?
• 38:09
  Speaker 1: And then they assign it to the relevant patent examiner based on the classification, based on the classification.
• 38:14
  Speaker 1: About a year ago they announced they had developed, they had trained, their own model internally for doing patent application classifications.
• 38:22
  Speaker 1: I don't know if it's 100% automated.
• 38:24
  Speaker 1: There's probably some human review after the fact.
• 38:27
  Speaker 1: But that's one way in which they're using AI internally.
• 38:34
  Speaker 1: And then the tools for doing patent searching have been AI-infused increasingly.
• 38:37
  Speaker 1: They're getting better all the time, primarily for doing natural language searching.
• 38:43
  Speaker 1: You can imagine, traditionally when the patent office gets a patent application, it gets assigned to a patent examiner.
• 38:50
  Speaker 1: I'll use the mousetrap example again.
• 38:52
  Speaker 1: They have to decide is this a new type of mousetrap?
• 38:55
  Speaker 1: How do they do that?
• 38:56
  Speaker 1: They search for previous patents.
• 39:02
  Speaker 1: They see if they find any that are for the same mousetrap that they're examining now.
• 39:04
  Speaker 1: So to do that they have to come up with a set of queries in a database.
• 39:07
  Speaker 1: There's an intermediate step of designing the queries and, as we know, that takes a lot of skill when you're using a traditional database, particularly if you have to use a query language like SQL or some other god-awful language you know.
• 39:21
  Speaker 1: But now, increasingly, that middle step is being eliminated.
• 39:26
  Speaker 1: You can just take the patent application that's been submitted and use it as if it's the query right.
• 39:32
  Speaker 1: Right, maybe you do something else, but things are heading in that direction and it's eliminating the need to have this special querying or searching skill, or at least reducing the importance of that.
• 39:45
  Speaker 1: And I know the US Patent Office is using AI, sort of AI infused or, as you know, it's hard to define AI.
• 39:52
  Speaker 1: You could just say it's based on natural language processing or more natural language-based searching for prior art.
• 39:59
  Speaker 1: But I'll say this, and I may need to be tactful, but it is somewhat related to associations.
• 40:05
  Speaker 1: There's a big association within the patent office, which is the Patent Examiners Union, and often there is a tension, to say the least, between the union and the development and adoption of new technology, at least in part because of the reason you mentioned fear of job loss, a concern about that needing to be retrained, change how things are done.
• 40:27
  Speaker 1: So I don't know a lot about the details of how that's played out within the patent office, but I know that it's a perennial issue there and I'm not making an anti-union statement actually.
• 40:41
  Speaker 1: I'm just saying that that is a fact and sometimes that comes into play.
• 40:45
  Speaker 2: Well, I think there's natural tension amongst any number of people or groups that have misalignment essentially in terms of what they're naturally thinking about, and so I think there's opportunity for unions to be incredibly forward-looking with AI and help train their legions of workers in their fields on AI to be more competitive and more productive.
• 41:07
  Speaker 2: And there's obviously the other side of it as well.
• 41:09
  Speaker 2: So I think it's going to be interesting to see how that particular sub-segment of the world we live in and associations moves ahead.
• 41:17
  Speaker 2: I appreciate you going down that quick path because those processes, those two examples of AI augmentation within the patent office, are super relevant to associations.
• 41:27
  Speaker 2: They're both examples where they've taken existing processes.
• 41:30
  Speaker 2: They have not automated the processes, but they've augmented human labor with AI tools to eliminate perhaps some of the least interesting parts of the work.
• 41:39
  Speaker 2: That classification element takes a certain level of skill to do it correctly and to route it down the right path.
• 41:44
  Speaker 2: So it's not, you know, if it's in the backlog of someone's queue the software expert and it was a drug application or something for a drug patent and it got there.
• 41:52
  Speaker 2: Now it's got to go to the back of the queue for the drug person.
• 41:55
  Speaker 2: So I think that's a really important process got there.
• 41:56
  Speaker 2: Now it's got to go to the back of the queue for the drug person.
• 41:58
  Speaker 2: So I think that's a really important process, yet one that is obviously very relatively low-hanging fruit from an AI perspective.
• 42:02
  Speaker 2: Similarly, for associations.
• 42:03
  Speaker 2: Many associations deal with high volumes of applications for content where, if they run a scholarly journal or if they run a large conference, they have submissions of proposals for speaking, proposals for journal articles, etc.
• 42:16
  Speaker 2: And they have to go through a very long process and a very complex process.
• 42:20
  Speaker 2: So similar opportunities exist in this realm.
• 42:23
  Speaker 2: So, but thanks for going down that path.
• 42:25
  Speaker 1: Yeah, I mean I wonder if it raises another point that's common to my world and your world, which is that this demonstrates that the kind of fixed mindset you know is a mistake, that there's probably a lot of applications in the context you talked about, maybe that people don't submit because they know it's going to take so long, let's not even bother, and then, by speeding things up, you could expand the number.
• 42:55
  Speaker 1: There are many ways in which AI actually can create increased demand or supply, and having this fixed mindset that, oh, if we reduce the amount of labor required per task, that means there's going to be less total work to do, I think is a mistake.
• 43:14
  Speaker 1: It may be true in some cases, but it's not always true.
• 43:17
  Speaker 1: I mean, in the legal context, by using AI, I think, to provide legal services, it may create an increased demand for either new services or the ability for people to purchase existing services at a lower price that they wouldn't have even bought before.
• 43:32
  Speaker 1: There's all kinds of ways in which the demand is elastic.
• 43:36
  Speaker 1: In a sense is what I'm saying from an economic point of view that people sometimes overlook in their fixation on this oh, the AI is going to take away work and thereby decrease demand for human labor.
• 43:49
  Speaker 2: That's a great segue into something we wanted to talk about, in that you know when we think about patent and patent both filing patent and eventually potentially protecting.
• 43:56
  Speaker 2: You know when we think about patent and patent both filing patent and eventually potentially protecting.
• 43:59
  Speaker 2: You know you typically think of large budgets.
• 44:03
  Speaker 2: You know it's expensive to get patents and you talk about your world of doing this work for organizations of various sizes.
• 44:10
  Speaker 2: You know.
• 44:11
  Speaker 2: So this, all this generative AI, is likely to make, you know, at least certain portions of your workflow, simpler, easier, faster, more scalable.
• 44:19
  Speaker 2: And so you know, if there was a fixed number of patents that could be created in the world, then of course you know you might have an existential threat to your fundamental business.
• 44:31
  Speaker 2: But if the theory is, you know, upheld and has so far throughout recorded economic history has been true, the demand is essentially insati, has so far throughout recorded economic history has been true.
• 44:37
  Speaker 2: The demand is essentially insatiable and correlates to lower cost, driving increased demand when there's a proper service.
• 44:44
  Speaker 2: That's novel, right.
• 44:45
  Speaker 2: So you have this happening over and over throughout time as technology disruptions have occurred.
• 44:49
  Speaker 2: You're going to have more companies essentially coming to you, more associations maybe coming to you saying, hey, robert, can you help me with a patent.
• 44:56
  Speaker 2: It'll cost now maybe a third of what it might have cost 10 years ago or something like that, and maybe there's more opportunity to get patents.
• 45:04
  Speaker 2: So have you seen any evidence of that in your own practice or in other practices kind of adjacent to yours?
• 45:10
  Speaker 1: Yeah, I mean, I'm just starting to see it.
• 45:12
  Speaker 1: It's really at the very early stages, and I would say that we are already working on modifying how we perform services, offering new types of services.
• 45:25
  Speaker 1: We're exploring all of the different ways in which AI can help us to evolve how we provide, make them more readily available.
• 45:33
  Speaker 1: It's early days, but I just want to.
• 45:36
  Speaker 1: I want to.
• 45:37
  Speaker 1: I always like using a couple of examples of what you talked about when you say through economic history and how easy it is to fall into this trap of thinking that there's fixed demand.
• 45:47
  Speaker 1: You know, bill Gates very famously said it might've been in the 80s, no one would ever.
• 45:53
  Speaker 1: Why would anyone ever need a computer with more than 640K of memory?
• 45:57
  Speaker 1: And so for people out there, I mean, that is what a?
• 46:02
  Speaker 1: Is it a millionth of what most computers have now?
• 46:05
  Speaker 1: It's a tiny, tiny fraction, right, and so it's.
• 46:10
  Speaker 1: And he's a smart guy, you know.
• 46:12
  Speaker 1: And yet he somehow fell into this trap of thinking how could we possibly expand?
• 46:16
  Speaker 1: I mean, when I was growing up, we had a handful of television networks and then cable came along and you had what?
• 46:23
  Speaker 1: 100 channels, and I think many people thought why would anyone ever need more than that.
• 46:27
  Speaker 1: Well, all of us now have the equivalent of what millions of channels through the internet television and there's no sign of the demand.
• 46:37
  Speaker 1: As you said, it's insatiable, it's insatiable.
• 46:39
  Speaker 1: So when people now are looking at AI and saying, oh no, this is going to eliminate the need for people, I'll just put it in context of invention and patents, which is that this is my belief.
• 46:51
  Speaker 1: It is backed up by history.
• 46:52
  Speaker 1: I can't prove it going forward that there is no limit to the number of new problems we will find to solve and tackle as we solve old ones.
• 47:03
  Speaker 1: So even if AI gets really good at solving our existing level of problems, people are really not just ingenious but motivated to try to tackle the next higher level problem, and then they'll sit on the back of AI to help boost them up to do that, and I don't think there's ever going to be an end to that.
• 47:23
  Speaker 2: To build on your Bill Gates story, in addition to his famous quote of 64K, should be enough for anyone, or something along those lines, which is kind of crazy to look back on.
• 47:32
  Speaker 2: He's also, I believe, out of the between him and Paul Allen.
• 47:37
  Speaker 2: He was the one who said our BHAG is going to be a computer on every desk and in every home, and that BHAG seemed totally outlandish at the time.
• 47:44
  Speaker 2: So the reason I bring that up is because you can both be extraordinarily visionary and have blinders on at the same time, and we all have those issues.
• 47:52
  Speaker 2: We all have these biases that lead us down these paths, and that's one of the things I think AI can help us do is to tell us if we're missing something right, to give us feedback on the work that we do.
• 48:03
  Speaker 2: That's one of my favorite use cases is work that I've done.
• 48:05
  Speaker 2: I'll feed it to the AI and say criticize this, give me a critique of this piece of software or this writing, or whatever the case may be.
• 48:14
  Speaker 2: One of the other things I wanted to quickly mention is in the talk that I give on AI, I start off with economic history and over the last thousand plus years, when we look at the growth of global GDP on an inflation adjusted basis.
• 48:26
  Speaker 2: It took basically all of recorded human history to get to through 1700 or so to get to 1 trillion in global GDP.
• 48:34
  Speaker 2: And then it took about 250 years, from 1700, roughly to 1950 ish to get from 1 trillion to 10 trillion in global GDP.
• 48:42
  Speaker 2: So you know basically infinity time of our species then compared to 250 years for the next 10x increase.
• 48:48
  Speaker 2: But what's even more striking, that is from 1950 through now, which is, you know, 65 plus 70 plus years, we went from 10 trillion to 100 plus trillion in global GDP, again on an inflation adjusted basis.
• 49:00
  Speaker 2: So the next 10x order order of magnitude increase happened in 65 years compared to 250.
• 49:05
  Speaker 2: So what's it going to take for the next 10x increase?
• 49:08
  Speaker 2: Right, and that's highly speculative to say.
• 49:09
  Speaker 2: As you just pointed out, we can't write the future the way we can look at the past, but there's a pretty good trend line there and, with AI being an accelerator for creativity and opportunity, demand is likely to drive that, because that's a story of demand.
• 49:22
  Speaker 2: It's not a story of anything else.
• 49:23
  Speaker 2: It's showing that 130 plus trillion global GDP, compared to 1 trillion not that many years ago, is a story of demand.
• 49:33
  Speaker 2: Being insatiable, really saying the same thing you just said.
• 49:35
  Speaker 1: Yeah, I mean I would think that for a lot of associations you tell me, particularly nonprofits, one of the things that they struggle with is there's people out there who could benefit from their services, who can't get access to it right For geographic reasons, for economic reasons, for all kinds of reasons, and so you know anything you can do to expand the ability to serve those people and legal services.
• 50:00
  Speaker 1: You know it's usually talked about in the context of people not being able to get their basic legal needs met for criminal defense or domestic relations or getting a will drawn up or anything like that.
• 50:13
  Speaker 1: But there's so many legal needs where people can't access them because it's out of their mostly economic reach.
• 50:21
  Speaker 1: In some cases there's not a lawyer locally who they can access.
• 50:25
  Speaker 1: Obviously, the internet has helped a lot with that, but there's so many opportunities to be able to expand the reach of services in law and, I would assume, for a lot of associations as well.
• 50:37
  Speaker 2: A hundred percent.
• 50:37
  Speaker 2: I think you've actually identified one of the core opportunity areas for associations to grow, which is to do a better job of connecting supply and demand in their verticals.
• 50:47
  Speaker 2: So if you take the legal profession, you say, hey, think about the bar associations that exist all over the country and all over the world, whether they're specialty around you know, practice area or their geographic base.
• 50:57
  Speaker 2: One of the functions they serve is a lawyer referral service, which exists also in various medical domains, and these things are very, very traditional domains and these things are very, very traditional.
• 51:09
  Speaker 2: You know you sometimes there you call up a phone number and someone from xyz bar association answers, talks to you, asks you about the matter in hand and tries to connect you with the right member.
• 51:16
  Speaker 2: Or you have an online version of that which is very much like click from a series of drop downs.
• 51:21
  Speaker 2: You kind of have to know the topics and then you may or may not find someone.
• 51:24
  Speaker 2: It's kind of like a google search where you know if I'm in Boston, I'm looking for an IP attorney, I might find 600 people right, or whatever the number is, and you and I, robert, met 20 years ago or whatever it was, because a mutual friend of ours had worked with you and he said hey, you got to talk to Robert about some software patent I was interested in and that was just by happenstance and you helped me with some software patents at my prior company and that's how we got to know each other.
• 51:47
  Speaker 2: But you know, that's very random in a way.
• 51:50
  Speaker 2: Right, it's this network and can we take that juice right that made that connection valuable and scale that right.
• 51:57
  Speaker 2: And that's an opportunity for associations because they're at that nexus of both sides of that micro economy.
• 52:04
  Speaker 2: So there's an opportunity there to create an AI powered, you know, professional networking resource that goes way, way deeper than like, oh, robert's been in practice for 25 years and has a CS degree and blah, blah, blah, blah, blah.
• 52:15
  Speaker 2: That's cool.
• 52:16
  Speaker 2: But what's his personality like?
• 52:18
  Speaker 2: What kinds of cases is he best at working at?
• 52:20
  Speaker 2: You know, who does he like to work with?
• 52:21
  Speaker 2: Right, and how does that fit with the potential clients?
• 52:24
  Speaker 2: And that choke point exists in a lot of associations I know in their various domains.
• 52:29
  Speaker 2: So that's just one opportunity category and there's so many like that.
• 52:33
  Speaker 2: Biggest associations can act as this superhero intermediary, if you will, to bring the quality of service higher and to lower the cost of getting to these people, because that transaction cost is a big reason why people actually have so much overhead in their businesses, because they have to spend so much time with the wrong kind of intake, and associations stand to benefit from that themselves if they do a good job of it.
• 52:56
  Speaker 3: Robert, I know you recently wrote a book.
• 52:58
  Speaker 3: I think at the time of this recording it's available for preorder, but perhaps by the release of this episode it will be fully available.
• 53:03
  Speaker 3: It's called AI Armor.
• 53:05
  Speaker 3: Can you share kind of a high-level overview of that book and maybe some insights that might be most relevant for our nonprofit and association listeners?
• 53:14
  Speaker 1: Yeah, the book is a strategy guide to using intellectual property to protect AI innovations, so the audience is primarily companies and other organizations that are developing new types of AI technology the types of companies that would be my clients who want to obtain patent or trade secret protection.
• 53:37
  Speaker 1: Leverage that to raise funds to grow their businesses to secure a successful exit.
• 53:44
  Speaker 1: One thing I'll say is that one of the motivations for the book was to address and bust some myths about patents and intellectual property, which often cause innovative companies to not pursue intellectual property protection because they hold these myths.
• 54:04
  Speaker 1: One of them is that you only obtain patents if you're planning to use them to sue your competitors, to stop them from competing with you.
• 54:11
  Speaker 1: And I deal with a lot of companies and they say, oh, we don't need a patent, we're not planning to sue anybody, we don't want to sue our competitors.
• 54:18
  Speaker 1: And I say, well, 99% of my clients never do that either.
• 54:23
  Speaker 1: Most of what patents are used for is to one signal to potential investors or partners, or to the public, or to your customers that you're an innovative, forward-thinking company, meaning when an investor says before I pump some money into you, what kind of reasonable assurance not a guarantee, but what reasonable assurance do I have that another company, particularly some tech giant of the world, is not going to see what you're doing, say that looks good, let's copy that, put it into our product and run you out of business overnight.
• 55:04
  Speaker 1: Now there's a lot of different modes.
• 55:05
  Speaker 1: Intellectual property is one of them.
• 55:07
  Speaker 1: There are others.
• 55:09
  Speaker 1: There's first mover advantage, there are network effects.
• 55:14
  Speaker 1: There's all kinds of economic mechanisms that can be moats.
• 55:18
  Speaker 1: We talked about data before.
• 55:21
  Speaker 1: Your trade secret, proprietary data, your customer list, your brand those can all be moats.
• 55:27
  Speaker 1: But patents serve that really valuable purpose.
• 55:30
  Speaker 1: They also are assets.
• 55:32
  Speaker 1: They can be sold.
• 55:33
  Speaker 1: They can be licensed fairly freely compared to things like trademarks, which you can only assign when you sell the business for the most part.
• 55:41
  Speaker 1: But I have quite a few clients who have obtained patents and then licensed them to other companies, meaning charge a fee for the ability to use that technology and then they also become really valuable in valuation of a company, particularly when you're looking to get acquired, go public or have some other form of exit.
• 56:02
  Speaker 1: So that myth that intellectual property is only valuable if you're going to use it to sue and if you're not looking to sue you don't need it or wouldn't benefit from intellectual property is one of the motivations I had for that book, but I think the way it would relate you said you know what would your listeners get out of it I think it's the higher level concept of a moat, of a moat, even though I think the moat I'm focused on in the book is primarily relevant to tech companies that are developing new forms of technology.
• 56:33
  Speaker 1: The concept is applicable and I mean I say here's where my book and your book are aligned.
• 56:38
  Speaker 1: It's relevant to any business.
• 56:40
  Speaker 1: Any business, really any individual has to ask what is my you could call it it goes by so many names unique value proposition as an individual, maybe, or what gives me my competitive advantage?
• 56:55
  Speaker 1: What is it that was going to draw customers or members or funders to me rather than my competitors?
• 57:02
  Speaker 1: Why am I not a commodity?
• 57:04
  Speaker 1: And if I'm not a commodity now, why won't I be a commodity in a few years as a result of AI right?
• 57:12
  Speaker 1: How can I maintain that moat, that distinctive competitive advantage, over an extended, sustainable period of time?
• 57:20
  Speaker 1: I think that's where what I write about and what, amit, you write and talk about and work about overlap with each other.
• 57:29
  Speaker 2: I agree completely and I think the idea of IP generally as an opportunity to build and strengthen a moat is a really critical concept, the intersection of our content areas.
• 57:42
  Speaker 2: And when I think about associations and their opportunities, looking ahead, I think about the types of services that their fields will ask of them or will ask of the world, the things, the unmet potential demand that's out there of just providing once a year conferences or multiple times per year conferences or fairly static educational courses.
• 58:10
  Speaker 2: There's a need for a much more dynamic, interactive, almost co-pilot, if you will, for that sector where the association has an opportunity to potentially publish a co-pilot for that field, an assistant for that field.
• 58:23
  Speaker 2: And they have the brand, they have trust and credibility in the space.
• 58:27
  Speaker 2: Oftentimes in that particular discipline they have deep content, deeper content than most others would have.
• 58:33
  Speaker 2: But the combination of the brand integrity they have in the field, coupled with expertise, could be very interesting.
• 58:42
  Speaker 2: Hundred smartest members you have and their collective make, you know brains being merged together and then being available as an ai assistant for your entire field.
• 58:52
  Speaker 2: That elevates the entire profession, whatever that profession is.
• 58:56
  Speaker 2: And so my question this is a very common discussion, right?
• 58:58
  Speaker 2: This is a common idea that people in the association market are thinking about.
• 59:02
  Speaker 2: There's the customer service side of it, which I find much.
• 59:05
  Speaker 2: It's.
• 59:05
  Speaker 2: It's interesting in terms of the value creates, lowers friction, etc.
• 59:08
  Speaker 2: But it's interesting in terms of the value creates, lowers friction, etc.
• 59:08
  Speaker 2: But it's much less interesting than the professional knowledge assistant that I'm referring to.
• 59:12
  Speaker 2: When you hear about that idea, do you think about potential patent protection or other forms of IP protection that could help an association establish a moat if they built that kind of a co-pilot or assistant around their professional expertise?
• 59:30
  Speaker 1: There could be something patentable in there, and I have been working with some clients on it.
• 59:35
  Speaker 1: I say could be because if the method you use to produce the co-pilot is generic, what I say you've got some core let's say chat, gpt and you use a process where you gather up your unique knowledge and you build the product, the co-pilot, in a conventional way, then that's probably not going to be patentable.
• 1:00:02
  Speaker 1: But the fact that you're using your unique knowledge, wisdom, data, other things to produce a co-pilot that is actually more useful than something else, you will have a commercial competitive advantage and you're going to most likely want to use trade secret protection, maybe some copyright protection on the input side.
• 1:00:28
  Speaker 1: And then you know, I don't know how associations are dealing with this, but in the private commercial context right now, there's a lot of talk and debate and experimentation with what happens on the output side.
• 1:00:42
  Speaker 1: Right, obviously, these systems evolve, they're living and they develop.
• 1:00:46
  Speaker 1: They're not static, as you've said already.
• 1:00:48
  Speaker 1: So once the co-pilot produces data, I mean very often you want to retrain based on that or update based on that.
• 1:00:57
  Speaker 1: And then there's going to be some question who owns the output?
• 1:01:00
  Speaker 1: Who has rights to use it?
• 1:01:02
  Speaker 1: Is it licensed?
• 1:01:03
  Speaker 1: Is that maintained as a trade secret.
• 1:01:05
  Speaker 1: Somehow, and very often that's going to be governed as much by contract, you know, an end user license agreement or something else as it's going to be by core intellectual property like copyright or trade secret.
• 1:01:18
  Speaker 2: Let me pare it back what I think I heard just to make sure I've got this right.
• 1:01:22
  Speaker 2: The example I gave patents are probably not the route that would make sense there, because there isn't anything in that mix that the association has created.
• 1:01:32
  Speaker 2: The underlying AI might be interesting, but they typically wouldn't have built their own model.
• 1:01:37
  Speaker 2: They would have leveraged something out there and then fed their content to it through trade secrets protection, possibly copyright.
• 1:01:46
  Speaker 2: That could indeed make something an asset of that nature, indeed a strategic mode, even though there's not necessarily patent protection available for that category of innovation.
• 1:01:57
  Speaker 1: Yeah, you said it much more succinctly than I did.
• 1:01:59
  Speaker 2: Thank you, and then the other related question is we have some clients we're working with who are associations and, to extend on the concept that is described, they actually want to build some additional capability on top of a base co-pilot, where they want to say, hey, listen, in my field I want my co-pilot to help clients their clients right now, the associations and customers with building new products in their field.
• 1:02:24
  Speaker 2: They might have this expertise around how to formulate a new product, a new chemical right Going back to the drug example or a new consumer good of some sort.
• 1:02:33
  Speaker 2: And these co-pilots go beyond the idea of simply knowledge bots that can talk to you, chat, gpt style but they can actually help you create something.
• 1:02:41
  Speaker 2: I thought that was an interesting thing to bring up in this podcast, because there's an IP question around that product itself.
• 1:02:47
  Speaker 2: That's like let's call it a super co pilot.
• 1:02:49
  Speaker 2: But then there's also the question of who owns the IP it generates and obviously by license you can determine if there's an assignment or whatever.
• 1:02:56
  Speaker 2: But what are your thoughts on that scenario?
• 1:02:59
  Speaker 1: Yeah, it's very interesting.
• 1:03:00
  Speaker 1: Going back to the book I wrote in 2009, the Genie and the Machine, that co-pilot would be an example of what I call the genie.
• 1:03:07
  Speaker 1: And then I explored do you patent the genie?
• 1:03:09
  Speaker 1: Do you keep the genie as a trade secret?
• 1:03:11
  Speaker 1: To use a different metaphor, I just published an article on IP Watchdog about evaluating patent versus trade secret for AI, which I use the analogy, if you remember the old Aesop's fable of the goose that laid the golden egg, which is the analogy there is.
• 1:03:27
  Speaker 1: If you have this co-pilot that can design new things or create new content or potentially create new patentable inventions.
• 1:03:34
  Speaker 1: One option is you keep that co-pilot secret and hidden.
• 1:03:38
  Speaker 1: That's the goose laying the golden eggs, right, and then it produces the golden eggs, the new inventions, the new useful things.
• 1:03:46
  Speaker 1: Maybe you sell those publicly, maybe you patent those, maybe you license them who knows?
• 1:03:52
  Speaker 1: But you might want to do that while you're keeping this goose laying golden eggs for a long time in secret without revealing to anyone else how it works.
• 1:04:03
  Speaker 1: There's a bunch of trade-offs involved because you might want to patent that as well.
• 1:04:07
  Speaker 1: One reason you might want to patent is if you think someone else might figure out how to create that same copilot in the relatively near future.
• 1:04:15
  Speaker 1: Maybe it's better to patent it, because a patent will give you protection against someone else who invents after you filed your patent application, whereas a trade secret won't.
• 1:04:24
  Speaker 1: So if you keep that co-pilot as a trade secret, you're sort of making a bet that the idea you came up with is something other people are not going to figure out in the near future, and it's hard to predict the future.
• 1:04:40
  Speaker 1: But you make your bets and you do your best.
• 1:04:45
  Speaker 1: But there's a bunch of these trade-offs in terms of patent versus trade secret protection, even including whether you you know you don't necessarily you may not have to make the output public.
• 1:04:57
  Speaker 1: I mean it depends.
• 1:04:58
  Speaker 1: You might privately license use of this co-pilot to parties who are going to agree to keep the output secret, as well as a condition of using the co, the copilot, in the first place.
• 1:05:12
  Speaker 2: There's all different kinds of permutations here.
• 1:05:15
  Speaker 2: Yeah, that's a really interesting angle on it and I think that for associations thinking through this, I want the listeners here to first of all recognize they have a valuable asset somewhere in there and that there are opportunities to create monetization schemes around this in various ways.
• 1:05:33
  Speaker 2: And there's ways to protect that asset in various ways.
• 1:05:35
  Speaker 2: But what isn't protectable ultimately would be a scenario where you do nothing, and I'm saying that from a business perspective.
• 1:05:43
  Speaker 2: So if you simply have your content sitting there and you're saying, hey, we've got a great brand and we've got great content, and someone else has OK brand and OK content, but they do this and they make their content supercharged with super co-pilot, even though your content is the excellent content and the next guy's content might be the very good content put into a novel format with better intelligence around it is likely going to run circles around what the better content sitting still will do.
• 1:06:14
  Speaker 2: You know, one thing that I wanted to extend in this conversation is when associations are thinking about their approach to IP protection, they're also thinking about the flip side of it, which is, how do they defend against other people taking advantage of them, taking advantage of their content or their assets?
• 1:06:33
  Speaker 2: What are your thoughts on.
• 1:06:34
  Speaker 2: If you're an association, you have years of journal articles and other things.
• 1:06:39
  Speaker 2: What's your point of view in terms of how you create like a defensive strategy against potential misuse in the age of AI?
• 1:06:48
  Speaker 1: Well, you know, this is a very active question right now in the copyright office in the courts.
• 1:06:55
  Speaker 1: It's really evolving very rapidly.
• 1:06:57
  Speaker 1: Probably you and your listeners are aware of you know, the big New York Times lawsuit against OpenAI, this question of well, a lot of the big models out there were created by being trained on publicly available data on the web.
• 1:07:15
  Speaker 1: They were made possible by the fact that the web exists and there is freely available content out on it lots of it and the fact that these models have now been created and are being used on their own, independently of the content that was used to right I mean newspapers grappled with it in the beginning.
• 1:07:36
  Speaker 1: Should we be putting our stuff up for free?
• 1:07:38
  Speaker 1: A lot of them, I think, answered that the wrong way for themselves, but it was a legitimate question.
• 1:07:57
  Speaker 1: They had a business model that worked for them for a long time decades at least, if not a couple of centuries and then they were up against a totally new technology and set of business models which relied on giving away or making available lots and lots of stuff for free.
• 1:08:13
  Speaker 1: And ever since the early web, journalists and content creators and content creation and owning organizations have been grappling with it.
• 1:08:23
  Speaker 1: But AI models take it to the next level because then, once the model exists, it potentially can be used and be valuable I don't want to say more valuable than the original content but the nature of the technological and business and economic relationship between the content creators and owners and the AI companies and the models they create that are trained based on them, is very much in flux right now.
• 1:08:51
  Speaker 1: I'll say that to say the least.
• 1:08:52
  Speaker 1: But I guess, going back to your question, it raises the question of I think you need to ask yourself, for every piece of content you create, should we make it available publicly for free?
• 1:09:04
  Speaker 1: More so than people have been asking that question, I think, for the last 30 years, yeah, that's a great point.
• 1:09:12
  Speaker 2: I think people need to put some thought process into it, and associations are famous for creating tremendous amounts of governance and structure around every decision.
• 1:09:20
  Speaker 2: So I think it's important to be nimble with this, because you have to have a volume of content constantly going through, so you can't, you know, run everything by a committee necessarily.
• 1:09:28
  Speaker 2: But I think some kind of a very light and nimble framework for conceptualizing what to consider, publishing, what not to, could be very, very helpful.
• 1:09:38
  Speaker 1: Something along those lines, I think, would be a great asset for associations to be, to build and to take advantage of, I mean there are people who think that even if you have an organization that's got really great proprietary data, that an AI system will sort of generate the equivalent of it in the near future or that someone else will generate that, and so you can't really maintain a competitive advantage based on your data.
• 1:10:06
  Speaker 1: I personally disagree with that.
• 1:10:07
  Speaker 1: I can't really maintain a competitive advantage based on your data.
• 1:10:09
  Speaker 1: I personally disagree with that.
• 1:10:09
  Speaker 1: I can't prove it, but I suspect and this is just a suspicion that might also be wrong.
• 1:10:14
  Speaker 1: You know, and I think this is where you're heading, but tell me, if I'm misinterpreting you that if you are an organization and part of your value is this unique, really valuable data that you have, that if you marry that with AI and think carefully about how you do that, both from a technological and business model perspective, you could then keep building on and developing and leveraging that original core that you had to maintain and increase your relevance and value over time.
• 1:10:44
  Speaker 1: Is that fair?
• 1:10:45
  Speaker 1: That's exactly right.
• 1:10:46
  Speaker 1: I mean, that's what we're proposing most associations.
• 1:10:49
  Speaker 2: Consider, because you know they're essentially associations act in.
• 1:10:52
  Speaker 2: You know there's several capacities, but one of the which is knowledge center for their domain, so being kind of the expert resource in a given field, the connector connecting people to people, and also people within the profession, the people who want access to professionals in the space.
• 1:11:07
  Speaker 2: And also there's a social element, but you social element, but those core functions, I think, are amplified by AI.
• 1:11:14
  Speaker 2: They can do them at scale better than ever before, as opposed to looking at it from the viewpoint of a limited pie or that fixed mindset you mentioned earlier.
• 1:11:22
  Speaker 2: So that's the opportunity that's out there.
• 1:11:25
  Speaker 2: The other thing is the association on a daily basis goes from being really administrative in nature, where they're processing membership transactions or event registrations, to actually being relevant in the eye of the profession on a daily basis.
• 1:11:38
  Speaker 2: Most people who deal with attorneys deal with their bar association to make sure they have enough CLEs to stay certified for the next year or whatever.
• 1:11:47
  Speaker 2: They're not really thinking about the bar Association as part of their daily practice.
• 1:11:50
  Speaker 2: Nor is the medical professional in a particular specialty area of medicine.
• 1:11:54
  Speaker 2: But if the association could leverage and harness that technology, coupled with the content they uniquely own, that could be a reason, a legitimate reason from the perspective of the audience, to engage with them more than episodically.
• 1:12:07
  Speaker 2: So I find that exciting and I think many associations have a significant opportunity in front of them.
• 1:12:12
  Speaker 2: The alternative is that people go to the good enough content.
• 1:12:15
  Speaker 2: What I mean by that is Google Gemini or ChatGPT or whatever else is contemporary in terms of the way you engage with it and not quite as good as what the association should be able to create, but good enough and increasingly better.
• 1:12:28
  Speaker 2: Right, you know CHATGPT?
• 1:12:30
  Speaker 2: I know there's a famous example last year sometime where an attorney composed some kind of brief or something in cited cases that didn't exist.
• 1:12:38
  Speaker 2: But of course, that person I don't think actually checked any of the work and that was an early version of CHATGPT.
• 1:12:43
  Speaker 2: But it's kind of a beautiful example of what could go wrong if you don't have reasonable discipline in using these tools.
• 1:12:50
  Speaker 2: I always tell people imagine you just got a really, really smart college grad, brilliant person, works hard, never rests, and they're amazing, they're encyclopedia kind of knowledge, yet they make mistakes.
• 1:13:03
  Speaker 2: And so would you take that person's output and just publish it directly on your website or would you actually take a look at it?
• 1:13:08
  Speaker 2: I'd recommend you take a look at it, just like you do with AI.
• 1:13:10
  Speaker 2: So anyway, I digress.
• 1:13:12
  Speaker 2: But the point is is I think there's this massive opportunity for associations to do exactly what you described.
• 1:13:18
  Speaker 1: Yeah, I would love, I would be drawn more to the bar associations if they did exactly what you just talked about.
• 1:13:26
  Speaker 2: Yeah.
• 1:13:27
  Speaker 3: I've got to insert a plug here.
• 1:13:29
  Speaker 3: If all our listeners are enjoying this conversation around what it means to stay relevant and thrive as an association, what it means to be exponential, we're having these exact conversations at Digital Now this year, which is October 27th through 30th in DC at the Omni Shore Hotel, and we'll be talking about this conversation essentially and leveraging your competitive moat.
• 1:13:47
  Speaker 3: So you can get more info on that at digital now conferencecom.
• 1:13:51
  Speaker 3: Robert, thank you so much for an incredible conversation today.
• 1:13:55
  Speaker 3: I know I have learned a ton, amit.
• 1:13:58
  Speaker 3: I feel like you've probably learned a ton as well, and I'm sure this will be really insightful for our listeners.
• 1:14:01
  Speaker 3: So thank you so much, and can you let everyone know where they can keep up with you?
• 1:14:06
  Speaker 1: Yeah, absolutely, the feeling is mutual.
• 1:14:08
  Speaker 1: I learned a lot from both of you today.
• 1:14:10
  Speaker 1: You can catch up with me on LinkedIn.
• 1:14:13
  Speaker 1: I post there all the time, linkedincom, slash in slash, robert Plotkin, follow me there, message me there.
• 1:14:20
  Speaker 1: And also on our website, blueshiftipcom, you can learn about how to use IP for your growing tech company or if you want to contact us.
• 1:14:30
  Speaker 1: If you might be in need of a patent or intellectual property services, you can contact us, schedule initial consultation with us.
• 1:14:36
  Speaker 1: Blueshippedipcom.
• 1:14:38
  Speaker 3: Thank you.
• 1:14:42
  Speaker 2: Thanks for tuning into Sidecar Sync this week.
• 1:14:44
  Speaker 2: Looking to dive deeper?
• 1:14:45
  Speaker 2: Download your free copy of our new book Ascend Unlocking the Power of AI for Associations at ascendbookorg.
• 1:14:53
  Speaker 2: It's packed with insights to power your association's journey with AI.
• 1:14:57
  Speaker 2: And remember, sidecar is here with more resources, from webinars to boot camps, to help you.
• 1:15:03
  Speaker 2: Stay ahead in the association world.
• 1:15:04
  Speaker 2: We'll catch you in the next episode.
• 1:15:06
  Speaker 2: Stay ahead in the association world We'll catch you in the next episode.
• 1:15:10
  Speaker 2: Until then, keep learning, keep growing and keep disrupting.