AI or Die: The Future is Now

Being a great IP attorney isn't just about knowing the law. It's a combination of your intelligence, your capability, and your capability is really your ability to apply your intelligence to whatever the problem solving is. And then of course, there's this third item that are the intangibles, right? So what makes us good, what makes us great is intelligence, plus capability, plus these intangibles such as our ability to have relationships, the way we communicate, these types of things. And your intelligence, if you're an IP attorney, really includes your domain expertise. How good of an expert are you at whatever niche of IP you specialize in? And then how good Can you apply that intelligence to analyze and solve problems? Your capability, as opposed to intelligence, your capability, it includes your skills, your work ethic, how hard do you work, your workflow, what is the process by which you get things done, your strategic thinking, your efficiency. And the best IP attorneys don't just rely on their intelligence, but they continuously And refine their approach and improve their processes to solve problems better. And then there's this third component, which are the intangibles. And the intangibles are really just who we are as people, the kind of special gifts that we each bring to the table that are unique to us, that translate into, for instance, strong relationships. So your intelligence, your capabilities, and your intangibles, These things often determine what level of professional success you can achieve. And they not only define our potential, our professional potential, but we are limited by some of the things that we're not good at, by the ability to learn. We don't know everything about intellectual property law. We don't know the case that came down yesterday. We can't remember everything. So we have this innate potential. We also have limitations to our potential. And before artificial intelligence, we all had a really good idea of kind of what our maximum potential was as an IP professional. We know that on our best day, how good could we be? And of course, tomorrow we might learn something new and become marginally better, but there's a cap, kind of like Magic Johnson in the, what was it, 19— '90 championship game, game 6, Kareem is down, their center is lost, the best player in the league is out, and Magic Johnson steps up and says, I'll play center. And he dominates the game, 42 points, 17 rebounds, and I think 8 assists. And he literally wins the game and he's a rookie. So he reached his potential, you can argue, In that game and may never have actually been able to exceed that over the course of his career. He reached his flow apex. He was in flow. And we've all felt that on certain days where just everything's clicking. Now, in the world of AI, these things have now changed. So I want you to imagine a world where you as an IP professional can be 2 times better than you've ever been, than you ever could be before, or 5 times better or more than you've ever been before, or 20 times more than you've ever been before. And this is not only the promise of AI, but this is the reality of AI today. The current capabilities of the present systems can make each of us 2, 3, 5, 10 times better than we otherwise could be. And so this dream isn't hype, and this amplification of your abilities using AI technology is happening all around you right now. And my purpose here today is not only to convince you that the future is now, but that you need to act today, that now is the moment in time. to make this move. Now, the title of this used to be AI or Die, the Future is Now, and that's the presentation I gave a year ago. I'm not so keen on the or die part because number one, obviously we all know from headlines that actually could happen, that AI could be the extinction of mankind. I don't worry about that. Who knows what's going to happen next? No one knows. I don't know. All I know is what is available today. And I know that I am going to take advantage of what's available today to explore the outer limits of my own human and professional potential. So let's take a look at this thing and let's think about the real world, right? And you're probably wondering who I am and who I am to be standing out here in front of you today. And this, I hate this part because this is the part where I'm supposed to tell you about myself and that I've been practicing for 30 years as an internet technology attorney, and that I've always been into technology and how it can impact the business of law, how it can change the way law is practiced. And I've been doing this since really 2004, innovating on law using technology. But that is definitely not the reason why you should listen to me today. Here's why I think you should take the things that we're going to discuss today seriously and follow up on them tomorrow. Not next week, not next month, tomorrow. I use AI every day, all day long. And on average, I interact with AI over 200 times a day. I am a power user of AI and I've developed skills and strategies and the right mindset to get more out of AI than the vast majority of people in the world. And so I'm living this. This isn't hypothetical of what this can do. I live it every single day and I've seen the impact it's had on my work product and my client relationships and the rest. So this is all real-world stuff. And over the last 2 years, I've really deepened my AI skills and abilities. And I've increased my flow apex, my professional potential by 2, 5 times, 10 times, depending on the task. I've amplified myself in ways that simply would never have been possible without AI. So I've also been training lawyers these last 2 years to integrate AI into their practice. And so lawyers who have an interest in it, they contact me and they ask me, what should I do next? Here's where I am in the process. How do I build this? How do I integrate AI into my workflow? And so I understand the challenges and the limitations and the approaches that various legal professionals are using to try and get up the ladder. And I'm not going to sit here and tell you that this is easy. I'm not going to sit here and tell you that it is going to be a cakewalk where you go to a prompt and you ask, question and it gives you the answers. That's not AI. That's not how AI works. And it is going to take effort and commitment. And I hope by the end of today, it's an effort and commitment that you actually want to undertake. So there's a couple different ways to look at this. And one of which is the AI imperative, why you must adapt now, right? This is the AI or die part, right? This is the argument As to why it is, it's really not a choice that you have to do this. And if you're a law firm, here are some of the arguments that you'll hear if you're in the flow of information about AI and especially with regards to law firms and other business entities and business professionals. So the first argument that people make is that your employees are already using AI without supervision. So if you're a law firm contemplating AI, Now is a critical time because you never know if your confidential proprietary customer information is going into some model that's going to get spit back out to the world and potentially to a competitor. So that's one argument that you'll hear. The other is that the AI revolution is unprecedented and picking up speed. Those are the headlines that you see every day. They're all over. There's never been something that penetrated the news cycle as much as AI has. And continues to do. The other arguments are also, I think, important to at least account for, which is the ability to catch up to early adopters will be limited. Let me say that again. Your ability to catch up if you wait may not exist. And here's why. When we've had these new technologies come before, such as cloud, case management where you could actually upload a document to the cloud and then access it on your cell phone or at your home computer. That provided us efficiencies and access and the ability to work from anywhere. There's no question that amplified us to some degree. Now, if you are Jane and you adopted early and you started to use cloud as a power user, you were able to get up to this level, right? And if you were Joe and you waited 2 years, yeah, you were behind, but you also got up to Jane's level because there was a cap on the types of efficiencies you're going to get out of that. There is no cap with AI. You're interacting with intelligence, intelligence which is already objectively, measurably as smart as the smartest of us and is exponentially getting smarter. Every month that goes by. So the reality of that is there's no limit on how good you can be potentially. Right now we have a limit because AI can only do so much. It's only so smart, but that keeps changing. Not only that, but the skills you developed and the strategies that you develop and the mindset that you develop, all of those are never-ending. So you, even if AI never changed, you could always get better and get better output. If you build the skills, you build the mindset, You build a strategy that's going to get you further up the ladder. So if Jane gets to here, because she started early and Joe waits, the reality is for Jane to get to 2x herself was work, but for her to get to 4x is less work, 10x even less work because those skills keep translating. So by the time you get to 10x, she's not 4x, she's 10x. Or 20x. And this is unique. There's never been a technology where this kind of paradigm existed. So there could end up being a great divide between those who start early and learn and those who do not. And one of the things that— this is one of the few things I'll tell you that is hypothetical. I don't know. No one knows. But I can see this reality where there's these In the years to come, there's an AI elite. There's not only the working class and the service class and the industrial class and the creative class, but there's now an AI class. And these are going to be not only the Sam Altmans of the world who create it and control these systems, who may run the world, but it's also going to include another category of people. And that is the people who know how to use these tools best, who understand how they work so they can get the most out of them. And those people are also going to be part of the AI elite. So those are your must reasons. And I actually find the should reasons to be way more compelling because at the end of the day, what is AI? It's allowing you for the first time to explore your potential, not only as a professional, but as a human being in ways that just Have never existed in the entire history of mankind. So why should you adapt right now? Because it is the most fun you will ever have in your career. Because doing great work is its own reward. Because having 24/7 access to intelligence and creativity never existed until today. And now it's there for all of us to use. So these are all great reasons to get in there and have fun with it and explore intelligence, the notion of intelligence, the notion of creativity. What are these things at its core? People think they sit down and they ask AI chatbots such as ChatGPT a question and they're supposed to get an answer. If that's what you're doing, you're not using AI. That's not what AI is. AI is an exploration of yourself. It's an exploration of intelligence itself. And as much as you may know, one of the things that I found over the last couple years is that I learned as much from AI about being a human as I've ever learned before. AI, the idea of intelligence, is worth exploring. The other really interesting thing is that you think, okay, it's question and answer and, oh, you know, something more than a Google search, right? And it certainly is those things, but that's nothing compared to what it really is. One of the kind of unexpected consequences of AI is that it changes your relationships with people. And I can give you an example. I've got hundreds of arbitration cases against Airbnb, and as a result of those cases, I've had a very challenging relationship with the attorney who's always assigned to the cases, who I always go up against. In the final hearing. And we've done pretty well, but we've struggled because at the end of the day, we're fighting each other. There are big battles going on. We're having the same battle on the same case with a different plaintiff again and again and again. And that can get frustrating and bring its own challenges to the table. When I started using AI, one of the first GPT assistants that I built that was specialized for me using ChatGPT Assistant Builder was For adverse counsel communications. And I gave it really tight instructions and I told it exactly what I wanted to do. And the very first email that I sent to her using that tool, I realized in looking at that email how terrible I am at writing emails. And those of my staff, they know that I'm a technology lawyer who loves technology, who can't type and never could. My fatal flaw. As an attorney, as a technology attorney, but that translated into really direct, short emails that really didn't move the ball. With ChatGPT and its ability to write and communicate on a human level, the email that I sent got back, it put in there, by the way, I hope you had a great weekend. And the email that I got back changed the relationship that I had with this attorney. For the last year and a half. We are now good friends. We now solve cases together. We now solve them earlier. We work together to try and solve problems. And it all started with one ChatGPT email that changed the dynamic between us. These intangible things you'll find are going to be way more important than the classic output that you get when you're trying to learn something using ChatGPT as a chatbot. All right, let's take a look at what ChatGPT can actually do today, which is kind of important, right? It's like, okay, what can it do? And the answer is way more than you think if used properly. If you understand how it works, if you understand how to navigate it and how to direct it and how to interact with it and how to share with it and have it share back with you, if you understand how to use it, There is no scenario by which it does not blow your mind and change your life. Part of my goal today is to inspire you to start on that journey in a really more comprehensive way than you may be doing already. So here are some things that you need to know about what it can do today. There are primetime tools called foundational models, which is ChatGPT, Claude, Gemini, these big models that cost hundreds of millions to develop that only a certain number of very few companies have deployed. And work really well and better every day. And then you've got all these niche products that maybe have been specialized for law or for IP trademark, you know, research or document automation, document drafting automation, or, you know, Westlaw's got its AI tool, right? Westlaw's got a lot of money, but none of those tools, in my opinion, are ready for primetime. You might get some value out of them, but nothing like you're gonna get out of ChatGPT, Claude, and some of these foundational general models that right now your only task is to learn how to use them and learn how to use them well. Niche AI products are just noise and you can't necessarily trust them with your data. So my advice, stay clear. If you use a niche product, that's where you're usually gonna get frustration. This didn't work that well. I'm frustrated. I can't get it to work. Don't waste your time. Your time is too precious. The other thing that it can do is, you know, instead of just asking a question of ChatGPT or asking it to do something, you can actually build customized assistants using ChatGPT that are designed to do specific tasks. So for instance, I have an assistant that does client communications. I have an assistant that will do discovery in Airbnb cases. I've got assistants that I've built to do these specific things that are really powerful. And also in the last 4 months, we've had the launch of Canvas, which is Microsoft Word within ChatGPT. Instead of a thread, you're actually editing within ChatGPT and working on specific paragraphs. You've got projects, which are the next version of GPT assistants, which are custom tools that you build to solve your problems, to do your And you now have not only generative models that we've all seen where it just spits out the next word really quickly, to reasoning models that'll take 10 minutes or 20 minutes doing a deep dive intellectual thought process into your problem and return with a PhD-level analysis of whatever issue you've asked it to analyze. And you could actually see how it thinks. You can see how it took your question, re-asked the question, and asked itself, do I understand the question? Then it came back and said, no, I think I need more information. And then it asked you a couple questions. And then once it knew it had the question, then it started down the reasoning, the tree of thought chain where it said, well, first I need to know this. And then it goes and research this, this, and then it takes all of the information that it gleaned and it puts it back into itself. And then it starts down another chain. And that's why it takes 10 to 15 minutes. But this This is not generative AI. This is not a chatbot. This is an analytical mind that's doing the work on a problem and having access to all of the world's data that happened to be on the internet in order to get the job done. And it's doing it in milliseconds. So your mind is able to send messages in all directions at the same time, which is why you're able to think so fast. Before AI, everything was linear. A computer had to do this, and then it did this, and then it did this, and then it did this. These neural networks are exactly like the human brain where they can do tasks in hundreds of milliseconds, literally all at the same time in all directions. And that's why it's able to do what it does. This is extremely powerful and it keeps getting more powerful. By the day. So what it can do is it can amplify you if you use it correctly today. So who are going to be the winners and the non-winners on AI? And the winners are going to be the people themselves. Remember Thomas Friedman, The Flattening of the Earth, right? Technology makes everything available to people and therefore you don't need big corporations. Well, guess what? Law firms can't use AI. Companies can't use AI, at least not yet until we get agents. Right now, only people can use AI. So the big winners are us, individual attorneys, paralegals, and staff. We're the ones that get to get powered up. The only thing law firms can do is empower And enable their attorneys, paralegals, and staff to power up. Here's something interesting. Winner, intellectual property attorneys and law firms. Why? Well, what is AI going to allow us to do? Innovate, invent, create, explore, to do things beyond the outer limits of human potential that are now available to us. Well, that sounds like intangible property to me. And I do think that we as IP attorneys are going to have an opportunity here that is life-changing and is novel in all of our careers. Clients are going to be winners when they use AI and clients are going to be winners when they use lawyers who use AI. The legal profession, I think, is going to be a big winner here because let's face it, we've been trying to reinvent ourselves for decades. Get out of the hourly billing model. No one likes it, especially lawyers. What fun is it to get up in the day and have your benchmark be how many hours can you bill in a day? We're smart people. That's your life. That's what your goal is. That's what you do during the day. What a waste, right? What a waste. So we are going to see a big paradigm shift in legal practice and law firms, whatever the business model of law is today. It will not exist. It will not exist at some point in the future, and no one knows when that is. Is it 20 years? Is it 10? Is it next year? I don't know. No one knows, but it is inevitable. The non-winners are going to be the late adopters, law firms who refuse to invest in the training, the hourly billing model. If I'm a client, I don't want to pay by the hour anymore. I want to pay by the second. Or the millisecond. For us, we've got to rethink our revenue models. Certainly, if you wait, you're going to be behind. Courts and ADR services, which are really rough instruments of justice, are probably going to be overrun by not arbitration clauses in contracts, but AI clauses in contracts where disputes are resolved by AI, which can take all of the data, That all the dispute resolution, all the outcomes for a particular class of cases, all the information about the specific case, and predict on a percentage basis wins or losses. Why is it, why is that going to happen? Because that's what AI does. It predicts, it's a predictive model. It takes data and it can see patterns way better than we ever could in our experience. And so eventually, We will see these ADR systems kind of overrun or integrated with AI. So you hear the word hallucinations all the time. Here's just one word of caution. If it hallucinated, that is not the AI, that's you, right? Because at the end of the day, AI isn't there to get everything right. None of us are. When someone gives me a project, am I guaranteed to get it 100% right? No. I make errors. I'm a level of good. I give you a level of work product back. Same thing with AI, but your job as the person interacting with the AI is to work with it to continue to improve the work product. And to the extent it gets something wrong, you as the user should be able to see that and easily correct it. And now they will give you, these models will give you all the reference materials that they looked at. When they gave you the problem so you can double-check that. So you've got a winner mentality and you've got a loser mentality. And I think that you all hopefully want to have the winner mentality here. And so here's the things that you should be telling yourself. So get rid of everything you thought you believed or knew about AI. Put that on a shelf just for a minute and think AI is a force multiplier on everything I do. AI doesn't work for me. It allows me to do more, better work. AI is the greatest opportunity of my career. How do I integrate AI into my workflow and rethink how I do everything? I will commit the time required to study, learn, and grow with AI. I won't be distracted by AI noise. And AI skills and strategy and mindset I learned today will be leveraged tomorrow in even more meaningful ways. And the last one is my favorite, which is Attention Is All You Need, which is the most famous paper now in the history of AI because it gave birth to ChatGPT. And it's only a few pages long. And what it said is, hey, if we have more data to train on and we build an attention mechanism into the model where the attention mechanism can compress the internet into a small file, relatively small file, And the attention between words, phrases, the association and relationship between everything is what is in that stored area. And now we can access it in milliseconds because we know what attention to give words and phrases to spit out the next words and phrases. So let me give you a little bit of a, just a basic level view of what AI is. Right now, Tepperdeth large language model AI is a token stream. You see the token stream when you go to ChatGPT and you ask a question. Those get converted into tokens. Those tokens then get put up into a cloud, which is the AI model, which has been tokenized, which has tokenized the internet and formed relationships that can be pulled in all directions at all times to see what the next word should be. And then it spits out the answer, which is the chatbot. But that's not what you want to be doing because that's not much intelligence. It's just the intelligence in the model, very generalized. What if you upload a document into that chat thread and tell it, I want you to review the document and I want you to tell me what you think The most challenging arguments are for me to overcome in defendant's brief. Now it's going to tokenize that entire document, put that up. Now that all gets processed, and then now your token stream goes from this to this. Well, what if you provided custom instructions that gave them, really gave the AI very defined role of you're the best attorney in the whole world at intellectual property. Your job is to do this for me. Here's what I think you need to know about this case. Here is the challenges that we think we are going to have based on what the judge has said. All that gets tokenized, put up, and now your token stream, your intelligence stream goes from here to here to here. The more you put into that token stream, the more intelligence you're going to get. So that's attention. That's the attention mentality. That is going to get you where you need to go. Late adopters, I think that if you're saying these things in your head, you're going to likely go slow or not at all. And I think it's going to be a challenge. So AI's impact on law firms and business models. So you're a lawyer, you work for a law firm, you may be a partner, you may own a law firm, you're a paralegal doing work. What is, what is the impact of all this on, on me? Right? On my firm. And AI is going to affect more than how lawyers practice. It's going to affect our revenue models, which may change dramatically. I don't know what they're going to be. They could be subscription-based. They could be lots of different things. What you need to do is have the right mindset so that whatever opportunity comes down the pipe, whatever that may be, whatever that new revenue model is, you're able to take advantage of it. Client expectations are going to change for lawyers and law firms, and I already see it, and I'm sure you probably do too, because clients are sending me emails that were generated by AI on their specific problem. They're already more educated and they're asking the exact same questions that I would ask based on my domain expertise. Here's an impact that I've already seen in my world, which is delegation of projects rather than tasks. So, With my associate attorney, David, for instance, I'm much less giving him discrete tasks to do than turning over a project to him and saying, run with it. Why am I able to do that? Because he doesn't need me anymore. He's got AI to direct him and manage him and empower him. And so on a lot of this stuff, it wouldn't make sense for me to give to David, who's got a full who might take a day or two to get it back to me, who I then might say, hey, go check this out, and I get 4 days later. I can get all that back in 5 minutes and then some. Why would I delegate that task? That is getting looked at. How are we going to do this moving forward? How are we going to scale? What are the roles going to be and how are they going to change? It sounds ominous, but my experience to date is that it's not, that actually We are all going to be more busy doing more fun, more important work, and be freed up from the mundane. So I think there is outcome here that's very good for everyone. So what do you need to do if you're going to implement AI? Okay. If you're a law firm, it's kind of organizational AI, right? What do you do as an organization? And you need to start an AI working group. You need a lead responsible for what are those short and long-term revenue models that you need to be thinking about. A lead who's going to do training and workforce development, AI integration across your firm. You're going to need a lead to incorporate AI into the firm's culture because this isn't a tool. This is a fundamental shift in everything about the practice of law. That means it needs to be ingrained into you. You need a lead to oversee the hiring practices, which will eventually become what are the characteristics that are going to make for an empowered AI-enabled employee? And you need a lead who understands the current state of AI and its trajectory. You put these things together, you now are started down the path of AI integration. So there are 4 different things that you need to be thinking about in order to get the most out of these AI tools. And so if you've not been satisfied with AI to date, here are the 4 things you need to be thinking about in order to really get better at it so it can get better for you. You obviously have skills. You need to have certain skills. You need to build those skills, but you also need other things. You need a strategy. How are you working with AI? What is the approach of AI with this particular problem? How am I going to integrate AI? So you need to think strategically about AI. We've talked about it today, but you also have to have the right mindset. Mindset may be the most important of all these things because you need to be able to rethink everything you have ever thought or done in your world, in your professional world. Whatever it is that you would do to sit down and make it happen in your day, given your job, you need to be able to step back and say, does any of that make sense anymore? Because I don't practice law even remotely like I used to. And my process is completely changed because I've integrated AI depending on the task into that process to make it different. We're not going to be able to do that if you don't have the right mindset and you're old school. And then the last one is measurement. You should find out, did I, did I, was I 2 times better at this task? And every once in a while check in so that you realize, wow, this is not only better work. in the same amount of time or better work in less time or better work in more time. And oh my God, the intangibles that I got out of this particular effort were X, Y, and Z. And you know what? At the end of it, this task, I was 5x as good as I could have ever been before. It's important to reinforce that because you want to be able to measure your progress. So quad core number 1 skills, prompt engineering, structuring those token inputs. How are you putting all of that tokenization into the AI in order to get intelligence out of it. Meta-prompting. Meta-prompting is asking the AI to— and I realize I've not been advancing the slides here, so I'm going to go ahead and do that. I've actually got slides going here for this dry run. So let's go ahead and catch up. Thank you, Stephen. All right, so I've been going forward on my end, but not your end. So apologize for that. Quad core skills number 1 is going to be these, this meta prompting is really important because what they've learned by testing these AI models is that AI gets better responses if the prompt comes from AI. So what's meta prompting? It's saying, hey, I need you to do this and here are all the parameters and here's what you need to know. And before starting, it's asking AI to improve that prompt and then you're going to get more quality out of it. So these are the types of skills that you're going to want to start to think about and develop. And you're going to see these skills are going to be depending on the task, they're going to be different skills. And part of it is knowing which AI tool to use and when, but all of that will come in time. All right. Strategy is the second quad core thing that that is going to power you up. And strategy is what is your workflow? What, when you sit down at your desk, what app are you opening? What are you clicking on? How are you navigating your digital world? You need to change that and you need to figure out where AI can fit into that. Not necessarily to do it faster. I think one of the great myths of AI is it does the work for you and it does work faster. When you have access to all the world's intelligence, That slows you down because you're interested and you dive in and, you know, maybe you're just doing a simple email and it's faster, but maybe you're doing a mediation brief and you need to understand something or brainstorm something or think about something. And that is something maybe you wouldn't have done at all. So you're spending time on that, but you're going to get work product, which is way better at the end of the day. So it's more than time savings. Some things faster, other things Just more. You need to think about your digital world, your desktop. How is it set up? What tabs are open? All of these things. Where are you using ChatGPT? Are you using Claude? All of these things are going to be important to get the most out of the models. What is the strategy for this? And so think about getting good at the tasks and think about strategically how you're going to go about reinventing your workflow to get the most out of these tools. And if you're a firm, you got to figure out how you're going to integrate into your law firm. How are you going to train? All these different strategic things are going to become important. Okay. So this next one is one that really has permeated the whole discussion we've had today, which is mindset. And we've talked about kind of the winner type mentality here, and that is mindset. But I do encourage you to spend some time thinking about rethinking how you approach AI and what it is you believe. It's moving so fast that very few people know its capabilities as of today versus last month and what's changed, right? So think about it in a positive way and don't be distracted by the AI hype and don't get sucked down the rabbit hole of some new promise by A legal AI vendor that doesn't meet the promise. The last one is measure. And one of the GPTs that I've created, the GPT assistance is called the AI Amplify Factor. And it basically does a couple different things. It does a normalized time efficiency calculation, which is if I were to produce the same work product that I did with AI, How much time would that have taken me versus how much time it did take me? Sometimes it's going to be less time, sometimes it's going to be more time. Then you're going to do, and you're going to say, okay, well, what kind of quality did I get this? Was I 2 times better or 5 times better in terms of quality? Or was I negative 2 times better? You should know that too, if it did not work, because that means you have work to do. And then the intangible benefits that we've talked about, Think about what did I get out of that email in terms of relationship with an attorney that I deal with every day? That intangible relationship item was 20x. You put all this into the calculator. I'm giving you the link to the calculator, and it's going to compute an amplification factor that's going to say this, on this task, you are 5 times more than you ever could have done before. And doing that measurement is important as you work through the process because Because you should be kind of figuring out where you are, how good are you, and be able to feel the benefits of getting better. And the calculator, if you touch in with it every once in a while, is going to help you do that. So here are the tools that I use. ChatGPT 4.0, 3.0, 1.0, and now the new deep research model. Whisper speech-to-text. I don't type much anymore. Almost everything I do is talking to the computer. Talking to the AI, having it transcribed, which transcribes nearly perfectly. We all have been dealing with Siri for all these years, which has taught us that speech-to-text is terrible. AI speech-to-text is nearly bulletproof. And you don't have to say period, comma, quotation mark. It knows and does that automatically for you. And if you get it wrong and you want to restart your sentence, it knows that as well. And it won't type out your misstart. Fireflies.ai records all of my meetings with my clients and other meetings that are important. And then I get a transcript. I download that transcript. I input it into my GPT assistant for that client. I've got a GPT assistant that does client communications, but I've also got a folder. One of the new features of OpenAI's ChatGPT is to create folders. I have a folder for every client matter, and every one of those client matters has custom instructions telling them about my client, what's important, what their goal is, what their budget is, what the challenges are, and the files that are relevant to that client. So the token chain for every interaction I have with that GPT thread not only has my question or my thought, but also has all of the tokens from the documents that are specific and the instructions that are specific for that client matter. The output that I get is way better than the output that you can get without that kind of instruction and context being added to the token chain. So you want to start building out. And by the way, you will say the very first thread is going to be for a project, for a client matter. You're going to upload the arbitration demand. You're going to upload the response to the arbitration demand. You're going to upload every transcript of every discussion that you've ever had with that client. That you are going to, a year from now, you're going to have perfect access Everything that you talked about with that client and everything that client told you will be there and integrated into all your future work products forever. And then you're going to say big instruction, which hopefully you're dictating in so you can get it big and bold, everything you're thinking about. The first thread is going to be, hey, take a look at all the documents that I uploaded and take a look in my instructions and I want you to improve Those instructions. And by the way, I forgot to mention that this client is highly motivated to get this matter resolved. Then it's going to give you a much improved better prompt. You're going to put that back into the instructions. You've just improved that assistant. When we work together with our staff, with our paralegals and our partners and our associate attorneys, what are we doing if we're good managers? We're training them, we're helping them, we're giving them resources, we're giving them insights. Helping them understand the problem, providing feedback on their work product. Your GPT assistants are no different than that. You are helping them help you, and they get better all the time. By the way, every threaded discussion is this giant token chain that keeps getting longer and longer. If you're starting a new thing for that client, do you want that little token chain? No, it's got a bunch of information in it that may not be relevant. You want to start a new thread underneath that project with the same context instructions because you want to go back to zero tokens. And you can see that this is an illustration of if you understand how AI works, you're going to get more out of it. When you understand how it is actually doing it, it's like, that's a key part of developing skills and strategy. So how do you go about doing this? How do you learn all the things that are out there to be learned in order to get better, in order to amplify from 2x to 4x to 5x? Right now, the answer is essentially YouTube videos. And eventually, right now they have tools where you can say, I'm giving you 4 YouTube videos on end user AI. I want, and here's who I am. And here's where I'm from in terms of my levels. Take a look at those videos and give me 3 paragraphs of what I need to know as a business professional about AI or how I should be using it. That and 1,000 things more are available right now for learning. Because one of the great things that AI does besides our job, which is words, strategy, and analysis, That's what AI is. It's us, right? We're at most risk of being replaced, I guess, because AI does exactly what we do, right? But what you want to think about is how am I going to take this next pivot and how do I learn about things and take small steps and keep getting better? And you'll find that as you get further down the line, it's going to accelerate and all of a sudden you're just going to be off and You're going to have fundamentals. So YouTube videos are really good. Watch them at night, watch them in the evenings, get training videos with our staff. We're about to launch them into a whole set of new training videos to give them the skills that they need to be able to figure out what they're going to do next. And once they get, you know, off the training wheels, then what we do is we interact with them. We say, okay, What is the specific job that you're doing in your world where you think AI might be able to help you? And then we help them build a GPT assistant for that, or eventually we'll build a GPT agent that does that and allows them to do more important things. These are the things that are strategic that you're going to need to implement in your world in order to really be part of the AI elite. Here's some resources that you can take a look at. These are all hyperlinks. Even the paper, Attention Is All You Need, you should read it. It revolutionized mankind as we know it, and it's 4 or 5 pages long. Here's the tools that I'm using. Here's my YouTube channel. Here's that Amplify Factor GPT, which I can share, which I share publicly so you can measure your own amplification. And Then you're going to be off and running. So I hope today I've changed the way that you feel and think about AI and that I've inspired you to start moving forward. Go home and think about what it is that we kind of reviewed here today, and tomorrow get up and start to do something about it.