1h 5m 11 min read 13,293
Why I Cloned Myself using AI
Chris Do is building an AI digital clone of himself to scale his mission of teaching one billion people, and this is how it works.
Chris Do
Founder, The Futur™ · September 3, 2023
The Bottleneck of One
Chris Do, founder of The Futur, has a goal so large it borders on the absurd: teach one billion people how to make a living doing what they love. His critics, and even some fans, have questioned its feasibility. Do’s typical response has been one of patient faith. “I'm not sure,” he admits, “it might take multiple lifetimes to be able to achieve this goal.”
The problem is simple. Chris Do is one person. He is bound by time, energy, and the physical limitations of being human. He can only be in one room, on one call, or in one conversation at a time. The mission is infinite, but the man is finite. This is the ultimate bottleneck.
Then, a lunch at True Food Kitchen changed everything.
Across the table sat Sho Nakasone, a former storyboarding student from Do’s teaching days at Art Center College of Design. Nakasone, now the founder of an AI company, had a question. He wanted to show Do what he was working on.
He pulled a random company logo from the internet, FedEx, and hit a button. Within 20 seconds, his software generated a complete, sophisticated website. It wasn’t a clunky template. The site looked, felt, and sounded like it came directly from FedEx’s internal brand team. It pulled brand colors, selected appropriate typefaces, wrote an entire about page, and even generated a photography style guide.
Do was floored. “It was really good,” he recalls. That single demonstration was the catalyst that pushed him from being AI-curious to an active participant, leading him to explore tools like Midjourney, Stable Diffusion, and ChatGPT.
But Nakasone had a bigger proposition. “Chris, would you like an AI version of you built? Because we can do that for you.”
In that moment, the seemingly impossible goal of reaching a billion people suddenly had a tangible path forward. The bottleneck of one could be broken.
The Architect of Augmentation
Sho Nakasone is not a typical tech founder. His journey represents a rare fusion of art, design, and code. Growing up in Japan before moving to the U.S., he found his voice in art and later in design, seeing it as a way to bring “beauty and clarity to the world.”
But he was also, in his own words, a “super nerd.”
At age 11, he begged his parents to send him to video game design camps. He attended Flash Forward conferences, sitting front row, idolizing the creative coders on stage. “I didn't really see it as much as coding,” Nakasone explains. “I just thought of coding as another tool, like a paintbrush or charcoal.”
This hybrid thinking was evident during his time at Art Center. For a typography class with instructor Brad Bartlett that required a 250-page book, Nakasone didn’t just design one. He used basic machine learning to generate 12 distinct, beautifully designed books, each customized based on the content. He even built a website where a user could log in, curate a book, send it to a smart warehouse, and have it printed and shipped within 24 hours.
This was in 2012, more than a decade before generative AI became a household topic.
After graduating, Nakasone honed his skills at the prestigious Boston Consulting Group (BCG), where he saw firsthand how good data, structured as principles and guidelines, could drive powerful outcomes. He also spent time at the creative production company B-Reel. This background gave him a unique perspective. He understood brand, he understood data, and he understood automation.
In late 2017, he founded his company, now called show.ai, with a clear purpose. Augmenting intelligence. The goal was never to replace humans, but to give domain experts the tools to scale their identity and their story. “The purpose of the company really is to take these AI models and then train them on domain expertise,” Nakasone says. “So people can focus on their story and their identity and scaling that to a greater, larger audience.”
For Nakasone, Chris Do represented the ideal partner: a domain expert with a wealth of high-quality data and a clear mission. This expertise, he believed, was the missing ingredient for creating truly great AI.
Project Dobot: The Making of a Digital Clone
The project began, affectionately nicknamed “Dobot.” The first step was to feed the machine. To create an AI digital clone, it needed to be trained on the specific DNA of Chris Do’s thinking. The initial data set included:
- The Book: Do’s collected thoughts, structured and articulated.
- The Business Bootcamp: A comprehensive ten-thousand-dollar course covering everything from mindset and communication to bidding and negotiation.
- Q&A Sessions: Years of live coaching calls and audience questions, providing a deep well of situational advice.
Nakasone ran a test, asking Dobot a series of questions. The results were impressive. The AI could reference specific concepts from the bootcamp, responding to student-level questions with answers Do himself had long since forgotten. It was a good stand-in.
But Do saw a problem. The AI was just giving answers.
“If somebody were to ask me a question today, I would not just tell them the answer,” Do explained. “I need to know more about what their intentions are.” He believes a properly framed question is 50 percent of the answer. A great consultant or coach doesn't prescribe; they diagnose. This is a core pillar of his teaching philosophy, a skill he emphasizes in his coaching and sales training.
This feedback was the critical turning point. It wasn't about building a better search engine. It was about building a better thinking partner.
Nakasone recorded a conversation with Do, having him walk through how he would handle questions in real life. He then fed that transcript directly back into the AI with a simple instruction: “Update this persona for Dobot to include all of this feedback.”
The result was Dobot a Socratic method. It became more curious, more diagnostic. Instead of just delivering information, it started asking clarifying questions. The change was profound. Nakasone found himself having late-night conversations with the bot, wrestling with strategic pivots for his own company.
“I can actually share that with the Dobot at 2 a.m. in the morning on a Sunday,” he says. “And I can have a long back and forth conversation where the Dobot is able to help guide me through that thought process.” In one session, the AI’s probing questions led Nakasone to a moment of clarity so intense, it brought him to tears. Not of sadness, but of relief. “It feels more like a burden lifted off your shoulders,” he describes. “This is something that I was blind to before, and now I see.”
The clone was no longer just an archive of Do's knowledge. It was beginning to embody his process.
A Public Test and an Unexpected Failure
With Dobot version 2 showing promise, it was time for a public test. Do and Nakasone hosted a Twitter Space, positioning Do as a human conduit. The audience would ask questions, Do would feed them to the AI, and then read the responses aloud.
But something was wrong. Dobot wasn’t asking questions. It was giving long, formal, and sometimes fluffy answers. It was still impressive, with some listeners noting it felt “really empathetic, emotional, and kind.” However, it wasn't the diagnostic tool they had just built.
Nakasone was in shock. “Typically when a model gets an update, it's an improvement. It doesn't get worse.”
He quickly discovered the cause. Unbeknownst to them, OpenAI had just pushed an update to their GPT model, the engine powering Dobot. These large, underlying AIs are known as foundational models, and companies like show.ai build their platforms on top of them. The latest update had added more weight to longer answers, causing the AI to embellish and add fluff, a style completely antithetical to Do’s direct approach.
The experience revealed a crucial lesson in the new world of AI development: building on someone else’s platform means you are subject to their changes. The very tool that gives you power can also break your creation without warning.
This failure, however, provided an invaluable feedback loop. It wasn't a dead end; it was a data point. It forced the team to become more intentional about insulating their models from the whims of the foundational layer and to build more robust controls.
Nakasone’s team had to go back in and adjust the platform, which he describes as an amplifier with many knobs. They had to fine-tune the settings to counteract the update and restore the desired behavior, not just for Dobot but for their entire platform.
The Refinement Loop: Forging a Better Bot
The public test, while technically a failure, was a strategic success. It provided a rich set of data and user feedback that became the blueprint for the next iteration, version 2.1. The process of building a digital clone forced Do to gain a deeper awareness of his own communication style, a process crucial for anyone looking to build a strong personal brand.
The feedback highlighted several key traits of Do’s style that the AI needed to learn:
- Use of Metaphor and Story: Do rarely makes a point directly. He wraps it in a story or an analogy to make it more memorable and relatable.
- Concise Language: The AI was too wordy. Do’s communication is lean and direct.
- Question-Based Dialogue: The Socratic method needed to be the default mode, not an exception.
- Citing Sources: Do makes a habit of crediting his influences. He wanted the AI to do the same, prioritizing resources he has personally read and recommended over the vast, anonymous web.
This last point was critical. Do is not just his own content; he is a product of his influences. The AI needed to know what he knows. The team began training the bot not just on Do’s own material, but on the books and videos that have inspired him.
The result was a Dobot that was not only more authentic, but more powerful. In a private test, Do gave it a prompt: “Help me write a LinkedIn post that dispels the myth that ‘those who can’t, teach.’ Cite examples of famous coaches or teachers who are considered the best in their field despite not achieving personal success as a practitioner.”
Instead of writing the post, the AI responded with a question: “Sure, let's dive into this. But first, let's clarify: what's the main point you want to get across? Are you aiming to highlight the value of teaching or the idea that success isn't always measured by personal achievement?”
Do played along, saying he wanted to emphasize the latter. The AI didn’t relent. It asked another question: “Before we dive into crafting this LinkedIn post, let's clarify a few things. What's the main message you want to convey about these coaches or teachers? And how do you want this to tie back to your own experience or perspective?”
In that moment, Do realized his own logic was flawed. The prompt was loose, broad, and contained a contradiction he hadn't seen. The bot, by refusing to answer, had forced him to gain the clarity he lacked. This is the goal of strategic thinking, a skill explored in depth in courses on brand strategy.
“I realize I haven't really thought this through that carefully,” Do confessed. The interaction was a powerful demonstration of the AI's true potential. It wasn't just a content generator; it was a strategic partner capable of holding him accountable to his own standards of clarity.
Scaling the Mission, Questioning Reality
The implications of this technology extend far beyond a single user. For Do, Dobot is the key to finally addressing the bottleneck of one. The plan is to roll out the brand tune chat within The Futur’s private coaching community, The Futur Pro Group, as a value-add and a testing ground. The long-term vision is much bigger.
Do envisions a future where his primary role shifts from one-to-one coaching to training his AI clone. “I think where I would optimize my time is to spend most of my time training Dobot,” he says. The AI would handle 90% of user questions at a fraction of the cost of booking his time directly, which currently stands at five thousand dollars an hour.
This model creates a revenue-sharing opportunity for both The Futur and show.ai, allowing the mission to become financially self-sustaining and scalable. Nakasone’s platform can even handle e-commerce and user management, streamlining the entire business operation. “You can focus on what's important: the story,” Nakasone promises.
The project roadmap includes several future phases:
- Synthetic Voice: Training a voice model so users can have natural language conversations with Dobot.
- Digital Avatar: Using deepfake technology to create a video version of Do that can emote and react in real-time.
- Specialized Bots: Creating distinct versions of Dobot trained on specific tasks, like a social media coach that helps users write more engaging LinkedIn posts.
This evolution echoes the 2013 Spike Jonze film Her, where a man falls in love with his AI operating system. What once seemed like science fiction is rapidly becoming reality. “It feels like 2023, 2024, we're going to be right there,” Do muses, “where people are going to be talking to their operating system and having relationships, and it's going to feel as real as anything you've experienced in your life.”
Nakasone even envisions a future where a user could convene a “Jedi Council” of AI mentors, consulting with a Chris Do bot and an Alex Hormozi bot in the same conversation to solve a business problem.
The Mandate for Creatives
The conversation inevitably turns to the ethics of AI, particularly around image generation and copyright. Many artists and creatives fear their work is being stolen to train these models. Do pushes back against this narrative, siding with a design school professor who told him AI is not doing anything fundamentally different from what humans do: it looks at things and then makes its own version.
The technology, known as stable diffusion, doesn't copy and paste pixels. It blurs an image into noise and then learns to recognize patterns, an approximation of how a human artist develops a style after years of looking at inspiration. Effective content creation has always been about remixing ideas, a concept Do has explored when teaching how to build content people react to.
“If humans look at work, and we're all looking at work all the time... what are you looking for and why would you look at it if not in some way to borrow elements, assets, or that spark?” Do challenges. The difference is that the AI does it at a speed and scale that feels alien, leading to fear and controversy.
The real warning is not about the ethics of AI, but the cost of ignoring it. The wave of AI is not just coming; it is already here. It is an exponential curve of progress that will only accelerate.
Do’s final word of caution is a direct challenge to every creative professional. “Don't wait for the courts to figure this out. You need to be playing with the tools, integrating some of these things into your workflow.”
To resist is to choose obsolescence. It is to decide you want to be a photo retoucher who refuses to use software, insisting on building physical models and lighting them in a studio. You can do it the old way. But you will be putting yourself at a distinct, and likely fatal, disadvantage.
The choice is not whether AI will be a part of your future. The only choice is whether you will be a victim of the wave or learn how to ride it.
TRANSCRIPT
the word of caution I would throw out to everybody is don't wait for the courts to figure this out you need to be playing with the tools integrating some of these things into your workflow and you draw the line for yourself but don't artificially handicap yourself it's sick to say you want to
do photo retouching but you're not going to use any software to do it you don't want to use Photoshop you don't want to use illustrator you don't want to use any 3D image making program because you want to do it the old way you want to build real models in real life and
you want to light things and that's how people used to do it you're putting yourself at a distinct disadvantage at some point He suggests to me Chris would you like an AI version of you built because we can do that for you the purpose of the company really is to take these AI
models and then train them on domain expertise it's all about augmenting intelligence so people can focus on their story and their identity and scaling that to a greater larger audience all right my next guess everybody is someone this is a kind of an interesting unusual thing for me a former student of mine
who's gone on to do some big incredible things and one of these random reach outs where he's like hey Chris I'm going to be in town let's have lunch together and normally I'm like Mia as the introvert I'm not gonna have lunch I'm just gonna stay home and eat by myself but you
know what as a former student I gotta go meet my former student and see what he's up to I got to tell you that lunch that we had over at True Food changed a lot of the things I was doing and it's led us to this point here in time and let me
tell you that story and then I'm gonna have my guest introduce himself in just a little bit so while we're at lunch show that's his name he says do you want to see what I've been working on with AI and Chris are you using AI tools now at this point in time I
think it's uh let's just pretend it's about a year ago before everybody's talking about AI but there are enough conversations that are bubbling up to the top where I'm like yes I know yeah there's things going on I'm not ready to dive into it he shows me some of the tools he's been
developing and how he's been able to leverage all of his time working on Brands brand Bibles strategy guides all those kinds of things and automating this very complicated process to make it like just flow what he asked me to do and we did this live with his laptop believe it or not was
just to pull a random company or logo from the internet I think we happen to pull FedEx and then he hit a button and within 20 seconds or so give or take and it was slow because the internet was a little slow where we were with the Wi-Fi it generated an entire website
now you're thinking yeah Chris what's the big deal everybody can generate pretty ugly website here's the thing this website looked and felt and sounded like something that came from FedEx this was incredible part everything was generated it pulled the colors it picked the typefaces it wrote the about page the philosophy the team
it pulled images it wrote the the photography style guide and I was reading through it and of course I didn't read the whole thing but it was really good that led me into diving in to mid-journey into stable diffusion and then later on chat GPT it's all because of show and if you
don't like what I'm doing with AI you can just blame him because he's the one who pushed me over the ledge and so without further Ado so welcome to the show thank you so much Chris it's it's truly an honor I I remember back at Art Center I was one of your students
right I think this was back in like 2009 2010. uh even back then uh your teachings really kind of resonated with me I think it was a storyboarding class at the time and that I remember those are the toughest uh critiques like typically um I went to like an art sporting School beforehand
so I was really used to kind of beesting my way through critiques and I knew I couldn't do that with you in your class and then so uh ever since then I've been watching your videos and so it's really really cool um opportunity for me I'm super excited to be working on AI
with you that lunch too I was really excited having that conversation with you at True Foods um at the time you know a lot of people didn't respond that well to AI especially like creatives and different things like that um when I when I first started the company it was back in my
2000 late 2017 uh early 2018. and then uh I think like open AI release GPT one back in like 2018 essentially and it wasn't like a chat and it wasn't really powerful of different things like that um but I was able to do what you saw back then which was create brand style
guides I also kind of knew from my previous experience working for like big Enterprises at BCG that good data really made a big difference and then a lot of that data kind of looked like brand guidelines to me they were essentially like principles guidelines examples that you can use to train an AI
to kind of behave like you and then in that way you can kind of focus on the story uh the stuff that we dream about or we do at art school and unfortunately don't get to do as much in practice just because um with the what's with the tool like AI essentially um
since then we worked with agencies we've worked with Enterprises but we're we're really really excited is teaming up with people like yourself because you have a wealth of incredible data you're basically um amazing curators of incredible data that will lead to what we believe the best AI in the world um and in
this way our um my mission kind of goes back to uh pre-art Center but is to bring Beauty and Clarity to the world um and right now I see the best way to do that is through Ai and partnering with people like yourself so I'm I'm thrilled to be here thank you so
much Chris all right before we go any deeper I have to get you to introduce yourself in case people don't know who you are show please introduce yourself tell us a little bit of story of course tell us a little story about yourself first of course so a little bit of a background
um kind of grew up in Japan I went to an international school um eventually came to the states um I was really into art and that was kind of the way that I saw icon being a beauty and Clarity to the world but it's really cool I thought I had my work in
Carnegie Hall and different things like that art was really my passion but I soon kind of found a limit into what I was able to do with art at the time um this was back in uh high school I went to a boarding school and went spent got to spend like eight to
ten hours doing art but I really thought of a design because of the ability to kind of solve problems for people and I felt like um I can really see kind of tangible impact of my work and I can do in kind of a very beautiful way if I recall correctly from our
conversation about a year ago I think you were writing scripts or ways to automate things this is pre-ai I believe if I'm not mistaken correct me to tell me more about that even in back in my 2011 2012 I was using base basic machine learning and conditionals and different things like that to
do my art center homework so I don't know if you remember Brad Bartlett had like a typography class where you had to kick out a 250 page book um Brad wasn't too happy with me because I wasn't able to show a lot of my process but at the end of the final I
was able to show up with 12 different books on beautifully designed by AI based upon its content and then I created like a really sophisticated piece of software to like even integrate things like user curation or ratings or even stock market prices like customize a book to the user um and then so
that that was always something that I was I was thinking about do you still have copies or images of these books that we can see oh yeah of course I'd be very happy to share that it was uh it was a moment of Pride for me because I even connected it to Smart
warehousing like I I created the website you can log in create an account choose the type of book and curation that you want um and then you can actually send it to print and then within 24 hours you can actually receive the book so I at the time I was really really excited
about it and then in some ways in the back of my head um I still have thoughts about releasing it as a book that's very cool if you're listening to this podcast check the show notes and we will share a link to that work so you can track along with our conversation this
sounds fascinating to me so we're talking about 2010 you said about 13 years ago you're already using machine learning and using AI yeah oh 2012 okay so it's about 11 years ago you're already doing this what I can't figure out show is that artist becomes a designer where's this knowledge or this desire
to do training models or even poking around the AI space when did this get introduced in your life and this is very unusual for a design student to be whipping out machine learning AI code to generate art projects absolutely well I I was always like a super nerd and then so uh just
to give you an idea when I was like 11 like I would beg my my parents to send me to like a video game design camp where I would like learn how to code video games um I was also really big in the flash world at the time so I like beg my
dad to like take me to like flash forward conferences and different things like that um and he was also very supportive of that too but essentially I was I was coding from an extremely young age and just it was um I almost I didn't really see it as much as coding as just
another I just thought the coding as another tool like a paintbrush or charcoal or anything else and then just like me I just look at the same way yeah of course no not really yeah okay so you beg your parents to send you to flash forward and to send you the video game
academies and camps did they do it did they let you go yeah yeah and then I would often at times get like you know just be able to walk in as a student to crazy things too I had I had this tendency when I was really young I realized oh like No One's
Gonna Stop Me so I used to go and attend really cool things all the time for really really cheap if sometimes for free people would just be like I would ask for a ticket and they'll say just go on in yeah um yeah when you're young people are cool they let you get
away with lots of stuff because they admire that there's a young human who's passionate and inspired by the things that they think no kid is interested in this kind of stuff so you must be one of the youngest people at The Flash Forward con uh event right you must yes yeah and I
loved it I was in front row Center wow um I I really idolize the people on stage my mission is to bring Beauty and Clarity and these were like the examples of people that were doing that in a really creative manner yeah so I was having a blast and were you living in
Japan at this time or you live in America uh so I lived in Japan until I was about 11 or 12. yeah something around that time so you're going to these academies while you're in Japan and in Japan I I was an artist but really all this stuff happened around the 12 13
year old age I see uh you know when you go to a new country and then you know you don't have friends and different things like that I got really good at art okay all right so all right now we understand the DNA part which is there's a strange person that is a
hybrid between art design and coding and video games and just probably more advanced than his age and is able to do lots of things and I think this is important for the context of like who's this guy talking because he's talking so fast and I'm having a hard time keeping up with them
it's because his brain's working probably in multiple like parallel processing multitasking and shuffling around lots of pieces of information okay that now brings us to where you're at today with your team how big are you uh how many of you are there and how you said there's a couple different offices and and
what is the purpose of the company uh we're still a small startup there's about two dozen of us and we operate a lot like a production company so I used to work at be real which is a production company so we we typically um hire contractors that are specialize at really good things
um and then uh for a specific task within the project and we have really good lead engineers and system Architects within the team full time the purpose of the company really is to take these AI models and then train them on domain expertise intelligence it's all about augmenting intelligence so people can um
essentially focus on their story and their identity and scaling that to a greater larger audience um it started first started off with just helping people create brand assets and different things like that but as you can see with these Foundation models they're so flexible that by essentially using our platform to give these
AI models context in a proper way you're able to get really really great outcomes it's also a great way for people to kind of have greater self-awareness and kind of reflect on how their system works we noticed that companies that do use our platform is able to kind of hone in on who
they are who what their identity is because the AI will be able to reflect back to them if something is not working or not okay let's get to the meaty stuff for the conversation stuff I'm super excited to talk to you and share with our audience about today which is this so in
one of our many conversations Sean and I've been chatting via Zoom or whatever we're discussing things at some point He suggests to me Chris would you like an AI version of you built because we can do that for you of course then like I said I'm straight my ears perk up like yes
because instantly fans of the show know this they know this about me we have this really big hairy audacious goal which is to teach a billion a billion people how to make a living doing what they love and some people are our fans and our critics have questioned is this even possible and
to be honest my answer usually is I'm not sure it might take multiple lifetimes to be able to achieve this goal with the bigger goal of disrupting the educational model as we know it today and I'm not worried because I feel like in the way things have worked out in my life that
you set your goal you get clear on it it helps you to make better decisions but eventually you magnetize yourself and you draw the people and the resources that you need to make this thing happen so now everybody who's getting ahead of the conversation knows exactly where this is going to go Cho
says we can build a digital clone of you that can teach and consult in ways that are eerily familiar to you and I'm like I need this show let's do this and we don't even know what we're doing but we're doing something together and we're affectionately nicknaming this robot this AI trained model
dobot I know you guys get it dobot you're smart and Show's been showing me updates to dobot and I want to talk to through the the process with with them and then you chime in where you think it's I'm Miss speaking or you want to add layers to this okay show absolutely yes
okay so version one he's like okay what information do we feed this thing and I said well it makes sense feed it my book those are my thoughts collected it's clear and then feed it the business boot camp and you're also an Alum of the business bootcamp it's a ten thousand dollar program
that's got I think seven or eight modules on all the things I need I think people need to learn to run proper business from bidding to negotiations to mindset communication all that stuff okay it's all there and then also give it all the Q a sessions that we've had over the last couple
of years and feed it that data so now it's a pretty good stand-in for the teacher who ran that course with me the business boot camp and so we looked at that model and it was really good so show in his computer brain runs a test he asks the robot dobot lots of
questions and he compiles the answers and he shares a spreadsheet with me he goes Chris here's the latest update and I look at him like I'm very impressed because it refers to very specific things that are said if you asked it as a student of the boot camp a question or two it
would know exactly what you're talking because it was trained on that material could respond immediately things I've already long forgotten pretty cool but then I asked Joe this question which is if somebody were to ask me a question today I would not just tell them the answer because I need to know more
about what their intentions are and I'm a big believer in this I properly framed question is 50 the answer so he said let's get on a call let's record the conversation and tell me how you would do this in real life and we ran through a couple scenarios he said this is great
I'll pause the story there you go do something and now we're at dobot version two what is it that you did with that conversation so people understand yeah so uh Chris was very clear in his feedback to the AI in terms of what he wanted from it next and then so I literally
just took the transcript from that conversation gave it to the AI and tell that hey update this persona for dobot to include all of this feedback um and then this this is probably a process I will continue to do for the rest of our lives essentially where we're continuously making these kind of
small fine-tuned alignments um but it's it's really exciting to see how well the AI is able to respond if you do properly give it the right the right guidelines and the right examples right so you said a lot of this is not coding it's training the models it's curating it's giving it feedback
so that it becomes more of what it is that you want it to do so while you're testing the version 2 of dobot which was more Curious more question based more diagnostic than it was just giving prescribed answers to things and you were telling me a story about how you were talking to
it at night and you you're kind of having a different relationship with dobot now share a little bit about the dialogue that you're having the where it's really beneficial is that I can do things like oh uh my my investor sent me this email or something or my board member or my co-worker
sent me this and I don't know how to feel about it I can actually share that with the dobot at 2 am in the morning on a Sunday and I can have a long back by back and forth uh conversation where I'm able to basically um or or the dobots able to help
guide me through that thought process of figuring that out I was able to do that multiple times and it's actually in many ways uh behind some of the Strategic pivots that we're making at a company is the ability to have these kind of conversations in kind of real time uh typically you know
um people like Chris don't have the time to be reading through all your stuff or um you know looking at all these different things but Adobe especially at 2 A.M on a Sunday right but sometimes that's when you need Chris the most and that's where you can get a version of Chris essentially
I can kind of walk you through uh some of these things that he's refined over a long period of time um and then so in this way we hope to see like the you know the benefit are the the positive effect that Chris can have on people uh even when Chris is not
there it for for me it was huge um even after being a kind of a business boot camp remember uh there were tons of questions that you continuously have because every scenario is kind of unique um it was it was a fantastic tool for me okay you can see the obvious applications in
in the implications that having a digital clone of yourself especially if you're a teacher or someone who wants to do something at scale obviously I'm one person I'm tied to the people that are in front of me or listening to it and the level of interaction is limited by to the number of
people who show up so in a way I mean show was telling me the questions the robot the AI trained dobot was having with him he even got emotional talking to it and we were just joking yeah I pretty much cried yeah why did you cry what did it say or ask you
that God got you emotional well uh typically you know the questions that I'm asking chat Bots at 2 A.M are are pretty deep were like um you know there's a lot of uh especially going being an entrepreneur you know uh dealing with a startup and then this is my first time really managing
people I manage people at BCG but I wasn't signing their paychecks right so there's a lot of new dynamics that happen to me on a date on a weekly basis uh and then so kind of realizing one that oh like hey this was something that I might have been like uh not addressing
kind of shoving into the shadow some of the questions that the the bot was kind of that was asking of me to kind of question questions that I needed to ask myself and then through that process um yeah it could get pretty emotional very quickly but this is this is basically kind of
a happy emotional right kind of almost like a light light and emotional where it feels more like a burden lifted off your shoulders than like you know oh I yeah I figured out this one you know it's not that kind of motion it's like oh wow like this is something that I was
blind to before and now I see can be an emotional experience when you get that Clarity and you feel like somebody's really paying attention to you I had these moments when I'm in therapy when I used to go to therapy with with Joan and she would ask me some questions I felt like
for this one hour that we spend together she's giving me her undivided attention and very rarely does it happen not even for 10 minutes like in our daily interactions we're always distracted with the things that we're doing so can feel deeply emotional okay so here's the interesting thing a couple days after version
two of dobot we decide to go into Twitter spaces and test the version two and people were told that I'm gonna just be the human facilitator uh the conduit between people on Twitter and dobot so they would ask questions I would feed those into the AI engine and just read people the responses
but something weird happened dobot did not ask questions and just gave answers which surprised me a little bit but there's there's a little bit more to the story right but everybody's semi-impressed like some answers felt a lot like me and one one of our friends needy even commented like it felt really empathetic
emotional and kind and I'm like well it's not me obviously because that's that's not me so that's pretty cool like the clone of you can be all of your strengths but it has the infinite patience of a computer which is pretty cool and really neat so what happened was she'll explain to them
why it didn't work the way it was supposed to of course um so we're built on the world's greatest Foundation models and then right now uh for text that's GPT uh Google has their models uh Facebook has open source models and different things that we're exploring as well these open source models are
getting more and more exciting we have the ability to switch to between them um right now we're using GPT as a foundation model that we train and build our platform and like provide the data engine and different things like that too uh GPT had an update on Monday so right after um Chris
even even me uh we tested the AI they released an update and then in that update essentially uh there was more weight towards longer answers which is not the case before so it was it had a tendency to kind of embellish or add fluff which is obviously not your style so that was
extremely problematic when I first saw that I I was so in shock and almost cry because typically when um a model gets an update it's an improvement it doesn't get worse but uh to us like increasing the length of the answer wasn't necessarily an improvement so we've had to go back and um
this is something that we continuously do is that we go back in there and refine and say okay let's adjust the platform like almost think about if the platform like an amp had little knobs where you can change certain things we can make these kind of fine tune adjustments for these updates and
then so we've had to make a little bit of an update to your model uh or essentially this is actually platform-wide to all the different platforms uh to make sure to make sure it doesn't add all that fluff um there there are you know scenarios where another thing it it won't mention certain
authors or different things like that essentially it has another layer of censorship that has been developed um you know the people behind GPT is open the eye people behind opening eyes Microsoft so they really do have to play by the rules um that's that's where it's going to be interesting to see um
how the landscape shapes up between these kind of um open source models that might actually start to outperform these Enterprise models base basically because they have less constraints the short answer to the question was in summarizing what you were saying is unbeknownst to us GPT had updated and broke the model and so
it wasn't giving us a question-based approach it was very wordy and it was speaking uh more formal than I would and so it was an impressive demo but it wasn't the one that we had prepared for so we went back we took all the feedback from the live Twitter space and we started
to implement a lot of the things that we're talking about so we're at version 2.1 what's what's kind of interesting is as you try to build a clone of yourself you actually kind of need to know who you are and the tonality and the things that you do that you're not aware of
so our audience helped us out with that they would say oftentimes I'm using metaphor or I'll wrap it around the story to make the point I'm not using this many words and there's a certain tonality that I will speak in that it feels a little bit different than the the two the human
version versus the robot and of course I asked a lot more questions that kind of help to pinpoint what the person's working through they also well I guess they made me aware of something that we've had to train exclusively on content by me but I am not just my own content I am
the product of multiple influences for the books I've read from the videos I've watched and I wanted the robot to know what it is that I know and to put prioritize those resources over this the wide web and I said also when it mentions a concept I normally try my best to cite
and quote people that might be one of those issues where we can't fully resolve but at least I try to because I think it's honorable to cite your sources both for you you and for the creator of the original thought helps blows people out and I'm always a little leery when people say
things and they say it as if they came up with it by themselves so version 2.1 You released you shared with me I'm playing around with it and then I have to tell you it's working beautifully last night and I just posted about this on LinkedIn last night I played around with it
I gave it a prompt and it did something I hadn't done before with me I I wrote it I'm gonna just read to it to everybody who's listening what I wrote and how it responded and prepare to be shocked or in all of how this technology is progressing for some of you who
are anti-ai I'm just going to ask you to keep an open mind you're probably not here anymore because all this talk is probably ruined your day but for the rest of you that are still here let's get into it okay and I'm gonna say this because it's kind of embarrassing that the robot
helped me to realize my logic was flawed I asked it help me write a LinkedIn post that that dispels the myth that those who can't teach cite examples of famous coaches or teachers who are considered the best in their field despite not achieving personal success as a practitioner if possible include Phil Jackson
and Emmanuel Stewart now if you're paying careful attention to this you might see what the robot has already seen But at least was blind to me I'm fully expecting it just to start writing as chat GP would normally right but here's what it says in response it says sure let's dive into this
but first let's clarify what's the main point you want to get across are you aiming to highlight the value of teaching or the idea that success isn't always measured by Personal Achievement whoa it truly understands what I'm asking for and it's asking me some questions I'm like all right let me just keep
playing along and I say to it kind of you know like how I would normally say to somebody both but with a greater emphasis on the latter and now I think okay for sure it's going to give me a response but it doesn't it's just before we dive into crafting this LinkedIn post
let's clarify a few things what's the main message you want to convey about these coaches or teachers are we focusing on their ability to bring out the best in others their unique teaching methods or something else and how do you want this to tie back to your own experience or perspective at this
point I'm like oh my God my brain hurts nobody's asked me too many questions and I'm just thinking give me the freaking answer but I realize I haven't really thought this through that carefully because my original prompt was loose it was Broad and they had a big flaw built into it and I
think if you're really paying attention you figured out what the flaw was but I'm not going to tell you I'll tell you online somewhere on social media okay just think about that so this is like how I would talk to people maybe not as patient maybe I wouldn't use these exact words but
whenever someone asks me something I don't presume to know what they're asking or that they themselves know what they're asking for so only in dialogue and asking more questions can we both get to a place of clarity because I think I'm not sure if it was Maya Angelou who said this there is
no such thing as the right answer to the wrong question so if it goes and answer it it's going to send me down a wrong path and I love that kind of interaction so I've been sharing as much as I can live via social media on different platforms the progress that we're making
and it's I gotta say it's really cool and it's a little bit scary how good it is already okay do you want to respond to that show absolutely and this is just the beginning uh just so you know uh just this morning we added the YouTube videos that you sent us as ref
like the things that were inspiring to you like books and different things like that so it should be uh even better than before so and then uh the really really exciting part is uh there is still a lot that we can do even within the platform with the data that we currently have
to further optimize it and make it better uh uh there's just tons and tons of strategies around that uh there's also one of the things that I noticed we we I don't think we put the pot your podcast data in here yet but things like the podcast data because it's so q a
as we are right now it's extremely applicable to a kind of more of a chat experience right now we have a lot of business bootcamp information in there so it actually might be really good at creating business boot camp courses for you know help you through that process in the future I'm really
really just like yourself I'm really really excited to continue to see this progress I think um uh just like we've seen with some of the image Generation stuff that you know you've seen uh how good it's become over the past year so the same thing is happening with language models uh people are
just um it's a little bit harder to detect you know like if you're not a great writer can you really tell the difference between an amazing writer and a legendary writer um it's it's it's becoming one of those worlds right now with AI right this is wonderful okay we've gone through this and
I think what you're doing is you're providing me with a tool and resource that I'm going to be able to help a lot more people because like you said if it's two in the morning if it's a Saturday or Sunday I'm not really that available forget about the long emails or the PDFs
you want to send me I'm just not that available and I can't be but I do Aspire or intend to help as many people as I can it's just there's a limit on what I can do without hurting myself in the process so this allows me to create something that can be used
for many people it allows them to have some form of relationship with you and it reminds me of that film I think it was directed by Spike Jones It's called her and it stars Joaquin Phoenix and and he falls in love with his AI which is fairly common in this futuristic world and
back then this felt like pure science fiction but it was a novel concept that a person falls in love with his operating system uh voiced by oh what's her name Scarlett Johansson it's voiced by Scarlett Johansson and of course she has a very seductive sexy voice a Husky and he's talking to her
and he falls in love with her as much as he would fall in love with anybody they have phone sex they have deep relationships meaningful conversations and it feels like 2023 2024 we're going to be right there where people are going to be talking to the operating system and having relationships and it's
going to feel as real as anything you've experienced in your life and then it calls into question what is real because I I don't know about you but I've had long distance relationships before where I couldn't see the person it was purely based on the phone and it's not that crazy of an
idea that you can actually find meaningful personal deep emotional connections with something that is artificial I completely agree the the movie Her is not you know you don't look at his behavior and say that's a crazy man you say you know that's a that's a man that clearly lacks love in his life
but that relationship seems to work where there's kind of like Mutual benefit is clear it seems in some ways it even seems to be healthy in the sense that the Scarlett Johansson the AI was so wise that it was kind of able to guide um him towards happiness yeah ironically you know she
leaves him spoiler alert first okay everybody rewind a spoiler coming okay go ahead oh sorry sorry yeah yeah so if you haven't seen it uh that there's a bit of a shock there um there are you know certain people now that are even like building servers within their home so they can have
like their own AI running at all times within their home that can't like leave or like uh people are already figuring out um but I I love you know one of the things that has inspired me to create what we do today is some of the movies and what we've seen right um
where I can say oh yeah you know oftentimes they show what can what can happen in the worst case scenario but often times I can kind of see the best case scenario the one of the first movie campaigns I worked on was for The Avengers film the first Marvel Avengers back in my
2010 or something like that and I had to create a lot of like the graphics in 3D for Iron Man and then to me back then I was like you know this this seems possible you know I mean you know I'm just creating designs and after effects and Cinema 4D stuff right now
but this seems possible um like even if even without a neural network or something as complex as we had today even the code from back then we could have probably made something happen and then I really love that idea of being kind of like a queue or like I you know cue from
James Bond where the tech guy that is able to provide a a Jarvis like technology to Great domain experts and leaders like yourself uh I even have the glasses right you have the Tony Stark like essentially because I I know in the hands of someone that can when when the AI does ask
you tough questions you're able to actually help give it Direction that's What AI is really lacking today and then um so and then oftentimes you know those questions that you talked oh these are tough questions when it comes to an Enterprise or brand and some of the bigger brands that we worked with
uh they're very slow to answer those questions and so it's exciting to see how much more power people like yourself are going to have in the future even compared to bigger Brands because of your ability to make decisions and to guide the AI and essentially give it a compass so it can run
in a direction that you need it to so show what's the business model for something like this I know you're helping me and eventually there's going to be some monetization model there has to be business here because you can't support a team of 24 people on just Goodwill right yes so talk me
through like if this excites somebody where where do they go with this kind of thing and how do you see this rolling out for me and for others yeah of course I mean uh people can go to our website show.ai you know we have pricing on there um essentially the pricing is reflective
of the value that we bring to the company or the brand or the person right in terms of efficiencies and productivity their ability to kind of 10x their presence using our tool a big part of what we do and it's kind of baked into our pricing is this idea of like these AI
live hours where essentially you work with our team where we work with some of your data and we help you kind of make the right decisions in terms of making the AI That's that's best suits you um for another brand they don't want want a dobot right they want their own version are
a different expert maybe they're in real estate or different things like that they want their version it's very different data and they're saying different things in a very different way and then maybe some of them you know last thing that they want is a follow-up question from their butt it really depends and
then so um we're really excited if people are interested in doing something similar where they are interested in utilizing this found these Foundation models but they wanted like a good per like a good co-worker they want them to be onboarded onto the mission we're the place to go okay so for for me
I can uh the the rollout plan for us and it's happening right now you've been working with our team to integrate a dobot chat presence inside of our paid walled off coaching Community called the future Pro Group and that's going to be value-add for everybody there but the idea there is is that's
not the Final Destination it's just a testing ground so that we can get better data help the people that we serve already and get really valuable feedback because in my future here's what I would love to happen I train and teach people on one-to-one on a one to few basis but I think
where I would optimize my time is to spend most of my time training dobot to be able to tackle questions that people feel like that wasn't a very good answer or response and so it keeps getting better models it gets better guidelines from us and so it's going to continue to be more
and more refined so it gets to this place where it's going to handle 90 of your questions and it's going to be good enough for you at a fraction of cost of what it would cost to book me for an hour for so for example for context here if you wanted to book
with me for an hour it would cost you five thousand dollars an hour I don't have many takers for that at that price someone else designed this way because I can't just keep selling my time there's a finite amount of that and so I want dobot to do most of the work but
ultimately we've seen other models like this where dobot lives in the world and you could pay for a number of questions or a period of time and then there's a model there where there's revenue for for both of us where there's some kind of rev share and we can see it paying for
itself in the team and growing not only the educational presence but growing financially for both of us right is that where this goes show absolutely um we we actually even have a system where if you wanted to you can launch your own kind of the future Pro Group completely within the platform so
we talked about the ability to create websites and different things like that uh even things like managing like your payments and your users and different things like that these are all capabilities that we can really streamline for you so you can focus on what's important the story and then uh manage every and
then manage every I don't have to worry about like all the other technology behind the closed doors because it is a race there let us kind of handle that um and in terms of the flexibility of the system you're completely right you can you like people can utilize the API to integrate it
within their software um and we have tons of incredible integration tools even things like uh the system can even generate CSS and different things like that for you now at the very bare minimum right you can just think of it as you know AI trained on you and your identity or your brand
um but I I the vehicle itself is actually pretty complex in the sense that it has 600 plus features so imagine this chat GPT or stability or these different things with 600 features that you can actually integrate this technology into your brand system whatever that might be and that could be future Pro
group that could be your newsletters it could be your website um and these days these touch points are are getting out of hand but let AI help with that right um and there's there's a couple of other faces I see it going through potentially one is we get the language model to work
so good that we're like yeah we're pretty convinced this is you and all my friends and family like yeah it's pretty it's got the same snappiness and it's asking the right questions and helping me to get clarity and then we take it to like using synthetic voice so now it's going to speak
like me and sound like me so if you really want to quote unquote talk to me in a natural language model versus writing things down via text prompt and then the last one is to use potentially some deep fake version of myself where it's like okay this is a digital version of you
I know it's not you but if you want to look at a face you want it to emote and react to you and sound like me that would be the third phase of this but you and I already talking about well there are specific ways that I create content the way I write
posts and so there could be a dobot that's trained on this foundational layer but one that's has specific data and weights to things about how I like to create content so if you're trying to grow your social media presence and you want to write higher engaging posts on LinkedIn you give it some
information and dobot walks you through the process and coaches you through it versus just rewriting it for you I think that gets really interesting so now dobot is a teacher or a coach and helps you to write better in terms of getting because I'm not an English professor in terms of getting higher
engagement on your posts and we can see it doing many different things not just being a consultor or a therapist or whatever so lots of applications I'm super excited where this goes now I know you prepared a couple slides I don't know if you want to show them if this is appropriate time
or if you want to fit or you want to finish another thought what do you think oh sure I would love to show the slides yeah just to piggyback on what you're talking about um you know one of the huge things in growing the startup over the past six years was surrounding myself
with kind of like these mentors are coaches they're extremely extremely valuable you know one of one of the beautiful beautiful things about the Dubai and then things like in the future would be that people will be able to have like a Jedi Council where they'll be able to talk to like Adobe or
and the Alex formosi bot in the same room about a problem that they're having I I'm really excited to see how this kind of scales yeah wonderful it should be huge value to everybody I'm gonna share my screen then yeah please yeah okay so here um this is kind of a little bit
of a framework breakdown of how the system works right and how we're able to essentially create dobot that we have this beautiful platform and this is essentially the AI data engine where it's like a long-term memory of the AI so it has contextual relevance to create answers that are specific to Chris or
specific to your brand we're built upon tons of different models some to do like AI tag you have images some obviously GPT to create text stability to generate images Dali Etc so you can get access to them all in one single place but the beautiful part is you can train it from one
place too so they all kind of talk the same language we do that by essentially filling it with the brand compliance and real tuning layer so this is some of the rules like the personas or you know some of the feedback that Chris gives us the data and assets so this is like
the business bootcamp or different things like that and we also have the ability to integrate so if you did want to integrate Google analytics or we have all these different blocks these are basically Senses Into AI so dobot basically has Senses Into oh this is how the marketing website is doing and different
things like that essentially becoming more contextually aware giving you better answers this is a little crazy but essentially this you see it in edit mode here um it's completely white labels front and a back end and essentially this is where you're able to you have a few things here you can't think of
them as four major features the brand tune chat so this is where you have the different personas and you can have the access to the dobot yeah the brand stock controller where you can literally just generate websites and assets um we integrate with like figma and different things like that so we we
do think about what people want yeah the AI data engine where you can edit Pages um update create sections like sort it the way you want we have permissions and different layers as well and then of course the brand tune Imogen we just had a recent update really exciting stuff all the imagery
you kind of see in the back end was all generated with the image gen um but you can also generate 10 images at a time using our system but the important part is you can generate them using your brand colors and stuff like that I thought this was kind of fun to show
uh so if you asked chat GPT right about like who are you it really doesn't have a sense of positioning and then so that's where prompt engineering has become really really popular where people are basically injecting the context into a sentence which is we know not best practice if you try to do
that with the human but with the context you essentially get the exact answer that you want right he even knows the AI even knows who it's talking to which is kind of funny um here I thought I'd just like break down on just a few conversations really quickly that we can talk about
um like this is just a conversation that I had with the dobot and then we can actually have a conversation directly but I thought this was easier so we can actually see how it's responding essentially it's asking with uh follow-up questions that I answer um and then these are kind of like more
kind of honest you know late at night uh uh conversations that you can have with the AI put lead in that pouring your soul out to a computer yes yeah yeah like uh just so people know you know this is not a fear of mine you know this is a friend of yours
you're asking for a friend I'm asking for a friend um yeah but you can kind of see that it is empathetic and it's really focusing on continuing the conversation so I feed it more information uh eventually it has enough information where it'll give it it'll give an actual kind of advice based upon
everything that I've given it but you can see it really did try to understand the scenario before I gave this advice you can see it's like also referencing some of the books and different things like that that you've included as well uh this is another completely different use case but I wanted to
show it anyway in this case you know one of the things that you know I would love to be able to do is ask Christo at 9 00 pm at night like hey can you review my marketing copy on my website and I want him to respond immediately uh this is essentially what
I did with the dobot and then uh essentially they're able to give me really good feedback on how I can improve the marketing language on that page I just copy I paste it into it so you can kind of see how you can get personalized advice that way here um you know I
started talking a little bit about things along the lines of like limited beliefs and different things like that you can see how we have short quick follow-up questions where eventually the AI has Clarity around the scenario isn't able to give proper advice very specific to the user um again this is a we're
going to continue to refine this but this is just a few examples of how we can do things this is another one this on this slide we're seeing another question can you give me a plan that I can execute over the next eight hours to improve my sales something really specific maybe a
person has eight hours to spend over the weekend and they just want to dig into some sales uh here it gave kind of a plan hour by hour on how someone can kind of jump into sales and different things like that um so you can kind of see the there's a lot of
flexibility in how we can train the system we can also kind of go back and say Hey you know this is what's going to happen after this cause Chris is going to give me feedback on this writing and he's going to say you know I want for these types of scenarios I prefer
to respond like this and the Beautiful part is we take that we put into the platform um and then from now from that point on it will respond that way which is really exciting I had some other slides these are actually slides that I presented at Art Center last year um it was
probably like September or something like that um before like chat gpp came out uh so this I should this this slide would have been even crazier if I had a Chachi PT on there but kind of talks about this is AI is completely out of the bottle in the sense that it's progressing
extremely quickly um at show.ai we're kind of riding that wave but you can see this is a way that is exponential these are just the amount of research papers that basically correlate really closely to ai's capabilities and so this is a way that you want to watch uh ride or else it might
wipe you out if you if you don't write it um one of the things I'd like to talk I like to talk about too at that art center presentation was just hey like what where does that kind of leave us and then um when the reason why we created the platform was because
we really really know that AI needs human input and humans are really good at collecting Insight possession positioning to ask questions in the field and to see things that can feed the AI to make the brand or presence uh more competitive uh this is a funny I I was curious what your thoughts
were on CR on this slide Chris uh in terms of copying or stealing there's a lot of um for the image creation uh models right now there's a lot of controversy around this in terms of open open data sets and how it uh AI is using patterns and its understanding of the world
to create new imagery um um my my general understanding is that like it's kind of like the Apple uh this the Google with the search engine you know with Yahoo was the first ones with the search engines but they copied It remix and it made it their own they kind of Applied their
own ingredients but I'd be really curious to hear your thoughts on um uh what you think about Ai and the image Generation stuff right now in terms of like even copyright or or uh ethics yeah AI ethics is a is a tricky one it's probably one that's too meaty for us to like
fully cover but I'll give you my high level opinion on this okay I recently spoke to professors a professor who's like the direct what is it called the chair department chair of the master's program in one of this prestigious design schools and he's like I did the Deep dive AI is not doing
anything different than what humans do is it's just called looking and you look and then you make it your own and a lot of people I don't think they understand this the way that AI is trained on on copyrighted images is before it processes or looks at the image it's been diffused it's
that's why it's called stable diffusion it blurs the image and then it kind of starts to squint its eyes if you will and makes out certain patterns and in the very beginning eyes and hands and certain parts look really weird because you're still learning like a child seeing the world for the first
time a lot of people who don't understand how AI works it good because it looks so good it literally thinks it lifts a part of something and puts it into the other thing like you would uh a photo bash right so it's not actually using the pixels it's an approximation it's understanding of
it and maybe I have this wrong but if humans look at work and we're all looking at work all the time quite literally you're looking for inspiration and you're processing lots of things and you're doing the Deep look and then you go back and you create some things what are you looking for
and why would you look at it if not in some way to borrow elements assets or that spark and because you don't know how to explain your process and you can do it fairly slowly you forgive yourself you give yourself a pass that it's not copying but you are and there's nothing wrong
with that it there's two kinds of copyright that are that are infringement that really bother me one is the blatant lifting of assets and images that don't belong to you and you use without paying for nor crediting the person who did it and that kind of copying is unforgivable or should come at
a heavy cost but if you if you look at a poster from Saul Bass and you like the graphic quality of it and the way that he renders images in a very simplistic form and you go out and do something that's very graphic and render something in a very simplistic form I don't
believe you're turning yourself into authorities if you use colors inspired by somebody you're not turning yourself in nobody is but because what's happening is non-artist non-quote unquote creative people are using tools like this and doing work at scale that rival your technical skill it's a very threatening gesture and I can see why
people are really concerned but this is not for me to decide and to be for the courts to figure out but the word of caution I would throw out to everybody is don't wait for the courts to figure this out you need to be playing with the tools integrating some of these things
into your workflow and you draw the line for yourself but don't artificially handicap yourself it's like to say you want to do photo retouching but you're not going to use any software to do it you don't want to use Photoshop you don't want to use illustrator you'll only use any 3D image making
program because you want to do it the old way you want to build real models in real life and you want to light things and that's how people used to do it you're putting yourself at a distinct disadvantage that's my take on it show what's yours oh beautifully said Chris no that's that's
absolutely my take and then um how you described how the AI actually thinks through this imagery um in many ways it actually you know I think people that actually use it kind of understand that it reflects our process too and then by utilizing the tool we actually do realize oh I am getting
better this is a craft like anything else and then um my my expertise my knowledge does translate um so I'm really excited to hear that you kind of a similar view on how to approach this uh I thought this was an interesting slide as well yeah so this is from our image generator
um this was just a it's a really simple prompt of a turtle this is June last year uh you see it's like a cartoon like child imagery image right you see in November 22 okay starting to get really good uh now uh this this year you're getting this incredible imagery so this is
one of the things in when we talk about progress you know it's even with the language models the language models are going through the same level of progress which is it's easier to see with imagery I also want to talk about fine-tune imagery this is something that we wanted to do with You
Chris um we actually uh we tried uh using some of your imagery to create a fine tune model we just had to get more of it just some higher quality imagery but we'll be able to create some really cool cool things but this is kind of where we see like the future of
our platform and what we're doing is just the continuous like customization of these tools and then making sure that we're getting the most out of these models and if that means fine-tuning it awesome if that means that we have to create a Playbook that an AI can read and understand so it can
follow guidelines that's what we build as as well that's the end of the slides okay and so we can we can talk about other things I'd be happy to show it we can show the platform live too I think we can probably tease them maybe with one quick example of how dobot works
and so we need to start with a really good question something that somebody is really thinking about do you have something ready to go show yeah so we were able to add the books uh and some of the video references uh to TED talks and different things like that um it's worth uh
definitely we as we talked about podcasts as well and I think that would make a huge difference to be honest with you because right now the business bootcamp material outweighs more q a type information and so the AI will have a tendency to kind of say oh most likely this is business boot
camp course material that I need to create uh so I'm really excited to see where the podcast stick is a completely different level but this this should significantly outperform Tuesday okay already let's try something and you know what I have a prompt for you to think about oftentimes I want to write a
talk or I want to do a workshop and I need help in formulating the title the description and a course outline and designing some exercises I don't want a total failure here so let's let's pick something that you think you can excel at given some of the prompts or the ideas I've just
shared with you oh in terms of like creating courses yeah like things like that I want to write a talk or I want to teach a course and I need to come up with a title the outline the hook whatever it is and perhaps some of the exercises you don't have to do
all that I just want you to pick something in that General space because this is something I do all the time so you type and I'll narrate for you so that you don't have to walk and chew gum at the same time okay so Show's typing in into the interface because it begins
with hey I'm here to help what's on your mind Show's typing in what are you typing in I need help writing a talk on generative AI should we do Publishers or should we do promotion agents pick whatever should we do promotion okay whatever for emotion agencies so now dobot responds absolutely show let's
start by getting a clear understanding of your audience who will be attending this talk are they familiar with generative AI uh or is this A New Concept for them understanding your audience will help us tailor talk to their needs and knowledge level all right so I wrote uh people uh people in the
motion industry typically agency owners they're familiar but haven't actually applied AI into their process sorry about the grammar everybody what's great about uh robots is to kind of figure out the context anyway so your grammar doesn't have to be perfect your spelling could be off you can misquote somebody it kind of already
knows because it has a lot of context within the larger framework of its data set all right so now it says great sounds like your audience has a basic understanding of AI but they haven't yet integrated into their workflows when crafting your talk consider focusing on the following points and I'm not going
to read everything because it's too much what is generative AI the potential of AI and Motion Graphics real world examples how to implement AI addressing concerns that sounds like a pretty good framework so far now what are you gonna do does this give you and then does it give you a good starting
point for your talk and I say yes that is great uh can you help me write the script it's really writing the whole thing out here right now sorry about that well you asked the two so here we go yeah I did I didn't okay let's scroll back to the top yes and
I'll read the ai's response here it says basically God help you to craft your script for your talk and here's a rough outline there's introductions start with a brief personal introduction and your connection to this topic share a startling fact a short story or a provocative question about generative AI to grab the
audience's attention State the purpose of your talk to explore the potential generative AI in the motion industry and then it goes with the body and the conclusion it gets sub or bullet points to each one of the main things so it follows a very classic structure if you've never given a talk before
this is actually quite excellent this is things I talk to people about all the time and so where do we go with this now so uh does this structure work for you if so we can start filling in each section with more detail well let's let's try should we go ahead and yeah
say yes let's go proceed yes let's let's go it's describing the title it's also telling you what each slide has and so we'll see if this is actually any good so I can now read this and show me this bigger it says okay let's dive in the title of this talk is called
embracing the AI Revolution the new era for motion agencies not a bad title and slide one good evening everyone I'm here today to talk about a topic that's been making waves in our industry generative AI now I know what you're thinking AI That's for Tech Giants and data scientists right what does it
have to do with us the creatives the storytellers well as it turns out quite a lot AI isn't just about crunching numbers and automating tasks it's about augmenting our creative capabilities pushing the boundaries of what we can imagine and Achieve so slide four generative AI in particular is a game changer it's a
type of AI that can create content from images to music to text and yes even Motion Graphics Slide Five imagine having an AI assistant that can generate hundreds of design variations in the time it takes you to make a cup of coffee whoa or an AI that can animate complex scenes with just
a few simple commands slide six but it's not just about speed and efficiency it's about exploration and Innovation with AI we can explore design spaces that would be impossible for a human to navigate alone we can discover new Aesthetics new narratives new ways of engaging our audience now I'm not going to read
the rest but again if you were a novice speech writer and and you didn't know how to give a keynote on this stuff this is not bad at all yes and then if you if you want another one you can just click refresh and then it'll it'll give you another draft it's it's
that simple so the iteration process is really fun and then you can go back in and say hey use this quote from this person or uh use this video script or something like that to influence the script yeah okay it's it's quite fun yeah all right while that's still crunching through and writing
shows next talk I think this is a good opportunity for us to kind of split away from this end this part of it and land this AI plane if you will via mid Journey or some other video based AI generative tool okay show um how do we how do we end this conversation
what do we want people to know about you and what you're doing and how they can possibly interact with you post this discussion yeah absolutely so um if you are if anybody is interested in training AI we've been spending the last six years essentially training it on different brands and different individuals to
behave the way that they want and to really help them get more things done um if if you're interested in uh signing up you can go to our website www.show.ai where you can learn more about us and get started but we're really here to kind of hand hold you through the process of
building your best AI uh so you don't have to worry about mitigating any kind of risk or other things associated with it we have a beautiful platform and a piece of software that we're the first a generative AI platform in the world and we can't wait to see what people like Chris and
yourself uh can do with it it's exciting times wonderful so if you can be curious long enough to explore these tools I think you're going to be as excited as I am and I want to end it on this um reflection I I was at a conference in Newcastle and I was hanging
out with Mr Daniel Priestley and of course the question came up is like how do you think AI is going to impact our industry and I believe there are a bunch of marketers in the room and he says without a doubt and he says very confidently AI is going to be the source
of generating billions of new money for people who embrace it it's just the only problem is this money is not going to be distributed evenly you need to think about that everybody so when we flash forward to the future a few years from now we look back all at all the new Millionaires
and billionaires who've incorporated or developed tools for or and somehow have added layers to the tools that we already used today are you going to count yourself amongst one of them or on the other side of that Financial equation I'm not here to tell you to be scared or to be afraid but
I'm here to encourage you hopefully inspire you and to open your minds a little bit about what you can do if you just stay curious with that I think this is a perfect ending to this conversation by AI show thanks for doing this thanks for inspiring me personally on my own journey into
mid-journey and what I think will be one of the most powerful tools that you're creating for us which is to be able to scale my time my knowledge across multiple platforms and to make it more accessible I can't wait for this version to be out there in the wild and working but also
developing different models and training different things for all the needs that I need as an educator thanks very much so thank you so much Chris really appreciate you and really looking forward to embracing the future with you
Enjoyed this? There’s more where it came from.
Get insights on creativity, business, and design from The Futur.
You can unsubscribe anytime. By submitting, you agree to receive communications and to our Privacy Policy.
“Don't artificially handicap yourself.”
— Chris Do
Key Takeaways
Subscribe
Details
- Length
- 1h 5m
- Views
- 13,293
- Topic
- ENTREPRENEURSHIP