As a large language model trained by OpenAI, ChatGPT has been designed to understand and generate human-like text based on a massive amount of data. From writing creative stories and poetry to answering complex questions and providing personalized recommendations, ChatGPT seems to be running on all cylinders, and still accelerating.
The possibility of ChatGPT extends beyond text, via Voice AI. John Miller, Co-Founder of Launchvox stops by to about ChatGPT’s potential, mainstream adoption, ethics, and security implications through the lens of a creator.
TIMESTAMPS
0:03:31 – Reflection on the Artistic Experience of the Renaissance Era
0:07:51 – War and Innovation in the 19th Century
0:12:50 – Ransomware and the Russian Revolution
0:14:16 – The Evolving Landscape of Security and Hacking
0:18:03 – The Future of Cybersecurity and the Use of Chat GPT and Resemble AI
0:19:53 – Voice Synthesis and Audio Processing
0:21:10 – Utilizing Virtual Voice Actors and Mastering Processes for Content Production
0:23:15 – Exploring the Ethical Implications of AI-Generated Voice Replication in the Entertainment Industry
0:27:50 – Security and Ethics in AI Development
0:30:52 – The Impact of Chat GPT on Business and Personal Use Cases
0:36:04 – The Need for Accuracy in Fact Checking with Chat GPT
0:38:03 – AI and Immersive Technologies with Launchvox
SYMLINKS
Launchvox
LinkedIn – Launchvox
LinkedIn – John Miller
Stable Diffusion
ChatGPT
Getty | Stable Diffusion Lawsuit
Resemble
Adobe Audition
DRINK INSTRUCTION
pic
EPISODE SPONSOR
N/A
CONNECT WITH US
http://www.barcodesecurity.com
Become a Sponsor
Follow us on LinkedIn
Tweet us at @BarCodeSecurity
Email us at info@barcodesecurity.com
This episode has been automatically transcribed by AI, please excuse any typos or grammatical errors.
Chris: I’m here with a very special guest today who is known to have an extreme amount of power in the industry, unmatched expertise, and life experience that spans like none other. Welcome and thanks for stopping by BarCode. You know, I’d like to first ask you about generative artificial intelligence. We are really seeing the power of generative AI with the emergence of programs like Stable Diffusion and ChatGPT.
Chris: But we can’t hide from the fact that it attempts to mirror the element of human creativity. I want to take it back to the Renaissance era. During that time, human artists used their own skill to create masterpieces that were renowned for their realism, technical skill, and beauty. As an artist yourself during that time, I’m curious if you can, using words, draw us your vision of what that experience was like for you.
Renaissance Artist: I, an artist of humble beginnings, take quill in hand to relate the story of my life during the grand age of the Renaissance. From a young age, I was enamored with the beauty of nature and the human form. I would spend hours sketching the flowers in the fields and the faces of my fellow villagers. As I grew older, I knew that I must journey to the great cities of Florence and Rome to hone my craft and learn from the masters.
Renaissance Artist: I was fortunate enough to find a kind patron who saw the promise in my work and provided for my apprenticeship under the cheetahlage of a renowned painter. I spent many years studying the techniques of the old masters and experimenting with new methods of my own. But the path of an artist is not without its trials and tribulations. I struggled to find commissions and secure my livelihood, and like many of my fellow artists, I had to contend with the demands of wealthy patrons and the church, who oftentimes sought to impose their own ideals upon my work.
Renaissance Artist: Despite these difficulties, I found great joy in my craft. The act of creating, of bringing to life the visions in my mind was a source of immense satisfaction. And to see my work hanging in the grand halls of the medici Palace or the sistine Chapel was a dream come true. I was also blessed to live in a time of great cultural and intellectual awakening. The world was full of new ideas and discoveries, and as an artist, I was at the forefront of this movement.
Renaissance Artist: I can say with certainty that my life as an artist during the Renaissance was one of both hardship and wonder. But above all, it was a life well lived.
Chris: That’s interesting. Thanks for sharing that. And congrats on the massive success that you’ve had while paving the way for other artists that have followed you. How would you compare artists of your time with cutting edge gen-AI technology?
Renaissance Artist: Oh, my good sir, let me take a moment to compare the grand artistry of the Renaissance with the use of generative AI. In terms of the creative process, we renaissance artists honed our skills through long years of apprenticeship and study, imbuing each stroke of the brush with our own imagination and vision. Generative AI, on the other hand, creates its art through the use of algorithms without the touch of human hands guiding the brush.
Renaissance Artist: In terms of style, our art was influenced by classical ideals and realism, a pursuit of capturing the world in its true form. Generative AI art can take on many shapes and forms, from abstract to lifelike, but it lacks the deliberate and conscious hand that guided us in our own creations. You see, my friend, the comparison between the art of the Renaissance and generative AI is one of human skill and imagination versus the diversity of the machine.
Renaissance Artist: Each has its own merits, but there is a timeless quality to the art created by the hand of a master that cannot be replicated by any algorithm.
Chris: Valid point. Yeah, it’s going to be interesting to see the development of generative AI art and how it evolves. Okay, so then you decided to leave the world of art to become a king, which is quite a career change. What’s life like living like a true king?
1600 King: In the year 1600 verily, my life is one of grandeur and opulence. My palace tis a grander boat with many chambers and furnishings most fine. I am blessed with luxuries aplenty, from rich apparel to art and jewels most precious. As king, I wield great political power wherewith I make laws and decisions for mine kingdom.
Chris: Very interesting. As I’m in the cybersecurity field, I’m curious about your positioning on safeguarding your critical assets. Tell me about the type of security controls deployed in your fortress.
1600 King: Methinks protecting my castle is a voters dimaport. My guards, they are ever vigilant, stationed at the gates and upon the walls. Strong gates of iron do we keep to keep out unwanted intruders. My men at arms, they are well trained in battle, ready to defend the castle at a moment’s notice. And should an enemy dare to breach our walls, a moat filled with water that’s encircled a castle making attack most difficult.
1600 King: I sleep soundly knowing mine castle is well protected.
Chris: Nice. So you employ a 24/7 SOC! Monitoring is key because threats don’t sleep. You know that. So let’s shift to cyberwar. There have been many major wars in the history of the world that we can learn from, including the revolutionary war in the 18th century. As someone who has served on the front line of battle, tell me what that grueling experience was actually like for you.
Revolutionary War Vet: The war. It was a brutal and harsh affair. We marched through scorching hot summers and freezing cold winters, always on the move, never knowing when the next battle would come. We fought against redcoats, the likes of which were well trained and well equipped. Our own supplies were often scarce, but we persevered, always pushing forward for the sake of our cause. The gunfire and cannon blasts were constant and the smell of gunpowder never left our nostrils.
Revolutionary War Vet: The loss of our comrades was a heavy burden to bear, but we fought on, determined to secure our freedom.
Chris: Damn! Definitely a lot of lessons there to be drawn from that in comparison to cyberwar. How would you describe what you experienced in war to how one should react in cyber warfare?
Revolutionary War Vet: I must say the war I fought in it were a far cry from the cyber warfare of today. Back in my time, we fought with muskets and bayonets on foot and horseback. The battles were face to face and the loss of life was all too real. Nowadays I hear tell of wars fought with computers and code, where the enemy is unseen and the battles take place in a virtual world. I cannot fathom it, truly. The idea of fighting without ever seeing the enemy’s face or feeling the weight of battle tis a strange thing indeed.
Chris: Yeah, very interesting perspective. So we’re coming out of a pandemic at this time, so we’re seeing a lot of changes. Discovery and invention, it no doubt is a significant time in history, similar to what you witnessed in the 19th century. Let me ask you this, what was that transformation like for you personally?
19th Century Inventor: The era in which we live, it is one of greater progress and innovation. The world is ever changing and advancing with new inventions and discoveries being made every day. The steam engine, the telegraph, the telephone all these have revolutionized the way we live and work. The factories, they are now powered by machines and run with efficiency like never before. The transportation, it has also been transformed, with trains and steamships now connecting the far corners of the globe.
19th Century Inventor: Truly, the world has never seen such rapid progress. And as an inventor, I am excited to be a part of it all.
Chris: Yeah man, that’s amazing. And people tend to fear new technology, and I think it’s just an evolution that will continue regardless of your personal stance. How did people react to your own innovation during that era?
19th Century Inventor: Society, they were both fearful and excited by the changes and innovations of our time. Many saw the advancements in technology and industry as a threat to their way of life and feared for their jobs and livelihoods. But many others, they embraced the new opportunities and possibilities that came with their progress. The factories, they brought new jobs and ways of working. The transportation, it brought people and goods together like never before.
19th Century Inventor: And for us inventors, it was a time of great creativity and experimentation where anything seemed possible. It was a time of great change and society had to adapt to it.
Chris: Yeah, that’s definitely a great takeaway and something I think that we should all learn from. So let’s talk about an epidemic that continues to plague us. Ransomware! In fact, I recently read that 74% of all ransomware revenue goes to Russian linked hackers. Although the Russian government officially denies any involvement in these type of activities, many of the ransomware groups that have been responsible for high profile attacks are believed to have links to Russia, either through their origins or through the use of Russian language ransomware strains.
Chris: As someone who has lived through the Russian Revolution, I’m curious how the country was during that time.
Russian Revolutionist: Conditions in Russia, they were most dire. The economy, it was in shambles and the people were suffering greatly. Food and supplies, they were in short supply. And the streets were filled with bedders and the homeless. The tsar, he was out of touch with the needs of his people. And his government was corrupt and ineffective. The war that had only made things worse. leaking young men away to fight and leaving their families without support.
Russian Revolutionist: The mood of the people, it was one of anger and frustration. And many were calling for change and revolution.
Chris: Damn. Well, as you know, the 1980s brought an increase in high profile cyberattacks, including those at National CSS, AT&T, and the Los alamos National Laboratory. The movie War Games, which was released in 83, was actually about a rogue computer program that takes over a nuclear missile system under the form of a game. Although in 1987, you saw the inception of commercial antivirus. Tell me about that.
80s Dude: Back in the day, the Atari SD was the bomb. But so were viruses, man. So this dude, Mark Minazi, created the ultimate solution. The first antivirus product for the Atari SD called the Ultimate Virus Killer, or UVK for short. It was the gnarly way to protect your Atari SD from those nasty bugs. All the cool kids were using it to keep their atari SDs running sweet. It was a total game changer, dude.
Chris: Totally, man. And today, we have our own game changers, aka video Game Hackers. If you were to hack a video game in today’s world, how would you go about that?
80s Dude: Yo, check it out. If a bro wants to jack into a video game and tweak the code, they got to find a glitch in the system. Maybe they reverse engineer the game’s software to find out how it ticks. Or they find a security hole in the servers or network and exploit it. Maybe they even use third party tools or mods that give them an unfair advantage. But here’s the thing, bro. hacking a video game is totally against the rules and can get you in deep doo doo. It’s like, not cool, man.
80s Dude: It ruins the game for everyone else. And you could even get banned or have your account suspended. And don’t even get me started on the security risk, dude. It’s like you could be opening up your machine and all your personal info to some seriously bad dudes. So what I’m saying is, stay on the straight and narrow, bro. Play the game fair and square and you’ll have a way better time. Trust me. Or maybe not. And you like being an asshole.
Chris: Dope advice. Speaking of technology game changers, both AI technology and evolving adversaries keep us security professionals on high alert. As a futurist, how do you see the state of cybersecurity today and what do you predict for tomorrow?
Futurist: I’m sorry. My knowledge cutoff is 2021, and as such, I cannot provide an accurate answer about the state of cybersecurity in 2023. However, I can tell you that in 2021, cybersecurity continues to be a major concern for individuals, businesses, and governments around the world. Advanced persistent threats, nation state hacking, and ransomware attacks are among the most significant security challenges facing organizations today.
Futurist: Additionally, the increasing use of cloud computing, IoT devices, and mobile devices has created new attack vectors for cyber criminals to exploit. As technology continues to evolve, the field of cybersecurity will also continue to evolve to keep pace with new threats and to protect individuals and organizations from cyber attacks.
Chris: Oh, no doubt. So what do you predict a future state of security to look like?
Futurist: It is likely that cybersecurity will continue to evolve and become more sophisticated in the future as technology advances and cyber threats become more complex. Some of the major trends that are likely to shape the future of cybersecurity include one, artificial intelligence and machine learning will play a bigger role in cybersecurity. Two, IoT will continue to grow, and with it, the number of connected devices that require protection.
Futurist: Three, cloud security will become more important as more organizations move their data and applications to the cloud. Four, quantum computing will bring new challenges and opportunities in the field of cybersecurity, with the ability to process vast amount of data that will be used to break encryption and also to develop new encryption methods. Five, cyberphysical systems, which are systems that combine physical and digital components, will become more prevalent.
Futurist: Cybersecurity for these systems will be critical to ensure the safety and reliability of critical infrastructure such as power grids and transportation systems. Six, virtual and augmented reality technology will become more pervasive, and with it, new security challenges will arise. Overall, it is expected that the future of cybersecurity will be characterized by an increased emphasis on proactive and predictive security measures, as well as a greater focus on protecting sensitive data and critical infrastructure.
Chris: I see. So with that…. <Glitch sound> Oh, man! John Miller, co founder at Launchvox is here! What’s up, John?
John: How’s it going, man? Thanks for having me. I’m glad to be here.
Chris: Hey, it’s good to see you, man. So the conversation that I just had was a result of ChatGPT and your expertise. I essentially fed interview text that I created into ChatGPT. Then, you took that output and applied a very realistic voice to match the dialect of the respondent. First, talk to me about the tech that you used and the approach that you took to create that.
John: So right away, I just want to say chat gpt the responses that you were getting from the questions, I was impressed with just on the level of we’re at the blend of how human and computer just sound like realistic conversations, and you chose the areas that you are wanting to go through. So the challenge for me was like, okay, how can we get a Renaissance voice? Or, how can we get a dude surfer voice?
John: Or how can we get, like, a Morgan freeman? All of these are very different, but the questions that you were having answered felt like they needed to have that conversational output. So I did some research. There’s a few out there, but the one that I ended up landing on was this company called resembles AI. And what they specialize in is making either A, you can clone your own voice. They really don’t need that much samples. I think about 20 different inputs of you reading lines they provide, and they take those samples and make a voice from them. And from what I’ve seen, they’re pretty good. They just can’t do draws or inflection points too heavily.
John: Then on the other side, they have what’s called a marketplace, which you pay into, like, zero usage of render time. And by the end of it, it’s not that much. So I took the content that you were giving and I just fed it into the marketplace voices they had. And for the Renaissance, it was pretty easy to find this guy. And I just was like, hey, I want you to give me the response in this kind of direction where you can go through and say highlight the words that you were saying and your answers like, I want there to be more emphasis here or I want you to have emotion here that you’re angry or you’re annoyed and you can go through and direct them.
John: It’s not one to one. Like, some of it, like, the direction you would say, I want you to be happy here. And it was literally just like, pitch and tone sped up, so I’d have to go and remaster that. But then after I got all those voices right, I then took that through adobe audition, and I added base to it, or I slowed it down or sped it up based on the flavor of the taste that I was looking for in more direction, which is really what a sound studio does. Now.
John: They have a voice actor come in, they do their part, they process it, they master it, and they send it out. So what we’re doing is effectively using a virtual voice actor, but then still applying that same kind of mastering process at the end of it to give you that flavor. And then that was really what I gave you. I think we were talking earlier maybe about 17 minutes of content was produced. And I think between each clip, I think that took about maybe 20 to 30 minutes. And then overall, probably, I don’t know, about a half day of work.
John: But what I didn’t have to deal with was, like, setting up a meet and greet of a voice actor. I didn’t have to go through the phases of vetting talent. I didn’t have to go through the phase of really making sure, like, hey, you’re not understanding this thing I’m trying to do. You’re just not nailing it. I kind of avoided the rift wrap that I see when working with other voice actors, which is a little bit of a negative. But what I’m not getting is a little bit of that personality that you miss. Right. Like, you’re not working with someone personable. And we’ve got a handful of talent that I’ve just loved, grown to love over the years.
John: So there’s that up and down, that draw. But the technology that I really used was Resemble and Audition and then just the flavoring in between.
Chris: Nice.
John: Yeah.
Chris: And I think the differentiator there was the emphasis that you put on it. It didn’t sound like your standard text to speech tool, which is typically very robotic. There were sections where it sounded nearly flawless to me, especially Morgan Freeman. Morgan Freeman was on point.
John: Yeah, the Morgan freeman was crazy for sure. The dude Rick, too. Like, he’s basically sounds like the Rick from Rick and Morty. The only time he would break down is when I was taking the questions, the answers you were giving me, and I would give it to him. And it had perfect punctuation in there that would break down for the AI. Because Rick doesn’t really have inflections where he stops. He’s just like a continuous stream of narration.
John: So I had to go and break your punctuation, which was written perfectly so that it would be conversational from ricky. So when he would say like, yeah, bro, we did. I took it out and be like, yeah, bro, then it was clear. Those are, like, the nuances that you end up find out. where’s Morgan Freeman? I just gave it to him. I’m actually the professional actor on this one. I just want you to know I’m not actually AI.
John: Okay, cool. Let’s just nail it. It was great.
Chris: Yeah. So, ironically, Justin Roiland, the show’s creator and voice of Rick from Rick and Morty, was just dropped by hulu and Adult Swim over charges of domestic violence, which we’re not going to get into, but it does raise an interesting situation. So gathering a large data set of voice files from someone like that is definitely obtainable. We obviously just saw the possibility to train a model to produce a very close, near, one to one replica of an exact voice.
Chris: Can you see this tech impacting the future of the entertainment industry with the potential of eliminating voice actors or narrators or even voiceovers in gaming?
John: I can’t see it. Not honestly, because most of these companies are businesses. And if you talk about a business, it’s looking at a bottom line. Right now there’s going to be companies that come out and they’re going to say we are human only and that will be great for them. But there’s also going to be companies that say we don’t want to deal with the negativities of working with sexual abuse people or we don’t want to deal with the fallout of people that just can’t, from our point of view, like they’ve got the talent but they don’t have the professionalism.
John: Is there something negative there? Is there something that ethical there that is to be decided by the consumer? Right? Because at the end of the day, it’s going to be the consumers that watch this content, that listen to it. And if they’re buying into it knowing that it’s AI bashed, then they’re telling everyone they don’t have a problem with it. So businesses are going to see that bottom line and they’re going to say, okay, we just got 30 hours of raw voice talent from Justin Roiland who is no longer someone we want to work with.
John: Now we could go out and find someone else and they may sound like it, or we can work with sound engineers and AI to produce almost a near flawless replication of that and we continue on with the show. And to that end, I don’t necessarily see how a business wouldn’t make that choice. Especially if you’re just coming from like we’re just experimenting right now to see how it goes. But if you’re consuming and your audience is loving it, it’s not like the show creator walked out, right? It’s still being written and it still has a team.
John: You just lost some of the soul of a guy that may, depending on what the jury decides, have been a bad dude or a good dude. We don’t know. That’s not the point of the conversation is that that voice is no longer there. Now, if you wanted to look at about it the other way, let’s say like Michael J. Fox, right, who wanted to continue acting, wanted to be the aaa star he was going to be, but he couldn’t, right? Like he has parkinson’s and he could now embody though content, voice content.
John: That his abilities. So like nvidia for example, and this is going to bit away from audio and focusing more on video, but nvidia for example, now has content where if I were to look away from the camera, if I were to look down or do whatever, it would just make it so that I was always looking straight ahead. They also have other capabilities where they could deep fake and to say like and everyone has seen deep fakes at this point, but they can be self trained ethically from someone who gives permission.
John: And if that person still wanted to be part of the I want to act, but I can’t do everything well we have technology that will kind of scrub out the imperfections of what parkinson’s might have prevented you from doing properly. I think there’s something ethically there. There’s accessibility there. It’s giving people that didn’t have the opportunities before an opportunity, especially if they’re loved and they can act well.
John: So, yeah, I definitely see going back to your original question, that why would businesses walk away from that?
Chris: Yeah, exactly. So I’d like to hit on the security aspect. I feel like I’m hard coded to always look first at the nefarious ways an attacker could use tech or use AI to negatively impact society. So I’m curious, as a creator, how much focus is placed on security when you’re going through the development cycle for similar systems?
John: That’s a great question. That is always being developed more and more, especially the clients we work with. Launchvox, who I didn’t really introduce the company that I co founded, we work primarily with a lot of medical clients in the animation and creative industries, and a lot of those clients have HIPAA standards. You can’t HIPAA compliance. There’s only so much you can do with patient data. There’s only so much you can do with information that might embroider on taking someone’s info and giving it out in a way that they didn’t agree to. Right. So with deepfakes, there’s a fringe there that could be equally dangerous but also equally wholesome.
John: It’s about how the tool is wielded at the end of the day. And we as a society have shown that we are willing to be pretty nefarious with pretty much anything that gives us the capabilities of doing so. As a studio, anytime we’re working with content like that, we always get permission. We try to be ethical. There’s always gray standards, like what we didn’t know that we didn’t know going into it. But the intention is always to be ethical, and we see a lot of that in data training as well.
John: How did mid journey everyone is hearing about like, they scraped their data? How did stable diffusion get theirs through the Lyon data set? I know that Getty is now suing stable stability AI for $1.8 trillion. So there’s an entire industry here that needs to be debated and precedence needs to be set almost to the point there needs to be like an AI bill of Rights. What privacy can humanity expect? There’s something we were talking about internally recently, about Sharon Tin, where one parent shares the pictures of their child online and then that’s their impetuity. But the child didn’t have any choice in that. And that can then be taken and used for things they had no permission of. Or years later, they see their face on porn that isn’t them, but it’s deep faked on there. These are things that there’s not really a way to massively at scale combat it.
John: I have no idea what the solution is. I know that we are getting to a ground zero of some turmoil.
Chris: Do you feel like there should be regulation?
John: Yes, absolutely. It should be. What that regulation looks like, I’m not entirely certain, but this is getting to a point where this can’t just be open and not really barriers of entry. There has to be some kind of protection for the consumer and what’s really nice. And I know that it’s been moving so fast that people haven’t been able to get ahead of it, but a lot of what I’m seeing is a positive discourse where people that are using artists primarily aren’t turning around being like, well, too bad, bro it’s out, let’s do this. They are reasons like, well, I want to use this technology and I want it to be ethical, so I will use the ones that are trained ethically. Like there are different ways of getting the same result without having it infringe on people’s rights.
John: I think it’s positive, but I don’t even think we’ve seen the worst yet.
Chris: Yeah. And how do you feel like ChatGPT has just changed the game for us, right? I mean, you got a mainstream audience running with ChatGPT now and continuing to use it. So do you feel like that technology has changed things for the better or for the worse? Or do you feel like it just comes down to what it’s used for?
John: I think it comes down to what it’s used for. Where I think we’re at is an inflection point to say if you think about where apps were, right, when the marketplace came out, like the iphone and how people started to consume their use of executables, right, it used to just be locked to a PC. Then it became something that followed you everywhere and a lot of people like one of the first use cases I saw that just clicked for people was language translation.
John: To say you’re in another place, I want to say this, and it was like all in the Apple ads, but it was a great piece of marketing because it showed you how you can what used to be a book and how you had to move around. You now have a way of having almost real time feedback between two people. Now, if you look at where we’re at with ChatGPT and vocal synthesizers, that same process cannot be instantaneous. It could be listening to two people talking and then translating real time. And if I had my ear pods in, that would literally take off transparency and noise canceling and I was only hearing what was coming through, I almost wouldn’t even hear the original person talking so that I wouldn’t have that language barrier or that confusion of the other dialogue coming through.
John: Those are two use cases to show like that inflection point can be used in a positive way. If you move further than that, though, it can also start to augment everything you do. And I say that. In an accessible way. There was something that I saw on LinkedIn recently about someone’s child has autism and they ask questions that can never be answered. But ChatGPT could enter the conversation and be an infinite loop, offering infinite solutions to the same question.
John: And that whether or not that kid or that adult with autism would feel that connection or that was answering it, it was still showing you there are ways of handling welfare and the mind of I guess you could say I’m firing on the word, but just general welfare of who you are mentally can be augmented. Now, if we were to step out of that and talk about, like, what about business? What about what can you do through the business?
John: We write stories for our clients medical. We call them rationale for Use videos, where they say, like, hey, we need you to kind of come up with a marketing collaboration to tell our clientele why they should use this and how to use it. So we generally become pseudo experts. We dive into their white papers. We really try to understand what it is and what their goal is and who their audience is. We turn that around, write a script, shoot it back their way, and we have feedback on it.
John: That is now, Jet gpt has allowed us to have a much smarter advocate in your court to say, like, oh, I’m going to write the script. Sometimes you can even say like, hey, write me an outline and then I’ll augment that. But in a lot of cases, we can now write a script and then pass it back to ChatGPT and say, what? Could I make more conversational in a demographic that’s surgeons in this paragraph? And it would give you a very formalized response because it has Cambridge data, it has medical data, it has lawyer data, it has regulation data, it has everything it needs to know, what it should stay away from and what it should really push in a way that you just told it in a small training simulation.
John: This is what I’m looking for and this is what I want you to do. Now, here’s where it breaks down for me. And this is something that we’ll see job positions grow from, like the way horse and carriers went, then cars came out and we had mechanics. I see there is going to be accuracy in fact checkers because the confidence of ChatGPT when it’s wrong, gives you no indication that it’s wrong or inaccurate or that the information was a fallacy.
John: You could ask it a question. I’ve seen this first hand and it’ll do math. One of the best examples I can give is recently I was trying to visualize how much in a volume would it take of water to fill all the drought in California, because California has all their data as far as all the reservoirs, how much it would take to fill them back up to normal. So I grabbed that number, it was like million acre feet, right? And I gave that to ChatGPT and I said, okay, now tell me in a cube what this volume would be.
John: And I gave it the number that I had and it turned it around and it said, okay, in a volume, it would be 57 cubic feet. What was crazy, it was like it broke the math down perfectly and said, okay, if you have this many acre feet, this is how you need to get in the volume. Here’s the math formula to do it, and here’s the answer which is completely incorrect. And then we told it, we were like, hey, the actual answer to what you’re looking for is this. And it said, you’re right.
John: That is the answer. It is incorrect answer. Again, we tried it one more time and it still didn’t correct itself. And it wasn’t using the same number it had before. It came up with a completely different number. So if people are using that online, for example, to say, like with bing, which just we’re going to start to see that now with ChatGPT integrated into the Edge browser, people might go and be like, hey, my son was just stung by a bee, what should I do? It’s going to be like, you should get a snake and it should bite that wound. And the venom between the two will cancel each other out.
John: Now, that’s a really extreme result, but it will tell you information confidently and people might react to that. People might go and do something completely terrible because of that advice.
Chris: Yeah, it could be dangerous. Now, I do want to ask you about Launchvox. Can you talk to us about what your company does, the projects you’re working on, and any future plans in the AI and immersive space?
John: Yeah, absolutely. So real quick about launchbox. We’re an animation xr studio with a lot of experience and just experiential applications. What that might mean is that primarily, let’s talk about animation. We have clients come to us and they need marketing videos, they need medical animations or even industrial animations about their products, or they have a product lunch. We develop those working alongside doctors or experts in their field to develop two or three minute pieces of content that they can either put on their website marketing show to a congress if they need some people that deal with law to understand more about how their product is going to be FDA regulated. So those are things that we work on that’s like what we’ve cut our teeth on.
John: And then there’s XR. It’s kind of taking those same applications and putting them in an experience you put your headset on. But we’re starting to see a lot of now is soft skill training. So someone that might be like, well, we’re trying to teach diversity and how we can make that a little bit more standard in a training facility. Rather than just having someone read the book, like, put them in an actual setting, have them go through the curriculum, but put them in a real scenario where they deal with those people and then we could see real time what they’re doing wrong and how to engage with that.
John: So soft skills training, huge break for us. We’re really seeing it grow now. We’re really diving into AI. We’re building a chat bot. We’ve been working on that pretty heavily. We hope to have something rolled out in the next month or two. And that’s going to be something we’re really the most excited about because that can go in many places. A sales trainer, a certification trainer, really, because all the content that we work with, with the clients, it’s all there, it’s all been written out, it’s all been drafted.
John: We can just now take that and feed it as a model and data set. So those are really awesome ways. Outside of that, all of us are artists and technical creatives. We enjoy the work we do and we try to put a really high quality content when we can.
Chris: Love it, man. Where can our listeners find you and connect with you and Launchvox online.
John: Absolutely. So we have our website, Launchvox.com. Or you can catch us on LinkedIn. I’m John Miller. That’s spelled exactly how it sounds. Our CEO, Bob dice and Ian brigitte. You can find them on LinkedIn as well. We are a small company, been around for almost two years now. We hope to be predominantly, like, creatives all the way through, but you never know.
Chris: Nice. And where are you based out of?
John: Sacramento, California.
Chris: Okay, so tell me a good bar there. Where do you go relax after a long day?
John: Yeah. So there’s a couple of places downtown. One of the favorites is The Republic. It’s like a game pub. And then there’s a really nice place, although actually it’s not around anymore. It’s called squeeze in. It’s a pretty nice place. You get some good people watching and some great drinks, but the Republic is pretty good. I would recommend that if anyone’s downtown.
Chris: So, John, I just heard last call here. You got time for one more?
John: Yeah, let’s do this.
Chris: So if you decided to open a Cybersecurity theme bar, what would the name be and what would your signature drink be called?
John: Cyber theme… I’m not too in the know of like, Cybersecurity, but what about AI theme? Do you cool with that?
Chris: Yeah. Let’s go. AI.
John: Okay. All right, let’s call it the data dive. Top drink. We call it the Data Delight. Or maybe, like, AI play we call it like the A.I.P.A. of the Week, right? Or you could even say like AI.PA if you wanted to really play on it. Maybe you have a cloud version.
Chris: Isn’t there like a cloudy ipa?
John: Oh, totally, dude!
Chris: Or, hazy IPA!
John: If you want to get real loose later, maybe you have cyber shots with your friends.
Chris: Cybershots?
John: Yeah, you can pull that from the Mexican binary if you want to chase the code.
Chris: Yes! Thanks for stopping by, man. I appreciate you sharing your knowledge and your insight with us. If there’s anything that I can do moving forward to help support you and Launchvox, you just let me know, and yeah, man. Can’t wait to see what’s next.
John: Yeah, man, I appreciate it. Take care.
Chris: Take care, man.