I recently read two simple tweets that changed my thinking in a profound way.
I don’t know Phillip, and, to my knowledge, had never seen him on Twitter before. But the Tweets happened to surface on my For you page under the Artificial Intelligence topic.
I think we are in for a very interesting future of creative expression. To me, these tools do change things. Something is lost and something is gained. I really enjoyed making this, but also feel the pain that certain parts of this creative process are no longer uniquely human.— Phillip Isola (@phillip_isola) April 29, 2022
I replied with two tweets of my own:
What will we lose, what will we gain, and when? The entrepreneurs and business leaders who find these answers, and adapt their talent and business models, will dominate the coming decade in every industry.— Paul Roetzer (@paulroetzer) April 29, 2022
So, why did these simple tweets affect me in such a meaningful way?
At Marketing AI Institute, we tend to focus on a theoretical net positive effect of AI in the coming years, a symbiotic relationship between humans and machines.
In this scenario, the machines will do all the data-driven, time-intensive work most humans don’t enjoy anyway, while the humans spend their days being strategic, creative, empathetic, and, well, happier.
But, what if the work that defines you, and brings you fulfillment, changes? What if AI moves further into the strategic and creative realms, much sooner than expected?
Recent breakthroughs in language and vision AI technologies are accelerating innovation, and creating increased uncertainty about where the limits of intelligent systems will be in the months and years ahead.
Every career path, business and industry will be changed by AI. It’s simply a matter of time.
And as designers learned with the introduction of DALL-E 2. And writers are learning with the rapid advancement of large language models from OpenAI, Google, Meta AI and others. The future may be closer than we care to believe.
So, what can you do to turn this uncertainty into opportunity?
Start by asking three questions:
- What will be lost?
- What will be gained?
If you’re curious enough to understand what AI is, and what it is capable of doing, you will be able to see the inevitable transformations that are coming to your profession.
Related: Want to understand AI? Register for an upcoming Intro to AI for Marketers Live free, online class.
Whether you’re a writer, designer, SEO or social media pro, content marketer, advertiser, data analyst, or CMO, AI will affect you in immeasurable ways.
But, if you figure out the answers to these questions first—What will be lost? What will be gained? And, when?—you have the opportunity to be at the forefront of the most profound technological shift of our generation.
If these ideas pique your curiosity, check out my conversation with Mike Kaput, my Marketing Artificial Intelligence co-author, on the latest Marketing AI Show Podcast Episode 14: Lost and Gained.
Episode 14: AI in the News: Lost and Gained
In this week's episode, show host Paul Roetzer sits down with Mike Kaput to talk about AI in the news. Coming off a string of tweets and articles shared among the team, it was a great opportunity for a podcast episode–and subsequent “AI in the News” series for the podcast.
[00:13:27] April 6, 2022 and the introduction of DALL-E-2
[00:19:34] Phillip Isola’s tweet that spurred this episode
[00:32:05] Things are going to change, and it’s not always going to be comfortable
[00:39:53] What will we lose? What will we gain? The reason we’re growing our community
Links referenced in the show
- DALL-E 2 and the Future of Design
- Paul’s lost and gained tweet
- Phillip Isola tweet that inspired the episode
- Marketing Artificial Intelligence: AI, Marketing, and the Future of Business by Paul Roetzer and Mike Kaput (Coming June 28, 2022)
Watch the Video
Read the Interview Transcription
Disclaimer: This transcription was written by AI, thanks to Descript and has not been edited for content.
[00:00:00] Paul Roetzer: Welcome to the Marketing AI Show. The podcast that helps your business grow smarter by making artificial intelligence, approachable and actionable. You'll hear from top authors, entrepreneurs, researchers, and executives, as they share case studies, strategies, and technologies that have the power to transform your business and your career.
[00:00:20] My name is Paul Roetzer. I'm the founder of Marketing AI institute. And I'm your host.
[00:00:27] Thanks for joining us for episode 14 of The Marketing AI Show. Before we get started, I want to take a moment to tell you about our show sponsor Market Muse. Market Muse Suite, an AI powered content intelligence and strategy platform analyzes millions of articles on demand.
[00:00:45] Uncovering gaps and opportunities for better content. Imagine an on-demand content audit that automatically identifies your best and worst pages. Content with high ROI potential quick wins at risk pages and more market muse uses AI to accelerate content planning, creation, and optimization. So you can build authority on your top.
[00:01:11] Get started for free today at marketmuse.com and I can personally attest to Market muse. We actually use market muse at marketing ions to now onto the show.
[00:01:24] Paul Roetzer: I am joined today by Mike Kaput . Mike is our chief content officer and my coauthor for our upcoming book Marketing Artificial Intelligence: AI, Marketing, and the Future of Business. See the link in the show notes. Welcome to the show, Mike.
[00:01:40] Hey, thanks.
[00:01:41] Mike Kaput: Glad to be here.I'm super excited for this format for the pocast.
[00:01:44] Paul Roetzer: It should be fun. So this is, uh, is the first time Mike and I are doing the podcast together. So Mike and I go back, what time would you have just started the agency? Like, oh, nine
[00:01:56] Mike Kaput: 11,
[00:01:56] Paul Roetzer: 2011. So yeah, so Mike's background real quick, but he journalists by trade.
[00:02:01] So I too, like for anybody who's not familiar with my background, I actually started out of journalism school, worked at an agency for five years, started my own agency. Um, in the process started the AI Institute. Actually Mike might as well tell the story, I guess we might as well set this up and then we'll come back to like, apply.
[00:02:17] We're doing this together. So I started PR 2020 in 2005, which became HubSpot's first partner agency. Um, and then we really kind of built the agency on the back of that. So tons of marketing automation, marketing technology, inbound marketing, Mike joins in 2011 is what a content specialist I think was the, yeah.
[00:02:36] Yeah, like
[00:02:36] Mike Kaput: an editorial consultant, basically just, um, focus more just on helping you get. Content marketing. Cause I did not have a, uh, kind of traditional marketing background of.
[00:02:47] Paul Roetzer: Right now. And that's, I mean, retrospect, my years may be off by one or two year, but it's around when Joe Pulizzi starts content marketing Institute.
[00:02:54] So like content marketing is becoming a thing. We're doing a lot of content creation. As an agency, Mike comes in to sort of spearhead a ton of the content creation and evolves into being like one of our top strategists. And the real key becomes, you know, I get this curiosity right around that same time as 2011 was I started studying AI, right.
[00:03:13] If I wrote the first book, the marketing agency blueprint. And so. I started spending a lot of time researching AI, right? The second book in 2014. And in that book, there is a small section about artificial intelligence. It was the first time I ever publicly wrote about or talked about any of the work I was doing.
[00:03:31] And Mike helped me do the research for that section. So it kind of peaked Mike's curiosity. I think, I don't know that you had been doing too much of that prior to that, but it gets Mike. And all of a sudden, like Mike and I just start talking about AI all the time. And so it became just sort of the happy hour thing we would do.
[00:03:46] We would chit chat about AI and, and eventually we're like, well, what do we do with this? Like, we're both kind of geeky about it, but like, does anybody else care about this topic? And so I think, you know, one of those, like, you know, open office hours kind of thing, we're just talk and it's like, well, why don't we start a blog?
[00:04:02] Like just start writing about this. And, and that kind of became. The origin of what today is the marketing eye Institute and spun into, you know, the marketing AI conference and the book we just wrote and all these other things really just organically happened because we were curious about artificial intelligence.
[00:04:16] So, I mean, that's kind of like the two minute version of the origin of all this. So the, the idea is like when we launched the marketing, I show the concept was for me to just interview really fascinating people. And we've done that. So for 13 episodes, we had incredible conversations with a Cade Metz of the New York times.
[00:04:35] And Karen Hao MIT Tech review, who just recently moved over to the Wall Street Journal. Mathew Sweezey at Salesforce, Christi Olson at Microsoft, Andrea Brimmer, CMO Ally, just like incredible interviews. And. Every intention of continuing on that path. And we have some amazing, uh, interviews lined up and that army common this summer and into the, um, you know, throughout this year, basically.
[00:04:56] So that format is going to continue. I'll still be doing interviews. We'll still have these amazing guests. But what we figured out is like Mike and I just constantly have these fascinating conversations. Like something happens in AI and I'll turn to Mike or Mike, you know, someone having a Michael turned to me and it's like, do what do you think, man?
[00:05:14] That's crazy, right? Like, like we'll look at something and some we sent, you know, Mike's written, what do you think? Eight, 800 articles or more on AI. Like there's very few people in the world that have like, from a business to marketing perspective, spent the time thinking about AI that we have collectively.
[00:05:30] And so. We're not the people building the AI. Like we're not the ones that tell you how to structure your data and how to do the machine learning models. Like, but we're really good at looking at opportunities where AI can drive efficiencies in businesses where it can create opportunities in your career.
[00:05:45] And so we just talk about this stuff all the time. So the idea behind this was like, well, what if just once a week we just kind of talk like pick one or two or three big things from the week that sorta caught our attention. And then. Have this conversation, but then explain to people why it matters to them.
[00:06:03] And so that's what this format is going to be. And again, like our, our intention is weekly. Um, you know, we're going to do our best to hold ourselves to that. But you know, last week we we've had this idea for about a month now. And last week something happened where I was like, okay, this is the, this is the starting point.
[00:06:20] We are, we're doing this podcast and it's starting now. So basically what happened is, um, Mike, and I have a sandbox of things that we could talk about. And then out of nowhere I read a tweet and I was like, this is like, I just want to talk to somebody about this. And I was at home when I saw this tweet last week.
[00:06:39] And, um, I remember sitting there. I have to say something like, I have to talk about this. I have to write about this. Like, I don't know what to do, but if I read a blog post, it's never going to happen. Like that it'll be two months and I want I've written a blog post. So I was like, okay, well maybe I'll do one of those like LinkedIn selfie videos, which I've never done.
[00:06:56] And I'm just, it's just not my style. Maybe I'll try one. Maybe I'll do one to promote this series. I don't know. But. And then there wasn't anything. So I go downstairs and I talked to my wife and I was like, okay, you probably don't care what this I'm gonna explain this to you. I'm gonna read this tweet.
[00:07:09] I'm gonna explain why this is this matter. She's an artist. So that will matter in a minute. I'll come back around to why it matters to my wife. Um, And I was like, I don't know what to do. And she's like, what sounds cool? Like you should do something. I was like, yeah. Okay. So then I reached out to them. I was like, I think I have a first podcast episode for us to do together.
[00:07:26] So here we are, this, this tweet occurs. I reached out to Mike I'm like, okay, let's, let's just do this. We're just going to talk about this topic because there's so many. Things to it. And to be honest, like Mike and I did a little prep for this, but this is more of one of those topics. When we tell you the tweet, and then you explain the context, you'll realize why, why it's just something that needs to be talked about.
[00:07:48] There needs to be more conversations around this topic happening and they're not happening. Um, and I think. Innovations are going to sneak up on people really quickly that are going to change businesses industries. And they're going to affect your career, whether you're an individual practitioner or you're an intern, maybe you're a student listening, or maybe you're the CMO it's coming.
[00:08:11] It's coming faster than we thought. So that's kind of teeing it up. I'll get to the tweet in a minute, but I think to, to make sure the tweet resonates with you the way. I think it should. Let's take a step back, Mike, and talk about a couple of key components. First is open AI. So if, if you're not familiar with open AI, it's an organization created, I think it was 2016.
[00:08:38] If I'm not mistaken, Elon Musk was it, does that sound right? So Elon Musk and a few other people fund the building of this open AI Institute, not an Institute, but open AI. Um, the, the leader of it is Sam Altman, who used to be the president of Y Combinator. So Sam comes in with Elon and filler people and they create this.
[00:08:58] Now the concept behind open AI. Was at the time. So 2015, 2016, we're like three years into the deep learning movement. So a lot of the capabilities in language and vision AI. So the things you experience every day where Gmail finishes your sentences and apple unlocks the phone with your face. You know, Facebook recognizes people in your photos, you upload and all these things where your, you know, your computer vision is coming into play and the ability to understand a general language like voice assistance and things like that.
[00:09:29] That's all like three years old at this point, like th th the, the breakthroughs in that technology had just basically happened. And so there was a land grab for talent. So like Google, Microsoft, Amazon, Nvidia Salesforce, like all these. Our re uh, Facebook racing to hire all the AI researchers, like the best of the best to build out research labs within their organizations.
[00:09:52] And in essence, then take all the innovation and keep it within the walls of those organizations until they figure how to monetize it. Open AI comes along and says, no, no, no, we're good. And we're going to do open source and that since the AI innovation, we're going to push the limits of what is possible capable of, and then we're going to release it.
[00:10:09] We'll open these models up and we'll let people innovate on top of it. So that's the core concept, but their mission is artificial general intelligence. So if you're not familiar with artificial general intelligence, what we have today, all the applications of AI in your life from Gmail to Google maps, to Netflix, to Spotify, to Amazon predicting purchase all these ways.
[00:10:30] AI is. Is what's called narrow intelligence. It's specifically trained to do specific things. General intelligence is the ability to do multiple things, to achieve multiple outcomes without having to necessarily be told how to do it. It can actually function more at a human level of intelligence. So.
[00:10:50] There's debate about, will we get there? How quickly will we get there? All of these things, but organizations like open AI and deep mind within Google have stated missions to get us there, to get machines, to be as smart as humans,
[00:11:03] which obviously has some significant effects now to get there. There is a belief that intelligence, as we know it, human intelligence.
[00:11:12] What makes us different from animals is things like language in particular. And the ability to see and understand the world around us vision. So the reason these organizations in large part work on these innovations and why they try and create machines that understand generations. Is because they're trying to create advanced intelligence, general intelligence.
[00:11:36] So when you hear about advancements like GPT three, which has embedded in, you know, dozens or hundreds of marketing tools that write copy, you know, the different things you hear about writing social posts and things like that ad copy, um, or you hear about things like Dolly too, which I'm going to let Mike talk about in a moment here.
[00:11:54] These advancements are being achieved because there are really smart people at really financially strong organizations who are waking up every day, trying to build intelligence into machines, advanced intelligence, and they need to, they need to reach certain levels of language, understanding, and generation and.
[00:12:20] Vision thing they need to give the machine, these things, these human-like abilities. Um, that's a belief again, not everyone in the AI. Deep learning community believes that language is necessary to general intelligence, but the prevailing understanding that we have from find the spaces, the majority of people seem to believe that AI is to get to general townships.
[00:12:43] You, you have to have machines that can learn language. So as mark. This obviously matters dramatically. So what we do is generate language. We read things, we write things, we created imagery. Um, and so it has a massive effect. And so Mike and I back in 2015, asked this question like, well, can AI write content?
[00:13:03] Like, can it write blog posts for us? And the answer to that of 15 was no, like we thought maybe, but the answer was no. So two, I'm going to come back again to the tweet that sort of changed. Everything for me a little bit in how I think about this. Now I want to talk about this. Um, but Mike, why don't you take a moment and talk to us about Dali two and what happened back?
[00:13:27] It was April 6th or seventh open AI introduced Dolly to, um, which was sort of the first domino in what led to this podcast. Yeah,
[00:13:34] Mike Kaput: for sure. So DALL-E-2 to that, if you want to spell it, it's got a weird spelling. It's D a L L dash E and then the number two. And that's like, you'll get kind of a clue as to what this does by, uh, the description of why that's the name.
[00:13:49] It's a combo of. The artist, Salvador Dali and Wally the Pixar or Disney robot. A fictional movie. Yeah. Has a great
[00:13:59] Paul Roetzer: movie, like no talking for the first 45 minutes. Really.
[00:14:03] Mike Kaput: So what Dali two does is, and this is just mind blowing. Like it. Straightforward, but when you actually see what this looks like, you're going to be pretty impressed.
[00:14:14] I think so Dolly too is AI that automatically creates original art and imagery based on natural language prompts. And what that means is you literally pull up Dolly tube, you type in. Anything show me a picture of their example is show me an astronaut lounging in a tropical resort in space and do it in a photo realistic style.
[00:14:40] And within a few minutes, the AI will produce. The image that you described in the style you described, and if you go to open ai.com, it's the main, if not the only thing right now on their site, um, check out the examples because they're stunning. They're incredible. Um,
[00:15:01] Paul Roetzer: and the original that the key here is they're not scraping an astronaut and scraping the moon, like.
[00:15:06] Mashing together, a bunch of existing images from a bunch of train it's creating, which we just wrote the book. And I was like, creativity is still a human domain. And now I'm like questioning. Did, you know, did it happen?
[00:15:19] Mike Kaput: Yeah. Yeah. And so the cool thing is this is the two indicates this, the second iteration of this system and open AI has made it just stunningly more powerful than the previous ones.
[00:15:31] And, uh, like I said, the results are incredible. It also has pretty incredible. And I think this is an underrated part of it editing capabilities. So if you tell. Hey, edit the Flamingo out of the background of this image. It will go do it. And it looks based on the examples. Really, really good. So that's at a high level.
[00:15:53] I don't want to like ruin the whole, uh, the whole reveal here, but that just know that AI can create photo-realistic art or original art and imagery in almost any style you
[00:16:05] Paul Roetzer: want. And now it is not readily available. So if you're a graphic designer or list or whatever, and you're like, oh my gosh, like I, I have to go test this.
[00:16:13] You can't, you can get on the wait list, but within like, Twenty-four to forty-eight hours. There was over a hundred thousand people on the wait list. So I put in to get on the wait list. So right now they've rolled it out to some let's select friends and researchers. And I don't know how they've chosen.
[00:16:25] Who's gotten it. But if you follow on Twitter, like the hashtag, there are people who have access to this who are creating things and they're sharing the things they're creating. And it does sound like they have every intention of making this an available tool later this year. So. April 7th happens or whatever day that was once I actually have the tweet up.
[00:16:45] Yeah. So, uh, April 6th, 2022 opening, I tweets our newest system Dolly. Can create realistic images and art from a description in natural language. See it here. So I have an AI list. It's actually public the Twitter list. We'll put it in the show notes. So it was like a hundred AI people in there, researchers entrepreneurs.
[00:17:06] So basically when something happens in AI and I'm not sure what to make of it, or like, I, like, I was like, wow, this seems like a huge deal. I'll go to that list and see if they're talking about it. And if so, what they're saying, and it was one of those days where. Within minutes, that list starts blowing up with Dolly to stuff it's like, oh, okay.
[00:17:24] Like this, this was a big deal. Uh, so I immediately look at it. Well, wait a second. Like what, what happens if you're a graphic designer or an illustrator? Like, again, go back to my wife is an artist. Now she's not a graphic designer and she doesn't do too many illustrations, but like we've had conversations about, should we do an illustrator, like a children's book someday?
[00:17:43] Like, I'll write it with my daughter. And, and then her and my daughter will illustrate together. Like, it's quite, it's like, well, we, we need to do that now. Like, can we just feed the story? To Dolly tune. It'll illustrate the book for us, like, is that where we're now at? So Mike and I go back to April 7th, we're just saying like, oh, what does this mean?
[00:18:02] Like w how, how, if you're a graphic design agency, do you have a job in six months? Like we always tell people that our general feeling is AI will have a net positive impact on jobs and the economy that more jobs will eventually be created by AI. AI is largely going to take tasks away in the near term.
[00:18:24] It's going to do specific tasks better than you like, maybe adding out the Flamingo, like, but that's not new. Like if you have filters on like Instagram and stuff, I think you have the ability to edit stuff out of that. That's magic. Like, it feels like magic, but that's not new generative from the beginning, from, from her description is new.
[00:18:43] So for the last like month, almost a full month. In the back of our minds has been this, like what, what, what does this mean? And then there's like GPT three, which is language generation, which just keeps happening. And so as we're finishing up the book again, which comes out in June as we're planning our marketing AI conference, which happens at the beginning of August in Cleveland.
[00:19:05] And we're thinking about the agenda, these sorts of things that are running through my mind, it's like, well, where are we at? We didn't think AI could do what we do, but should we take a step back here? So that's, what's going through my mind. And then on April 28th, this dude Philip I solo who, I don't know the, the, the AI algorithm of Twitter, which surfaces things that it thinks I would like on my homepage or in my search, which AI is one of the categories.
[00:19:34] Pops up this Phillip Isola a guy I've never, ever, never heard of the guy before for this moment. So I don’t click on his profile. All I read is the tweet and it says back in 2018 at open AI, a few of us wrote a story with GPT. So this is the original, not GPT three. This is the predecessor to two generations ago.
[00:19:54] I wrote a story with GPT as an AI quote co-author we didn't have an AI illustrator back then, but now we sort of do so I tried plugging the text into Dali. Here's the result, the BS quotes, which is the title of the. Story. The poem is a short story by humans and AIS, and then he links to the story. So now again, we'll put this in the show notes.
[00:20:24] So all the images in the tweet is the B's the title, and then an illustration that obviously Dolly created with the first passage from the story. So I don't click on the story yet. I have not actually gone and read the story, scroll down his next reply right after. I think we are in for a very interesting future of creative expression.
[00:20:48] To me, these tools do change things. Something is lost and something is gained. I really enjoyed making this, but also feel the pain that certain parts of this creative process are no longer uniquely human. That was the point where I was like, oh my God, like I. Stop. Like, I, I sorta grew back from my computer and I re-read that line a couple of times something is lost and something is gained and it's a, it's like this feeling I've had.
[00:21:21] Many times throughout the years, like I remember Mike, you were there for Angela Pham’s keynote. She worked at Facebook, still works at Facebook. I believe she did a keynote on at our first marketing conference in 2019. And she talked about this, the loss of humanity as AI gets better and better. How a friend of hers who wrote in broken English, all of a sudden started writing like perfect emails.
[00:21:47] And she realized that he was using smart components. And like finishing it was finishing a sentence for him. And she said it was like, it was like a piece of him wasn't there anymore. Like this wasn't him writing to her now it was, it was the machine and him. And so like, there's been a few times in the last few years where I've had that feeling, but I've, I've struggled to explain to professionals, to like writers, designers, business people, whoever, whatever your social media people, advertising people.
[00:22:16] Who don't care about AI, who aren't following it yet, how to make them care. And that line just kept coming back to me of like, that's, what's going to happen. Like whatever your career is, whatever industry you're in, you're going to lose a piece of what you do and who you are, especially if the job. Is a true career for you and it's defining for you.
[00:22:39] Like it is who you it's part of who you are. It's how you get fulfillment in life. You're going to lose aspects of that. The question is. What do you gain though? And when does this happen? So I want to come back to that mic and get your thoughts. I know you wanted to actually read the poem, so why don't you go?
[00:22:58] And like, again, kind of like keep filling in the blanks here for people, but why don't we take a step back and like, okay, now read the poem now, you know where this goes? Let's listen to the actual story share, and
[00:23:09] Mike Kaput: it's super short, and I want to read it for a very specific reason, which I'll tell you at the end.
[00:23:13] So here it is very quickly. Once upon a time, there was a beehive that produced the most delicious and most golden honey in the entire universe. But lately something strange had been happening. The bees had been sleeping long past the usual end of their winter slumber. And by the time they wake up. There is not a single flower left in the forest, except the flowers have an extremely spicy pepper plant.
[00:23:37] They found out someone was drugging them. Oh, my drug bees is the most hilarious spectacle to watch. They're buzzing orchestra was completely off key and the bees dancing was crazier than ever. But, oh, how did they dance? There was so much dancing that you couldn't make it out. It was like they had a life of their own and in some mysterious way, they were happy.
[00:24:01] VN. And the reason I wanted to read that is because of course there's like some weird language in there. There's some that doesn't make sense. Um, there's some just strange sentence construction, but I got to the end of it and I kind of felt like, I felt just like a little tinge of like, I was like, oh, that's kind of a cool little story.
[00:24:22] It's probably no more nonsensical than your average kid's book, to be honest
[00:24:27] Paul Roetzer: poetry. I mean, let's be honest. I love poetry. I wrote poetry in college. Half it doesn't make any sense.
[00:24:33] Mike Kaput: So it's one of those things where I was like, wait, I just felt like a very smart, like, I felt something from this. I don't always know why.
[00:24:43] And they don't know if, what I felt was because of what the human wrote or what the machine wrote. And so, look, I'm not going to like overly critique this story, but it's like, you're reading this. You're seeing drawings that look exactly like illustrations and children's books that were generated by Dolly too.
[00:25:02] And I kinda just got to the end and was like, it doesn't really matter who did what. What matters is like, can you say this is not art when you get to the end of it? Like, just because the machine did it. So what it made me feel a little something, it was visually appealing. Isn't. A piece of art. And I think at that moment, I was like, oh, this is really going to change a lot of people's definitions about what things are.
[00:25:32] And it's easy to laugh off. Right. When you say like, oh, okay, machine did it. They can't do it the same way as humans, but it's like, that's simply no longer
[00:25:40] Paul Roetzer: true. And I don't think I said this. So fill up, who I didn't know is an associate professor at MIT, and he's spent time as a visiting research scientist at open AI, um, from 2017 to 2018.
[00:25:56] So that's kind of his, his background, but so, you know, I read this, I try and figure out what do I do with this information I wrote down. What is lost and what is gained. And when question mark, and what I realized is it is the fundamental question. Every one of us needs to be asking of our, of our chosen career path and the, and potentially of our businesses.
[00:26:25] So me as an entrepreneur, as a writer by trade, um, as you know, I do, I still heavily involved in the marketing of the business, like, you know, marketing strategy and what we do there. And it's just like, you look at those things and you say, but what does, what does my career look like? So I'll bring it down another level.
[00:26:43] So it's not as obstruct from a writer, you know, in three to five years, like right now, language AI is moving faster than Moore's law. Like I just, this morning we'll put this meta AI. Introduced an open source version of something, basically as powerful as GPD three, if not more powerful, except they're going to open it up to everybody.
[00:27:06] Well, right now researchers, so if you're a research Institute, you'll be able to get access immediately. I think to the GPT three tool that meta has built where GPT three is being commercialized. So it's being kind of put up. This is being designed for researchers to work on. That two years ago would have cost millions of dollars to build a good access to.
[00:27:31] And now it's just like a free tool. So the accessibility and the capability of language generation is moving so fast. That there's no way Mike or I could sit here and predict where it's going to go. Like if you would have asked people in the industry six months ago, when do you think we'll get to a point where a machine can generate original illustrations and artwork from a text description?
[00:27:56] I mean, maybe there's a handful of AI researchers that I don't hear from very often that would have guessed sometime this year. But my guess is they probably would have said maybe in the next decade, that five to 10 years, like
[00:28:09] Mike Kaput: I saw someone in the industry tweeting the morning of Dolly two's announcement, something to the effect of between Dolly to announce this week and also in the last week or two.
[00:28:22] The further development in language models. It's a real tough month for people that say machine learning has hit a wall,
[00:28:29] Paul Roetzer: or that it's only for repetitive data-driven tasks or that it's not going to affect creatives. So. What is lost, what is gained and when it's like, there's so many people just don't care, like don't follow what's going on in AI think, ah, whatever.
[00:28:44] It's just the next trend. I'll worry about it in three years, five years. And we've said this line over and over again, it's like this, it changes everything. And it's like a really hard macro level idea to get people to care about. But it truly does from a career perspective. And again, creativity is one of the things that I thought was uniquely human.
[00:29:06] I still, I mean, I still want to say it's uniquely human. That what AI does, isn't the same. But it, the lines are getting blurred. Like I, I think I, I wrote a passage in, in the marketing artificial intelligence book. That's coming out something about like, cause I was actually using Google deep mind and AlphaGo as the example and whether or not AlphaGo was creative and its moves against the world go champion.
[00:29:30] And I said, this story about like, basically like, okay, if you, if you try and. If you try and get a machine to create an image, like does it, and it's an image of a dog. When, when we do that as humans, like you understand dogs, you had a dog, you have a personal connection to dogs, you understand their emotions and why they bark.
[00:29:49] And like, it's like we have a human understanding of a dog. The machine doesn't have that a machine just sees pixels, like generates things through math. Like it doesn't experience in creating the same way we do. Like it's, if you, if a machine creates a music, it doesn't have a life of experiences and pain and joy.
[00:30:08] And like all the human emotions that go into human generated music, but to the ear, can you tell the difference? Like, can it connect? So is it actually creative? What I think I understand. It's not creative like we are, but I don't know. You could say it's not creative. And now you look at this, it's like, wow, could you possibly say that's not creative?
[00:30:31] Like, it's just how you define creativity. I guess that becomes the key
[00:30:35] Mike Kaput: for sure. Yeah. It's one of those things where I think you really hit the nail on the head. It's what, what is surprising about this? Isn't necessarily just the fact that a machine can create art or that a machine can generate language.
[00:30:50] It's the question of, okay, you can sit there and turn your nose up at math driven art or machine driven. But does it matter? And this technology and the rapid development of it doesn't really care about our feelings. Uh, what of, what is, and isn't possible at the end of the day, if you read a perfect article or see a piece of art that moves you.
[00:31:15] You don't know anything about the human that wrote it or their experiences? A lot of times, unless it's a very famous artist or author yet, it's the power of the actual outcome that deliverable the product that is what moves you or gives you ideas or excites you. And we're very rapidly getting to a point where it's like, people are going to ask, like, what is the difference?
[00:31:38] Who cares if the machine wrote it? If it's. Achieves the same effect. Right. And I think that's where we're going to get into really, really murky
[00:31:48] Paul Roetzer: territory. Yeah. I mean, I guess the whole point of this episode for the people that are curious for the people that are already paying attention, and maybe, maybe you weren't before this, maybe this is the first time you're ever really given, you know, care about AI for some of you've heard from us and things like that.
[00:32:05] I think what I'm trying to say is. Humanity is gonna, like, things are going to change and it's not always going to be comfortable, but it's sort of inevitable. So this technology is going to raise forward, Google, Amazon, Microsoft Nvidia by do like Oracle, Adobe. Like all these companies are racing to build more advanced language envision, AI capabilities.
[00:32:37] Trying to hire as many people as they can to create this stuff and move as quick as possible. So as a, as a business person, as a professional, as an artist, as a creator, you don't get a say in this, the technology is going to be created and it's going to advance. All I think we can do is accept it. I'm not saying you have to necessarily embrace it.
[00:33:00] But you have to accept that the change is coming and it's going to be coming faster and faster. And going back to the question of what will we lose, what will we gain? And when, if you're willing to constantly ask that question, because again, going to change it, like I, if I was a graphic designer, illustrator, I would not have thought I may lose the need to create original works.
[00:33:21] That that may be the machine does that. And then I curate an edit it, like, I don't know, like, I don't know what an illustrator. How this affects that I it's not my world, but I, I think about it. Um, but to ask those questions and then say, so what's the opportunity like, okay, I run an agency that writes blog posts.
[00:33:40] 90% of our revenue comes from blog posts, writing. I run a graphic design studio. You know, builds logos and corporate collateral, or I'm an illustrator for books and things like that. If you're constantly asking this question and then the next question is, okay, so what's the opportunity. If Dali too can do this what's opening eyes plans.
[00:34:00] Okay. They're going to release it later this year. Okay. How do I get there first? Like Gary V I'm not like a huge Gary V. Guy. I'm not up here. Like, you know, shouting out Gary V for everything. I think he's grabbing God bless the guy. Like he is made. Like he's just genius. He works his ass off. He deserves everything.
[00:34:17] He gets. This is the kind of thing where it's like, he's always at the forefront. I'm not saying be Gary V on this. What I'm saying is your industry is going to get disrupted what you do today. Someone else is going to find a smarter way to do it with better technology. But if you're constantly asking yourself those questions, what will we lose?
[00:34:38] What we gain when, and then what is the opportunity? Then you can actually get out ahead of this stuff and maybe live a fuller life, have a better career. So I don't know. I mean, it's kind of like my I'm just sort of like big picture. What I probably wanted to say when I first read it. I'm like, if you have any additional thoughts on that topic.
[00:34:54] Mike Kaput: No, I couldn't agree more. And I think as people wrap their heads around this, as they look forward to the opportunities, I think Sam Altman open AIC. Had these two quotes that I pulled out of the initial long form post, he wrote about Dolly to describing it, what their plans were for it. And the first quote is.
[00:35:16] Well, a decade ago, the conventional wisdom was that AI would first impact physical labor and then cognitive labor. And then maybe someday it could do creative work. It now looks like it's going to go in the opposite order. And I think that's super interesting because one not to scare creative people, but the fact, the bigger point that even the people at the forefront of this, of these developments, Are not able to see into the future.
[00:35:43] It is outpacing some of the smartest people in the industry, even the ones making it, which means it's going to be a very strange and very odd. Knowable or uncertain future. So it's like these expect thought experiments. Aren't just thought experiments. They're critical to start trying to figure out what's coming next and how it's going to affect you.
[00:36:05] And then I think this is where I really saw the opportunity is that later in the post, he said, This, you know, the technology is quote, it's an example of a world in which good ideas are the limit for what we can do, not specific skills. And I think that's really what I sometimes see as like a marketer.
[00:36:24] Sometimes creator is. I don't have a ton of design background. I don't have a ton of visual background, but you know, if I had a designer sitting across from me right now, I could really like vaguely describe like, okay, here's kind of what I'm thinking for the art on this blog post or ebook or report it's like this, this and this.
[00:36:47] And like kind of like this other thing and this other style and the designer would hopefully listen, have some really good. Creative ideas and then use their skills. To go create that for me. I can do that with technology like Dolly too. And for me, that's super exciting because I'm like, oh my God, I have all these cool ideas.
[00:37:10] Like for every one of our blog posts, like I've never been, I mean, like you said, we've done 800 blog posts. I've never been more. Sick of stock art in my life. Like we do, I think a pretty good job of trying to select some really engaging imagery, but I, if your stock art site, there's no point it's like, you're, you might want to start thinking about this.
[00:37:30] So it's just one of those things where I get excited about the opportunity, but my opportunity, the thing I gain actually may end up in parallel being someone else's law.
[00:37:44] Paul Roetzer: Exactly. That's, there's always a balance. And I, you know, as you're were talking, I was even thinking, I mean, I'm going to go down this on this episode, but so like my daughter who's 10.
[00:37:56] Like she wants to be an artist like her mom, and she's amazing already at it. And I could see her wanting to do illustrations and things like that. Like if I were to show her this technology right now and say, Hey, you know, you love unicorns. Like let's make a unicorn dancing on rainbows. Like she has a song for unicorns dancing on rainbows and it did.
[00:38:13] In front of her eyes. What does that do to her? Does that de-motivate her to be like, well, what does it need? What does the world need me as an artist for, if I can just ask a machine to do it. I have no idea what the answer to that is. I'm just actually like, I'm starting to kind of like process. This idea of like this isn't cause he's got it.
[00:38:34] We understand like, again, if you don't know me and Mike, you don't know the Institute, we are not all about AI for a bunch of hacks and shortcuts and cost savings. That is not what we teach this for. The tagline of the conference was more intelligent, more human. What we were trying to find is what are the ways AI can make us more human?
[00:38:49] How does it free us up? To do things like strategy and creativity and use empathy. And you know, the things that are really hard to give to a machine, if not impossible, how does it free us up to actually live more fulfilled lives in a way, like at a very high level. So I tend to think about things at these different levels of okay.
[00:39:08] Professional first higher education. If I'm, if I'm a professor of graphic design and this comes out in the fall, what in the world? Tell my students. And then I think about the next generation who are right now or inspired to be creatives and to do many of the things that by the time, honestly, they're in the working world, they probably aren't going to need to do.
[00:39:29] Um, I mean, my daughter is 11 years. It's like, I can't even fathom that the innovation, like, again, deep learning as an, uh, capability in language and vision. All these things we used was basically 2011, 2012 was the tipping point 10 years. What happens in 10 more years? The innovation is to be a hundred times as powerful as what we see today.
[00:39:53] So I, I think like, I don't know, probably wrap this episode, like again with no real profiling, like, okay. Here's how to solve this again. I, I think the call to action is just think about that. What would we lose? What will we gain and win for, for you for your business, for your career? Personally, um, because we don't have the answers.
[00:40:12] And I think our whole thing is we're trying to build a community of people who are asking the hard questions and maybe collectively, we start to find some really interesting answers. If nothing else, we get a bunch of people that care about using AI for good working in the same direction. Um, because we're, again, we're not the ones building it, so we can't change the way they build it per se, but enough voices that are trying to, you know, get brands and tech companies to think about these things as they're building it, um, can only do positive for the.
[00:40:43] For sure. All right. Well, hopefully this was interesting to you. Hopefully you will join Mike and I again, um, again, like we're gonna try and do this every week. We, um, we hope you'll join us. Check the show notes. We we've kind of talked about a lot of things we'll go through and put all the links to everything in the show notes.
[00:41:01] Um, this gets published on the marketing Institute blog. So. Podcast networks, but always post a summary on the blogs. So if you want jump on the Institute of marketing AI institute.com and subscribe to the newsletter, um, yeah. And you know, we hope you keep listening to the podcast are 14 episodes in and we're planning to do a lot more.
[00:41:21] So, um, and we'd love to hear your feedback. And Mike and I are both pretty good about responding on LinkedIn. I mean, LinkedIn and Twitter, I'm pretty good at both of those. So we'd love to hear from you, you know, if you love this. Don't love the show, whatever, whatever we can do to make a better. So thank you, Mike.
[00:41:36] It was enjoyable doing this for the first time and not standing around the coffee machine, actually like the mic in front of us.
[00:41:43] Mike Kaput: I loved it. Yeah. If we can actually, uh, instead of us just coming up with a bunch of ideas, we can actually share them with others
[00:41:50] Paul Roetzer: to a crowdsource stuff. Like here's something they're going to get to do this.
[00:41:54] All right. Well until next time, stay curious, explore AI and uh, we'll be in touch. Thanks.
[00:42:01] Paul Roetzer: Thanks for listening to the Marketing AI Show. If you like what you heard, you can subscribe on your favorite podcast app. And if you're ready to continue your learning head over to marketing AI institute.com. Be sure to subscribe to our weekly newsletter. Check out our free monthly webinars and explore dozens of online courses and professional certifications until next time, stay curious and explore AI.
Paul Roetzer is founder and CEO of Marketing AI Institute. He is the author of Marketing Artificial Intelligence (Matt Holt Books, 2022) The Marketing Performance Blueprint (Wiley, 2014) and The Marketing Agency Blueprint (Wiley, 2012); and creator of the Marketing AI Conference (MAICON).