Not all AI news impacting the marketing world involves generative AI, and this week on the podcast, Paul and Mike discuss three artificial intelligence developments, breaking down their importance to marketers. In a word (or two): buckle up.
They start with well-known marketer Neil Patel’s recent blog post where he revealed the results of Google’s latest algorithm updates on sites he owns that have AI-generated copy—and the results weren’t pretty. Patel disclosed that he has “100 experimental sites that use AI-written content.” He claims the sites are simply to figure out how Google perceives AI-written content, not to “game” the algorithm. Regardless of the motivation, he sure found out.
Next, Boston Consulting Group (BCG) recently released its guidelines for how companies should approach responsible AI based on its Responsible AI Leader Blueprint. BCG defines responsible AI as “developing and operating artificial intelligence systems that align with organizational values and widely accepted standards of right and wrong while achieving transformative business impact.
And finally, earlier this year, the Partnership on AI did work on better understanding how AI will change the local news landscape by talking to nine different experts in the space, including prominent media outlets and technologists. The Partnership on AI is a major nonprofit that was founded by Amazon, Facebook, Google, DeepMind, Microsoft, and IBM to research and share best practices around the development and deployment of artificial intelligence.
00:02:18 Neil Patel analyzes the impact of Google updates on AI-generated content
00:14:37 BCG publishes guidance on how to do responsible AI the right way
00:24:34 Partnership on AI and media experts look at how AI will change local news
Disclaimer: This transcription was written by AI, thanks to Descript.
[00:00:00] Paul Roetzer: You've got to be looking around and understanding how quickly this space is moving and you need to be proactive in doing things the right way.
[00:00:09] 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.
[00:00:29] Paul Roetzer: My name is Paul Roetzer. I'm the founder of Marketing AI Institute, and I'm your host.
[00:00:38] Paul Roetzer: Welcome to episode number 23 of The Marketing AI Show. I am your host Paul Roetzer, along with my co-host Mike Kaput, Chief Content Officer at Marketing AI Institute, and co-author of our book, Marketing Artificial Intelligence: AI ,Marketing, and the Future of Business. What's up Mike? How's it going?
[00:00:57] Paul Roetzer: Good. Mike and I were actually in office together today. It's sort of a rarity these days where the team, the whole team was in the office today. Yeah, no kidding. That was kind of nice. I'm now no longer at the office , but for a moment in time we were all back together again, hanging out and having meetings.
[00:01:12] Paul Roetzer: So, before we get started, take a moment to thank our sponsor, Persado. Persado is the only motivation AI platform that generates personalized communication to immediately motivate each individual to engage and act. Organizations that use Persado benefit from an extensive customer motivation knowledge base, enabling them to hyper personalize their communications at scale.
[00:01:37] Paul Roetzer: Clients who incorporate Persado's innovative motivation AI increase conversions by an average of 41%, unlocking millions in unrealized revenue. Visit persado.com. That's P E R S A D O dot com to learn more about their motivation AI platform. And with that, Mike, I'll turn it over to you, and if you're new to our weekly format, we pick three topics, Mike and I just kind of keep track of the trends throughout the week. Take input from the other people on the team. And then we pick three things to talk about that seem top of mind and really relevant to our audience of marketers and business leaders. So, Mike, take it away. All right.
[00:02:18] Mike Kaput: So first up, we have an interesting article from Neil Patel.
[00:02:24] Mike Kaput: Neil Patel is a well known marketer, pretty widespread presence in the marketing community, and he owns both an agency and a number of other sites and businesses, both for clients and for himself. And Neil recently revealed how Google's latest algorithm updates, including mostly a recent spam update, but also the helpful content update that happened in the last couple months, as well as a core algorithm update actually impacted sites he owns that have AI-generated copy.
[00:03:01] Mike Kaput: And to put it lightly, the results were not pretty. So Neil Patel disclosed that he has over a hundred what he calls "experimental sites" that use AI-written content. He claims the sites are simply to figure out how Google perceives AI-written content not to game the algorithm. Now, regardless of motivation, he sure found out how Google perceives AI-written content because as a result of these updates that Google has rolled out in the last few months, Patel saw his AI-generated content sites get crushed by the most recent spam update, which just completed, I believe, in the last week. And he saw for AI only content--- content, not modified by humans, just produced by AI, and published--- an average drop of 17% in traffic and an average drop of almost eight positions in search rankings for AI only content.
[00:04:03] Mike Kaput: Now, he actually compared that to AI-generated content that he did have a human team edit, add to, and kind of polish for publication. And those sites only dropped 6% over that same period and saw about a three position drop in search. So a dramatically lower impact on the AI-generated content that has been also augmented by humans.
[00:04:33] Mike Kaput: Now, Patel readily admits that it's still really early to kind of draw ironclad conclusions here, but it certainly seems like purely AI-generated content may have some blind spots when it comes to search. So this is an area we obviously have experimented in, kind of. Paul, what do you make about of this development and of Neil's findings
[00:04:56] Paul Roetzer: here?
[00:04:57] Paul Roetzer: I mean, I love Neil. His content's amazing. We've followed Neil for years. I love that he describes them as experimental sites that use AI-written content. I assume that means he was making no commercial gains from these hundred sites that were using AI-written content for however long he's been experimenting with this.
[00:05:19] Paul Roetzer: But anyway, so experimental sites or commercially experimental sites, regardless, I think it's interesting that there's data to it. You know, as soon as we saw the helpful content update, and it was very clear that there was an attempt to weed out or penalize AI-generated content, you know, it seemed Google was definitely moving in that direction.
[00:05:42] Paul Roetzer: And then with the the spam update and the different things that they've been pushing, it's pretty obvious that there's a bit of an arms race developing where Google wants to know if AI wrote the content and the platforms enabling the content creation with AI and the people using those platforms probably don't want Google to know that AI wrote most of the content.
[00:06:08] Paul Roetzer: And so it's one of those where right now Google appears to have found ways to know if content, or predict that content, was likely, largely written by AI. And my guess is that the algorithms on the other end are to keep getting better and keep making it harder for Google to know AI wrote it. I know that there is a, what was it?
[00:06:31] Paul Roetzer: Is it Writer that has the AI copy checker where it actually, like, you can put the content in and it'll analyze how much it seems that AI wrote of that piece of content. So I think there's going to be this really interesting phase where Google's going to keep trying to find ways to say, "make your content more human."
[00:06:51] Paul Roetzer: And machines are going to keep trying to find ways, or the people to build the machines, keep trying to find ways to make the machine content sound more human . And somewhere in this is probably the moral of the story, which is, "don't rely on AI to write your content purely." It is increasingly valuable as an assistant to writing content and generating ideas and building briefs and developing drafts, but do not remove the human from the equation.
[00:07:21] Paul Roetzer: I think we knew all that before, it took Neil building a hundred experimental sites to figure that out statistically, at least in his world here. But I think it's been pretty generally accepted for a while that letting AI write all of your content was going to end poorly for everyone.
[00:07:41] Paul Roetzer: And I wouldn't advise it. I would think of AI as an assistive writing tool. And I would keep humans in the loop at all times. It might not be as much as we used to need 'them, but the humans need to be there and they need to be helping craft the final content, I would say .
[00:07:59] Mike Kaput: So it sounds like you don't see this as necessarily invalidating the use case or the value of AI content generation, but more approaching it with a little more nuance than subtlety. If I'm a company or a brand, I imagine, you know, we both used to work in the agency world, in the client services business, and the first question I feel like we would probably get from a client after reading this article is something like, "Okay, well, should we be using it? Should we not? How should we be using it?"
[00:08:33] Mike Kaput: Yeah, I mean, it goes back to, I think we've talked about it in this podcast before, or maybe I did it in a recent interview, you know, you and I started this Project Copyscale back in like 2015- 16 at our agency, and the question was, can AI write content? Like can we as a content agency, use AI to help us write blog posts?
[00:08:53] Mike Kaput: And the answer then was a resounding no. Like it was nowhere close. The answer today is yes, it can write blog posts like, I mean, I've been playing around with Jasper recently. You know, we've been looking at a lot of tools. Looking at Writer is fantastic. We've got Hyperwrite, GoCharlie, I'm sure I'm missing some.
[00:09:12] Mike Kaput: We've tested Frase, we've played around with a lot of writing tools. Many of them are GPT-3 based, and I can tell you right now, like it's pretty good at writing drafts. It's shockingly good actually. Like if you just give it a prompt, like write me a article about, AI-generated content, it's going to give you 400- 500 words if you go long form and it's going to read, to the average person, like a decent human writer wrote it and so there is absolutely going to be a desire for brands and agencies and publishing companies to just say, "Well, if it's that good, like let's just let the AI do it." So it's going to be really challenging because it's pretty good now, and I can't even imagine like six months or 12 months from now, like these tools we're playing around with and that everybody has access to.
[00:10:09] Mike Kaput: I know they're getting better fast. So I don't know. I mean, I just, I look at this space and it's, it is every day. it is more fascinating to see what's going on in the language space. Yeah,
[00:10:23] Mike Kaput: absolutely. I mean, it's obviously still really early, but given how fast these tools have developed, I imagine there's a lot of people including us, who are sitting back and really thinking about, so what do you see as like the human's best and highest role in content creation moving forward? Because there's no doubt it's changing, even if it will take a couple more years to truly sink in for a lot of other people that are skeptical and this will happen. I mean, where do we fit in?
[00:10:58] Paul Roetzer: I mean, I think about editors more than the writers sometimes, because I think that's almost easier to understand because I think a lot of writers are going to spend more time as editors, I guess is the simple way to say it. So if I were to write a thousand word blog post in a traditional process, I would start with my idea. I would build an outline. I mean, you and I do this for a living, have done this for a living for decades. You build your outline, you know, you figure out what your lead is, you figure out your key points within the story.
[00:11:34] Paul Roetzer: You kind of have an idea of where it's going to go. And then you start writing and you draft and you write the sloppy first draft. And then you go back around and you know, I'll usually do three, four revisions of my own. And then I'll normally have someone on the team edit it. It's like that thousand word post where I can knock out the draft, maybe, in an hour and a half, like hour, hour and a half, depending if I do any research related to it, the end product is going to take me five to seven hours before I hit publish, you know, and upload the thing and do all this stuff. I could absolutely see that entire process taking 30 minutes to an hour on the high end.
[00:12:10] Paul Roetzer: But the research phase is going to be automated. Like I'm just going to say, "Hey give me five bullet points on the problems with using AI-generated content." Boom. Like instantly, 10 seconds in, I've got my five key points. Like those are really valid and it might give me an idea for like two other ones. Whatever.
[00:12:25] Paul Roetzer: I got my points. Okay. Write me the first paragraph about the opportunity that exists to write content with AI. Boom. Instantly. Like, I got my paragraph. Okay. Keep going. Like, that's pretty good. And now I'm just like, I'm letting the machine, I'm just prompting it, like, keep going, give me, give me more content, and then I'm going to step back.
[00:12:41] Paul Roetzer: And I haven't written the draft yet. Like the AI has now done it. I've prompted the draft, but I haven't written anything yet. Now I'm going to step back and I'm going to read it and like, this is actually really good and I'm just probably going to highlight, I'm going to update this paragraph. I'm going to do this paragraph, do a plagiarism checker, make sure it's like legit and like original.
[00:13:00] Paul Roetzer: But my role as a writer is largely now an editor. And maybe I'm going to jump in and rewrite my own lead because I got some inspiration now, or I'm going to do a conclusion in a different way, But, the process is completely different. And that's today. Like I mean, we're not talking about future stuff here.
[00:13:19] Paul Roetzer: This is the tech that you can go get for $49 a month or whatever it is. And so I think it's going to be very disruptive for large content teams, you know, if I think about brands that have dozens of writers or hundreds of writers, you have to start thinking immediately about advancing and upskilling or reskilling that team and thinking realistically about, do you need all those people on that team And if, if you're going to redistribute, or what can those people be doing if you're saving 50% efficiency on the content you creating? Are you just going to create 50% more content? The same amount? Like there's major decisions to be made around this stuff and I just don't feel like people are ready.
[00:14:03] Mike Kaput: Yeah, it seems like step one for many, many, if not the majority of people, is understanding that this is here.
[00:14:10] Paul Roetzer: You've gotta go try it. Go get a tool. Like we say this all the time, you have to experience it. Go test a few GPT-3 power tools and see for yourself. This is not BS. We are not overhyping what it can do. This is legit tech that does this stuff now, and it's only going to get better and it's going to happen fast.
[00:14:31] Paul Roetzer: You've got to try it, and you have to think about what does this mean to my team and my business model.
[00:14:37] Mike Kaput: So, zooming out a little bit on our next topic, we've kind of just talked about really in the weeds of, okay, here's how this type of technology, AI in general, in marketing and business is going to really impact individual roles.
[00:14:53] Mike Kaput: Now let's look at kind of the macro picture here, because Boston Consulting Group, BCG, recently released some guidelines for how companies should approach artificial intelligence, and they base this on something they call The Responsible AI Leader Blueprint. So this is essentially a set of steps and frameworks and guidelines to start thinking about how to use AI ethically and responsibly. And the way they kind of define that is responsible AI is "developing and operating artificial intelligence systems that align with organizational values and widely accepted standards of right and wrong while achieving transformative business impact." Now, we'll kind of unpack that in a second, but basically they recommend that any organization considering or trying to apply AI, which eventually will be all of them, take four key actions and very quickly, they are one, establish responsible AI as a strategic priority with senior leadership support. Number two, set up and empower responsible AI leadership. Number three, foster responsible AI awareness and culture throughout the organization. And finally, number four, conduct an AI risk assessment.
[00:16:12] Mike Kaput: Now, we're going to talk a bit at a high level about this, but really, here's the kicker: BCG isn't just publishing these guidelines on a lark or for itself. They say these are more necessary than ever given the imminent arrival of the European Union's AI Act. Now, this is one of the first broad ranging regulatory frameworks on AI that from governments that is expected to come into force next year.
[00:16:43] Mike Kaput: BCG actually really expects other governments to follow suit with their own regulations once this act is out as well. So what we have here is a major consulting firm offering advice on responsible AI to help businesses get ahead of what it sounds like will be a near future with much more increased AI regulation.
[00:17:04] Mike Kaput: Does this kind of gel with what you're seeing in terms of the market?
[00:17:09] Paul Roetzer: Yeah, I mean it's definitely what we're hearing as well from our friends and thought leaders in this space that pay close attention to the regulations and the Privacy Act and things like that. I mean, what we keep hearing is just behave as though you're under the European Union's AI Act guidelines, whether you're in Europe or America, whatever, that stricter regulations are inevitable, and you should start behaving now in the best interest of the consumer that you should not try and cut corners.
[00:17:43] Paul Roetzer: Take advantage of lax guidelines or oversight here and just be good corporate citizens when it comes to using this. Now, you know, I had bold faced in, in some prep notes this, widely-accepted standards of right and wrong. Yeah, in today's society, I would love to know what is widely accepted anymore.
[00:18:04] Paul Roetzer: Yeah. I feel like everything is 50/50. What one side thinks is right, everybody thinks is wrong on the other side and vice versa. So that is the challenge from a brand perspective and from your company: what is right and wrong, and hopefully you have core values and hopefully you have ethical guidelines within your organization and moral guidelines for guide your company.
[00:18:27] Paul Roetzer: And so this becomes an easy conversation. But I think the real key for us, and we talked about this in the book and the More Human chapter. From the beginning you need to think about the ethical use of this stuff. It gives you superpowers, which you can use for good or for evil, and it's only going to get more powerful.
[00:18:47] Paul Roetzer: And you have to make decisions that are in the best interest of your stakeholders, your employees, your community, your customers, your partners, whoever they may be. I think just at a high level, while your organization may be still at the phase where they're trying to pick a couple pilot projects, and this seems far off, it's not because depending on the use case, you're choosing to pilot even, you may already start running into some ethical decisions around: where that data's coming from, which model you're using, what company built the model you're using, the company behind the AI that you're choosing to buy.
[00:19:25] Paul Roetzer: It's probably best to go ahead and have a basic framework of your own ethical guidelines in place as you start moving into working with AI technology.
[00:19:38] Mike Kaput: That's a really good point. And, kind of one of the things that BCG recommends, and I know we have recommended in the book and to partners and people, audiences that we talk to,
[00:19:51] Mike Kaput: you mentioned kind of a framework we would call that, another word for that is an AI ethics policy, which is essentially a framework to say how your company views AI, what its position is, what it will and won't do with the technology. Could you talk a little bit about what you see being important in an AI ethics policy and why this even matters?
[00:20:14] Mike Kaput: I mean, some people might hear that term and think this is kind of a vanity exercise, but in fact it's a really important document to not only have but actually
[00:20:25] Paul Roetzer: publish. Yeah, I'll, I'll kind of answer by going through the two examples we use in the book. So in the More Human chapter, we analyzed from the perspective of Google, so like a platform company, cloud company, technology company that's building AI and their guiding principles to what they will build and how they will use it.
[00:20:46] Paul Roetzer: And then actually we do cite BCG a 2021 report they did on elements of a responsible AI policy. I don't know if there's a rule of seven tied to ethics, but they're both seven. So I'll just kind of run through these really quick. So on the Google side, the seven principles that guide its objective for AI applications and you can take these and apply them to your use of the AI.
[00:21:09] Paul Roetzer: So be socially beneficial. It's pretty straightforward. Avoid creating a reinforcing unfair bias. Good guidelines for marketers. In any case, be built and tested for safety. In that case, you're relying on the vendors that they've done it, but you need to also then do it yourself. Be accountable to people.
[00:21:26] Paul Roetzer: That's the human-centered idea of AI. Incorporate privacy design principles, so make sure you're not dehumanizing people within the process and invading their privacy. Uphold high standards of scientific excellence. Be made available for uses that accord with these principles.
[00:21:42] Paul Roetzer: So that's Google's. And then real quickly I'll run through BCG, which they analyze seven generally accepted dimensions of responsible AI, how we phrased it in the book. Very high level. It was accountability, transparency and explainability, fairness and equity, safety, security and robustness of the AI systems, data and privacy governance, social and environmental impact mitigation.
[00:22:08] Paul Roetzer: You don't have negative impact on society and the environment, and then the human plus AI, that the AI is not there to replace the human and you make this tangible in that writer example: the AI doesn't exist to cut your writing staff from 10 to five. That is not why it's built. It's not why you should be using it.
[00:22:26] Paul Roetzer: That is not a human-centered approach to applying AI. The human-centered approach is saying, we have 10 writers. We actually could produce the same output with five. How can we redistribute the other five people in their time to create more fulfilling work to do interesting things we didn't have the time to do before?
[00:22:46] Paul Roetzer: To be a better social citizen, to be a better brand. Like that's what you should be doing it. And so that's why I say you need to start with a framework of an ethical guideline because if you allow yourself to look at the efficiency gains, and say, " Well, we can drop our staff 20% next year." If we can gain 40% on content output, that's the wrong way to look at this,
[00:23:07] Paul Roetzer: and in the end it's a losing approach. So it's just critical that these conversations are headed at a high level within your organization, that you realize what this technology is going to do, and you have frameworks to help you do it in an ethical and human centered way.
[00:23:27] Mike Kaput: Yeah, that's a great point.
[00:23:28] Mike Kaput: And I think as kind of a final thought here, one really important thing BCG mentioned is they encourage people to think of the wide ranging impact of GDPR, which as every marketer knows, caused a lot of headaches, a lot of soul searching, and a lot of sleepless nights for a lot of brands that had to adapt to a European law, even if they didn't, if even if they weren't based in Europe, because it applied to all EU citizens, many of whom they were marketing to. The same thing applies to AI regulations in the EU. I mean, you're going to be using tools that market to certain people in the EU that you're going to have to follow regulations for. You're going to potentially be using data on models trained on data from EU citizens. This will have an impact on
[00:24:19] Paul Roetzer: you. Yep. It's a global economy. You've got to be looking around and understanding how quickly this space is moving and you need to be proactive in doing things the right way. Absolutely.
[00:24:33] Mike Kaput: All right. Last but not least, a really interesting, article came out written by the Partnership on AI.
[00:24:40] Mike Kaput: So the Partnership on AI is a major non-profit. It was founded by major tech companies like Amazon, DeepMind, IBM, Google, and it exists as an industry non-profit to research and share best practices around the development and deployment of AI. The group actually did some work on understanding how AI could change local news, and they actually talked to nine different experts in media in the space, including people from prominent media outlets, people with serious reporting backgrounds, and also technologists at some of the major tech companies and you know, you and I come from journalism backgrounds. The takeaways here should be required reading for any journalist or media professional, and we'll include the full article on the show notes. However, I thought these also kind of shed light on the opportunities that AI presents to transform how businesses at large communicate with audiences.
[00:25:39] Mike Kaput: So I just wanted to share a really quick quote from this work from an expert at Google. Her name is Natalie Gross. She's a product manager of local news experiments at Google. And when asked by Partnership for AI on AI about how AI would impact local news, she asks us to quickly imagine a scenario.
[00:26:00] Mike Kaput: And I'll paraphrase just very briefly. It's 2026 and a person wakes up in London and is greeted with a tailored news summary specific to their day and their interests. First, an algorithmically assembled alert that their bus line is likely to be delayed due to road work, and this analysis is created from a merger of London transport data paired with their geolocation data.
[00:26:24] Mike Kaput: Next, they actually get an advertisement from a local venue that a band they'd like to see is playing tonight, and tickets are 10% off. A reader-reported insight comes up next. There was a car break in very close to their apartment yesterday, but the data widget beside the insight puts in context that car break-ins are actually dropping in their specific neighborhood.
[00:26:45] Mike Kaput: Finally, they see an in-depth article about air quality in London, a topic they read about often by their favorite journalist. So Paul, you graduated with a journalism degree. I worked for a bit as a journalist. This is a topic near and dear to our hearts. What is your perspective on AI's potential impact on local news?
[00:27:06] Paul Roetzer: The journalism schools aren't ready. I mean I think about this all the time. I yeah. So you said, I came out of journalism school, I stay in communication with them. I pay attention what's gone in the space. And you know, I've written about this before a couple years ago, that the media industry, the local news in particular, I mean, our local news in Cleveland has been decimated.
[00:27:32] Paul Roetzer: The newsrooms are almost nonexistent. And my dad worked for the Cleveland Plain Dealer, so like when I was growing up and I would go to the printing presses like, I still remember the smell of like the papers coming off the press and I used to ride the truck with them and deliver papers like I around journalism of my whole life.
[00:27:51] Paul Roetzer: There were like 350 writers at the Cleveland Plain Dealer when I was a kid. I don't know how many there are today, but I think it's approaching about a dozen. So local news from The Plain Dealer is not coming from local journalists necessarily. And then you start looking at the local papers that existed and those are a shadow of their former selves.
[00:28:12] Paul Roetzer: So the industry has already been completely transformed and that had nothing to do with AI, that was just a pure business problem when you add in AI's ability to do what you just described, that excerpt from the product manager of the local news experience at Google that it's hard to comprehend what local news looks like, or what journalism looks like in what, four years from now they're talking about? Yeah. I mean, as a reader, that sounds amazing. Like if I wake up and it's just sending me that in an email or I just ask Siri or Alexa, read me my local news for the morning and you know, a minute later I've got all this context, or I get my car and I just hit a button and it starts talking to me and telling me all this stuff, that sounds amazing to me as a consumer.
[00:29:03] Paul Roetzer: As a business person, as a journalist, it's like, well where's the human's role in that? Because what you just described, I don't need a human do any of that. All that data can come from data sources and can be processed and generated as text or voice with zero human interaction. There's nothing you just described that needs a human writer.
[00:29:22] Paul Roetzer: So I don't think that there's enough conversation happening in the world of journalism about this imminent future, and I don't know that we're solving it, especially for the next, phase of writers coming out of schools who are starting. I think about freshman at journalism schools today, like four years from now.
[00:29:43] Paul Roetzer: This is it. Like, where's their job in in that future? What are they doing? I don't have the answer to that, unfortunately, but it's a question I think about a lot and I think we maybe could be doing more to help in that area. And I think it just needs major, journalistic outlets to be challenging it. I'm glad to see Google's at least thinking about it, but they're thinking about it from a technology perspective, not necessarily the human impact. I'm not saying they're not, but I'm worried about the human impact. The tech's coming one way or the other. It's the human impact that worries me.
[00:30:22] Paul Roetzer: That's
[00:30:22] Mike Kaput: a great point. And as we kind of wrap up here, I to kind of bring that home, I'm curious what kind of areas of local news, given your experience there, do you see as particularly ripe for either AI disruption and or AI enhancement?
[00:30:37] Paul Roetzer: The enhancement, I mean the disruption seems obvious, the enhancement one is interesting when you think about the ability to train these models on specific journalists or publications. So if I'm a journalist, been doing it for 20 years, say I'm a sports beat writer or local news writer, business writer, and I can feed it the 10,000 articles that I have written, the AI can actually learn my voice and my style and, you know, the nuances and the way I write and everything, and it can start to output that.
[00:31:08] Paul Roetzer: And again, you could look at that as well aren't I just replacing myself? Maybe, I guess if that's the perspective you take, but if you think about that assistive writing engine, that's not just trained on all this general knowledge, it's trained on you, and now the AI not just helps you write, but it helps to write in your style, then maybe you can now put more, or maybe you focus on the editorial stuff that's different.
[00:31:32] Paul Roetzer: So, I think that's the key is just that there's a world of opportunity people need to think about that's beyond just the automation of this stuff, and I think it opens up some interesting opportunities for people.
[00:31:46] Mike Kaput: That's awesome. And I think that is a really great place to kind of wrap that discussion.
[00:31:52] Mike Kaput: I mean, things are moving so fast. It's really kind of a diverse array of topics this week because I wanted to really show that it's not just image AI or text AI. This stuff is permeating every area and every asset of our lives. And it's just incredible. And I think we really have an exciting time ahead. A scary time ahead. And I think our rapid fire topics here that I want to throw at you, they might just describe to us exactly how fast some of this stuff is moving. And, you know, to kick that off, there was a series of tweets from a guy named Alexandr Wang, who is the CEO of Scale AI which is a prominent AI company.
[00:32:35] Mike Kaput: And what he actually was promoting is that Scale AI in a few weeks is releasing a tool designed specifically for marketers that lets you generate unlimited creative for product ads, marketing campaigns, and social media posts. He said in these tweets, we're enabling marketers to AI generate unlimited and infinitely creative images of their products for AI creatives, brand campaigns, and social media.
[00:33:00] Mike Kaput: Simply provide your product and where you want it to be placed. And he basically is saying AI will do the rest and create all these various generations of product images in whatever scenario image, place in the world, environment you want. And he says AI can give every marketer superpowers, and we're excited for that future.
[00:33:21] Mike Kaput: What do you kinda make of this?
[00:33:25] Paul Roetzer: The examples we're going to start to see will just blow people's minds. We talked about it I think last week about architecture, room design, product design; it's just going to be whatever your imagination allows you to think about creating, you're going to be able to create.
[00:33:43] Paul Roetzer: Or some entrepreneur is going to have the vision to build that tool. That we call no-code tool. You have no need to understand how the AI works or what it does. You just need to be able to explain to it what you want in a very concise way, and you get the magic of this. And I think Scale AI, if I'm not mistaken, I was just watching a YouTube video last night, they just announced a partnership with Stability AI, who we talked about recently that raised the $101 million to feed some of their data into the Stability AI models. I forget the exact details, but Scale aI is a major player in this space. And then it made me think of, I don't think we have this one on a rapid fire, but there was another one I just saw.
[00:34:26] Paul Roetzer: Oh no, we do the Runway ML one. I don't want to jump ahead, but I guess I'll jump ahead. Yeah, let's go for it. In that one, it's the same idea of like what they're doing is you can create an image, but then you can keep creating. So it's like, show me a bird and a tree in a park in New York. Now add a street landscape around it.
[00:34:44] Paul Roetzer: Okay, Now put buildings around that and you just keep telling it what you want and it can infinitely generate off of that first image. Yep. So again, this stuff is four months old like this tech, The idea of generative AI and image generation is new to the world, largely outside of AI research circles that have known about it for a while.
[00:35:05] Paul Roetzer: But the average business person and marketer had no clue what this stuff was before May, and all of a sudden you look next day, and something else will be like mind blowing. So it really is insane to be watching this space right now. You and I have been theorizing about this stuff for 11 years and now all of a sudden it's moving faster than I probably ever imagined it could.
[00:35:31] Mike Kaput: Yeah. I think that Runway example is really great because like you said, that was specifically about their infinite image generator, which infinitely expands any image in the same kind of style seamlessly. But to your point about the speed, if you go to runwayml.com, their website, you'll see that that infinite image is just one of many different tools that they offer as part of their content creation suite. And if you go under product, we have everything from video editing, green screen motion tracking, text image generation, image generation text to 3D texture you can erase and replace based on text prompts you have at your fingertips.
[00:36:14] Mike Kaput: This crazy, powerful AI powered suite of content creation and editing tools that, I'm sure they've been doing this work for several years, but seem to have almost materialized out of nowhere. And the pricing on these is crazy. I mean, the team license is like 300 bucks a year, $28 a month. Yeah, and that's probably going to even go down, I would guess as competition occurs.
[00:36:42] Mike Kaput: It is wild what you have at your fingertips on a timeline that is just breathtakingly
[00:36:49] Paul Roetzer: fast. Yeah. I mean it really is. I remember Runway ML, I learned about right before our Marketing AI Conference, MAICON 2019. The first one we did, and I actually, in my opening keynote showed, a process with Runway ML where I took my daughter's Super Everything Girl drawing and uploaded it and then I turned it into a Monet and a Picasso. And so I showed, as an example of AI, look, I just took my daughter's thing and like reimagined it as these famous artists. So we've been following Runway ML for years. But yeah, I mean the last few months it's just blowing up.
[00:37:31] Mike Kaput: I think this last rapid fire link I've got for you kind of really wraps this all up and sums it up. It's Sequoia Capital who we talked about last week a little bit as well, who are very big into Generative AI. They actually released a market map of possible generative AI tools. The first one came out a few weeks ago.
[00:37:51] Mike Kaput: They're already on v2, and we're looking at a logoscape of essentially prominent generative AI applications in different areas, and it's constantly updated. It's got a easily over a hundred tools on the current one, but what I thought was really interesting is they're already starting to break up generative AI into interesting categories. So you've got your basic ones like AI that can generate text is its own category. Video is its own category. Images, code, speech, 3d, and then other. But what I found particularly fascinating, especially for what we do, Is within text, which is an area we've talked about a lot, and have seen massive, massive innovation in the last 12 months.
[00:38:36] Mike Kaput: They're already starting to break up solutions by marketing specific. They have a lot of tools that we've profiled on here, including Jasper Phrase, Writesonic, and Writer. They have also categories specifically for sales. Specifically for support and then things more like general writing, which I found really fascinating.
[00:38:56] Mike Kaput: What did you make of this? I know we had discussed it a little bit internally.
[00:39:02] Paul Roetzer: First, everybody loves their logoscape. My God, this thing was blowing up on my LinkedIn feed, at least. Everybody's sharing it. It's just, it's always so funny to me, such a simple concept of just putting stuff into things and people just eat up logoscapes every time.
[00:39:17] Paul Roetzer: That being said, it is a really interesting logoscape and for many of the reasons you just highlighted, the one thing I'm thinking about as I'm looking at it is, not only are the number of these companies going to expand dramatically, like Y Combinator's just going to be pumping out language generation companies like it's like their job, which it will be. So that's going to expand. But then you look at like a Cohere, which I'm really hot on, and they have them under support chat, email. Well, cohere is going to be huge way beyond this. I mean, you could make an argument that Cohere might end up being the winner in the language space based on its backing and who's, who's involved. You could see DeepMind make a play in here. You could see Meta get in here, You could see Google get in here like I mean, they're going to have a heck of a time keeping this logoscape properly updated. It's going to turn into the MarTech landscape of 9,000 logos within 18 months,
[00:40:17] Paul Roetzer: I feel like . So I think it is a very helpful visual and I think it helps people to start to wrap their mind around all the ways these tools are going to be used. And I think, as you said, going from V1 to V2 in like the first seven days or however quick it happened, is just an example of the, the pace of change here.
[00:40:39] Paul Roetzer: And, I don't know, like buckle up. It is just going to be a wild time in marketing and business.
[00:40:48] Mike Kaput: That's awesome.
[00:40:50] Mike Kaput: Alright, that's all I've got on the rapid fire side. Paul, both on the kind of main topics and the rapid fire, really appreciate your thoughts and your perspective. I think you bring a really unique perspective to all of these topics and I would say our audience probably agrees, but yeah, I think that the advice I'm taking away from this week as I do many weeks is buckle up.
[00:41:11] Mike Kaput: Things are moving really
[00:41:12] Paul Roetzer: fast. I mean, I'm sitting here like, god, I can test some more tools. We have this conversation, I don't know how other people feel when they listen to it, but like we go through this and I just have a list of 10 things I want to go do.
[00:41:26] Paul Roetzer: I'm just looking at the text section of this landscape and thinking, oh, there's a half dozen on here I haven't tested yet. I wonder who these companies are. And I'm like, oh wait, we talked to those three last week. That's interesting. So yeah, it's such a fascinating space.
[00:41:40] Paul Roetzer: And the think my final take would be, if you figure this stuff out or you have the passion and the curiosity to stay at the cutting edge of what's going on and how quick it's moving, you are going to have insane career opportunities. Like if you get this stuff and you can go test a few GPT-3 tools and you can make business cases to infuse things within it and find image generation tech and infuse that, you're going to be unique within your business and you're going to open up doors for yourself. The other thing I'll tell you, and this actually came from a conversation we had earlier today, if you are working at a corporation That doesn't want you to push forward with these ideas and doesn't want you to drive efficiency and productivity, because it makes everybody else look bad,
[00:42:30] Paul Roetzer: I would go find a different corporation to work for that embraces this stuff because they're the ones that are going to be around five, 10 years from now. The ones who refuse to find improvements in productivity and creativity and decision making are going to struggle to remain competitive. So just seek the knowledge, test the stuff, and find ways to advance your career.
[00:42:52] Paul Roetzer: They're going to be plentiful, in, in the months and years to come.
[00:43:02] Mike Kaput: That's awesome. Well, thank you again for your thoughts. Really, really appreciate it. And thank you to our audience for your time and attention. As always, we want to thank you for being a part of the community and being involved with Marketing AI Institute. We're all trying to figure this out together.
[00:43:20] Paul Roetzer: Thanks everybody.
[00:43:22] 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.
[00:43:43] Paul Roetzer: Until next time, stay curious and explore AI.