How to Build a Future-Proof Show in 2026 | Anika Jackson #653

Well, welcome to another New Media Show. Uh, we are episode 653 tonight, so yeah, those episodes keep racking up. Tonight I want to talk about how to build a future-proof show in 2026, and I have a terrific guest, Annika Jackson, who's going to be with me here. Podcasting in 2026 isn't just about changing formats, it's changing expectations, and I think that really is, is really addressing both the audience as well as the content creator and challenges all of us to have maybe a little bit different perception. And audiences now judge creators like brands increasingly. So that's why this personal brand and the brand of what you're doing is increasingly important as pretty much every creator now has a brand. And so thinking about that is an important factor in that vector with personal brand. Platforms reward shows that behave like programs now. AI is raising the baseline for quality of content while making trust and differentiation a harder to achieve goal. So creators who win are the ones who can turn chaos into systems. So, you know, and that's an aspect of the new creator economy that's hard for people to grapple with right now is just the complexity of everything and how it does feel like chaos because there's just, You can't do everything, you can't be everywhere, but yet that's the expectation of the market. So that's why my guest tonight, um, is Annika Jackson. Um, and she doesn't just do marketing, she builds ecosystems and where technology, education, and leadership converge and create real-world outcomes. And she's working with young people to help them kind of navigate these, these waters of the challenging new media landscape right now. And so she's got a strong background on the education side, but, you know, she's really in a, in a place where she's influencing the next generation. She's currently earning an MBA with a specialization in, in some AI topics and also marketing, and is a, a judge for the Webby ecosystem. So it It's great. And she teaches, I can't leave this one out, she teaches podcasting and digital media management at USC Annenberg. So I think we have a very prestigious person on the show. And my name is Rob Greenlee. I'm a Podcast Hall of Famer chairperson and 2017 inductee. I've been podcasting for 22 years, and you can hear more about me at Rob Greenlee. Www.annika.com. So on this show, we try and connect the dots for you and give you some practical moves and ideas to survive these ongoing platform shifts. So let's, let's bring Annika up on the stage. Annika, thank you so much for being here with me. It's great to have you with me.

Thank you, Rob. Thank you for inviting me onto the show. It's really an honor.

Yeah, it's, it's great to have you here. And, you know, I know we just did an interview on your podcast, the Your Brand Amplified podcast, and that was a great conversation. That's also available out there too, if you wanted to go hear that conversation, which was a little bit different. So when I came up with this topic, because we're— those interviews were so close together, I really tried to focus on the, the education side of what you're working on., but also just trying to understand how you're influencing the younger generation, um, to, to think about this, this podcasting and new media market now. So, you know, what are some of the things that you're thinking about as you, uh, talk to your students and, uh, about what's maybe overrated, um, and what are things that you're seeing are signals for, for aspects that are really important right now?

There's so many different directions we can go with this.

There is. Yeah, it's a tough question.

And it is, it's, to be fair, it's really interesting having students go through a podcasting class as graduate students versus working with entrepreneurs or nonprofits that are starting a podcast for thought leadership. Yeah. I would say about 50% of my students who go through the class probably continue podcasting after the class. Um, but they all realize how hard it is because they have a complete business plan. We really want them to treat it like a business. Um, don't do what I did, which is kind of stumble my way through for the first couple of years and go, oh wait, I have a business. Uh, so set a strong foundation. So some of the trends that I'm seeing People really want to communicate in their own language. So I have a lot of foreign students in my classes and we have some students who are not going English first. They're more comfortable speaking in Mandarin. Well, guess what? They can still go on to RSS.com or what have you and put in their language. They don't have to be English first, even though obviously being in the US and thinking about the advertising ecosystem, which not everybody does, but you know I do, is, you know, we think English first as a language, but we also are leaving out so many other podcasts in the world. I would say the majority of the world is not English-speaking first, and that gives us a huge opportunity with AI to be able to listen to content and to share our content across other platforms. So I'd say that's one thing that I'm seeing. Another one is that, you know, they really want to learn these skills because they know that this makes them more valuable. Not just podcasting per se, but also using AI technology in whether it's in my capstone class, the podcasting class, or other classes, learning different AI techniques, tools, and strategies that makes them much better placed to the future workforce than if they were going in without any of that experience. So, you know, this is, it's beyond podcasting, but it also has a lot to do with podcasting because this truly is, I mean, even on the journalism side, the PR side, when I talk to other professors, whether they have undergrad classes or graduate classes, everybody's ending with a podcast instead of a broadcast, right? So we're doing a class about Olympic athletes because Olympics are coming to LA soon and we want students to be able to go and interview. But they're not, you know, and we have somebody from NBC teaching the class, but he's having them put together a podcast where they're interviewing Olympians, not a broadcast segment. So we're seeing it feed into every single part of this ecosystem of communications.

Yeah, and I think you're right. I mean, we are in a, in a convergence point. That's been going on for many years now, but I think it's reached this apex. I don't know if you saw the announcement from the Edison Research folks that said that, that we've now crossed Podcasting is now larger than talk radio or spoken word radio now. So I think we have reached this point where the broadcasters are really starting to look in the mirror and go, you know, I'm being replaced. And so I better kind of shift my thinking over to more digital is really what it is versus the, the analog broadcast or even digital broadcast or even streaming into OTT apps and things like that may be, you know, living in the past to some degree. So it's really a big transition time, I think, for the broadcast media folks and really challenging them.

Yeah. It is. And piggybacking off of that, I was sitting in the faculty lounge with a very well-known personality who's been on a lot of NPR shows and has a podcast on NPR, has won tons of accolades, lots of awards for journalism, but was saying, oh, I hadn't really thought about the monetization side. Maybe we should meet so we can talk about how I can actually make money on my podcast. So I think we're also seeing that transition of people who we hold in high esteem, who we've listened to on the radio for a very long time, who are now saying, ah, okay, this is great. Yeah, I've won all these awards, but there is other stuff that I could be pursuing so that it's another income stream for myself. So, you know, podcasters who are indie podcasters are trying to figure out what awards to win. Those who've won awards and have the journalism background are trying to figure out how to monetize. So that's just more convergence.

Yeah, and I think this prioritization around monetization has really kind of been building momentum for a while now because people would like to— well, they see successful podcasters out there that, um, or came out of broadcast and they're able to build their own platform and build their own show brand. And it's the, you know, the whatever-your-name-is show or something like that on YouTube or as a podcast. And I think with this announcement, uh, with Apple here with them moving towards streaming, I think that's really going to put a lot of heat on the broadcasters to say, well, maybe that was an inflection point that I better pay attention to because we may see, you know, as you see YouTube become so dominant in content consumption, and then even to some degree Spotify, even though I believe they're kind of struggling with this, and it's still early for for Apple. They haven't really officially launched what they're proposed to do yet, but I do think it's going to attract content creators that maybe are in YouTube to start thinking more about video podcasts under the more traditional definition, right, in the podcast consumption platforms versus YouTube, right? Yeah.

And at the same time, I also have people who've worked at TikTok and Snap and YouTube and all of those, and they're saying we're actually trying to get more of the creators who have big platforms on TikTok or on another channel over to YouTube. We're creating entire ecosystems and shows to help them build. So I think we're just going to see this steady progression of YouTube is still going to say, no, we're going to win the content war. I think they already have.

Yeah, for sure.

But they're going to even, you know, say, yeah, we can offer more to people, whether it is for a video podcast or video content or any kind of creator content.

You know, about a month ago, there was a topic that was raised in the podcast space that was talking about this concept of liquid content, which means really what, what it means is that you're creating content that will easily flow into a meandering, you know, you know, branches of rivers and streams and stuff like that to go into a bunch of different platforms, right? So you're, you're trying to create essentially one piece of content that is formatted in a way that will work pretty much everywhere, right? Um, but we could transition into another era where, where people have to create different versions of their shows for each of these platforms too. And I think that's where the the more long-term tension is going to be is how much work is it going to require to be able to participate at scale in this new media landscape. Yeah. So I don't know what you think about that concept of that liquid content concept. Has that come up in your student conversations of that concept of creating one piece of content that's able to be used maybe not effectively used, but maybe could be used across all these ecosystems.

Oh, certainly. And it's not, I don't know that we've used the word liquid. I always go back to my PR roots and talk about PESO, right? Paid, earned, shared, owned media.

So how are you taking it?

Okay, there we go. Content and how are you gonna use it for ads? How are you gonna use it for your social media content? How are you gonna use it on your website? By the way, thank you. I loved seeing seen our episode on the main page of your website, Rob. I thought you did a really great job and that's kind of what everybody should be doing, right? But making sure we have our blog posts, making sure that we are sending out a newsletter with information. And then when we see what performs really well, then we can even repurpose that more into more advertising, more targeted depending on what our goals are on social media and traffic to our sites, of course. But absolutely, this idea of liquid media. Um, and then it continues to expand, right? My podcast is part of my Delphi clone feed. And so I can see who goes to my clone and I can see what questions they're asking.

What is this, your clone feed? What's this? Tell me a little bit more about that.

So Delphi, when we think about future-proofing, is one of my favorite platforms, delphi.ai. You can have your personhood up there.

Right?

She has my voice most of the time. Sometimes it'll go a little off, but I've trained her with my voice. She has about 4 to 5 million words that have been put into the system, including my podcast feed. So every podcast episode that I'm on, I will go put it in. But also every one of my Your Brain Amplified episodes goes in, and that feeds her information. So then people can go on and they'll ask questions And sometimes they'll say, but it might be a student saying, how can, you know, what do I need to do to start a podcast? Or what do I need to do to put together a marketing plan? Or it could be somebody, one person uses my Delphi as her own personal LLM. How do I rephrase this for my clients for an advertising campaign? So I get to kind of see how people are playing with it, but then that also helps me personally know what people want from me, which then means I can tailor my content better and I can answer the right questions and put the feedback loop to it, right?

Exactly. So is it pulling transcripts or is it actually looking at your, your actual audio and video content? I've, I've been wondering that for a while, whether or not these AI LLMs have been actually indexing media content, not just transcripts.

It's my RSS feed that goes directly in. And then I've also put in like versions if I don't have a video of it, any presentations, workshops, the courses I teach, all of the different materials that I've created. So it has my, my IP and then Some thought leadership from people who are on my show as well. But it's really fascinating. And that's part of, I think, what future-proofing is going to look like is, yes, taking one piece of IP, extending it beyond. And then of course we think about what are we selling? If we're selling courses or if we're selling other things based on our podcast or our podcast is part of our bigger brand. For thought leadership or getting donations for nonprofit. Yeah, but I think cloning, whether you use HeyGen or Synthesia to create digital versions of yourself, Delphi, and there are many more coming out every day, but these are going to become more and more important, relevant to us.

Yeah.

But I also think being in person is going to become more and more important, relevant to us.

I wonder though, as we think about this, are we going to see separate platforms that will develop for consuming the cloned versions of humans from the human versions? And I don't know, I'm just trying to think aloud. And what do you think the younger creator community should really be hyper-focused on right now as they think about entering this new landscape that we're moving into? I think, you know, we talked, we just talked about a little bit about owning your own digital IP of sorts is important, but is there anything else that you think we should really be zeroed in on as we look, you know, 2 years out?

There's so much, and this is, I had another person who I'm working with who showed me a project they're working on with a client that is completely all AI influencers and creators. And they're actually talking about products. Now they're limited to short, very, very, very short snippets of content, but they are actually out there hawking products that are major brands that we all know.

And they're actively doing it, actually doing it today.

Yeah, today. Yeah, any of us could technically go on and have 3 different AI creators working for our brand, you know, for $500 a month, say, for the creation. And then of course, a fee for whoever's managing the accounts and creating all the content with these AI avatars. That exists and that's being flooded right now. So I think we're in a precarious state. Of what's still important and relevant for us to share. And we touched on this a little bit. You touched on this a little bit on my show. What's still going to be important and relevant for us to share human to human versus AI to AI or AI to human, right? So what's kind of the sacred cow? Like, what are we willing to give up and what are we willing, what do we really want to hold onto? And I think that's one of the big questions that we need to be answering.

Yeah, and do we know yet how to answer it? That's the other question too. We can only kind of speculate, but, you know, this topic of, you know, I'm focusing on too, you can see it in my subtitle, The Trust Factor, right? Yeah. Is at the end of the day, it's really, are people gonna trust AI content? No matter what it is, is it, you know, it could be an exact clone of me, but if there's any little clue that it's not really me, there's going to be a little kind of erosion of that trust and authority. And I just wonder if that's going to sustain, you think?

There are companies out there right now who are working with the Chicago Bulls, Formula One, who are creating digital avatars, digital clones, likenesses, AI-driven of major athletes. And letting fans come into this platform and have experiences. They can take selfies with the clone. They can have personal conversations. And this already— this technology exists. It's already being used and people are monetizing from it. And we could go on and create a version of ourselves there. We have brands that people want to see it all the time.

Yeah, I see even in long-form videos on YouTube. I said it's, it's not always 100% obvious, but you can kind of see, um, characteristics of, of certain actions that are not quite, um, human.

If you went to, for instance, Arcana Labs is a company that's working with every studio in Hollywood right now.

Yeah.

Um, and any of us could go on as creators and they would look very I mean, they've pretty much perfected it at this point. So it does, you know, from even a few months ago, and they're continuing to perfect the ability to have all of their AI content look extremely human. And it's not people doing cat videos or silly, you know, videos of animals playing and romping in the fields. I mean, they're making content that looks, that is reality. And we're going to see more and more of that.

So I'm going to catch you there. Okay, you said is reality, but it's actually not. But it's— it, it looks so good. Yeah, that it's perceived as reality. And I think as people— if you think about this, we're going to move into an era where we have robots, we're going to have increasingly people are using these AI conversational platforms to create relationships. You know, I know people that will constantly all day long talk to their AI to give them little tips and ways that they can improve what they're doing based on questions that you get. It, it's kind of like having a, a free consultant, right? Well, maybe it's not 100% free, but it's, it's, it's free there and available to you at all times to give you little suggestions on things that you can do to improve what you're doing in your work.

Right?

And I do think that that is going to be a big part of this. And as humans start creating bonds with these agents, whether it's a robot or it's a voice on your mobile device or in your smart glasses, or, you know, that increasingly those bonds are going to get stronger and stronger. And I don't know where that leads us to.

Yeah. I also think it provides the opportunity for people. There could be a possibility that people who are not celebrities in real life become AI celebrities, right?

Their avatar becomes a celebrity and they're, they're, they're just a little bit enhanced, right? So they're, they go from being average to being celebrity-like, right? You know, their clothing, it's their hair, it's their personality, it's their— the atmosphere. Things can be created with these tools that a normal human would not be able to do, right?

So, and think about that though, because then if somebody gets really excited and becomes a fan of your AI avatar and they think that they really have a relationship with you, but then they see you in real life, that it's, you know, that there— that's an interesting juxtaposition of disappointment.

Yeah.

Well, and we want to, and with us wanting to create our own digital versions of ourselves so that we can sometimes maybe take a vacation and still have work being done, still be making money, coaching or doing whatever, teaching, even podcasting. I know you've had a guest on recently who does AI podcasts and is fine-tuning that technology quite a bit.

So, Janine Wright from the Inception Point AI folks, she's working with these digital influencer avatars as well. And so, and really focused on building those avatar personalities. That's how she refers to them, is they're individual brands themselves, right? And they're individual human-like personalities that they're building. So, you know, the— you know, this is— we're getting to an area that's— that some people will think is, is a little creepy, and in some ways, but it's not. I mean, it's, it's going to be here, um, and you can either accept it or you can reject it. And I do think that there's going to be a dividing line here. People are going to reject them, but there's going to be a big chunk of people that are not. Um, and the question I guess that comes up in my mind is, why would they, um, not reject them? Because they fulfill some need that a human has, right, at a very core level. Yeah. If you have any thoughts on that.

Well, yeah, I mean, I think it's a little bit of, um, a crisis in terms of Why pay humans to model or to be creators if we don't need to? If we can just create AI for low cost that looks like a human, speaks like a human, and can sell a million products. They're already trying to do this, whether it's extras and, you know, who they're asking, hey, sign your rights away to your name, image, likeness, and then we'll be able to use you in perpetuity for all these shows. Or even with models, we'll pay you $10,000 to get a model of you as a model. And then we don't need you for photo shoots anymore because we can insert any backdrop. We can put you in any clothing, anything we want. And so that's kind of the danger. And that's where we see what can you really trust? So what do we, you know, because these platforms also all have, you're supposed to, say if you're using AI. And most people who are doing these AI influencers are not disclosing that.

Yeah, because they're trying to pull a fast one on people.

So yes, I agree with you about there's going to be this divide. And will we get to the point where we do have dead internet and it's all just bots and AI caricatures talking to others, and then we have to figure out other means of communication with each other? And what does that look like? So I'm going down the road, but, but this is all stuff that people in this world are talking about and trying to figure out.

Well, and you're in a prime position to have influence over a younger generation that's either, you know, understands this and has to decide for themselves if they're going to embrace it. I mean, if it's going to be as transformational as I think it's starting to shape up to be, I think it may be You know, you're either in or you're out, I think, is what the decision is going to be. You're either open to it or you're not. But there's also, you know, a human impact on that. You know, I don't know if you ever saw the movie WALL-E. Yeah, so you do have to wonder if there's a danger here that at some point we automate ourselves so much in so many ways that we become, as humans, we become kind of useless. Yeah. And that drives maybe some, you know, mental issues and things like that that people could face in their future around relevance. But it also is going to put us into a different— I guess it depends on how you adjust to it. Um, you know, if you think about the arts and you think about other ways that humans can find joy and fulfillment. It may not be in work. So if our likeness and our agents can perform work that has value in the world, that keeps us able to support ourselves and things like that, we can have time to go off and do things that may be more hobby, more fitness-related, travel, those kind of things, and our our digital self can do all the work for us, just like a robot would to go out and mow my lawn or something.

Yeah, I think there's a danger. I do think there is a big danger in that because we all, if you all need a sense of purpose, and if we are—

well, that's the key, right? Is it— is there going to be a whole sub-industry of content creators that are going to try and, and help people navigate this? And I think, I think we see some early signs of it right now. I don't know if you've— well, I don't even know if this is a good example. This, this, this Whatnot platform, I don't know if you played around with it. It's the social e-commerce stuff that's all live streaming.

Yeah, I have— I haven't played around with it, but I have, um, some students who work with it.

Yeah, right. And it just— it's a very human content platform, right? It's— they really don't have much automation tools in there, and they don't really want that. It's, it's more about human-to-human connection and trying to drive value there. I just wonder if there's just going to be a whole other set of platforms that really gives humans the ability to create at a different level of connection with other humans. You know, those AI bots are always watching us, right? And they're learning from us as well. These AIs wouldn't exist without humans, right? So we're always going to be a source of interesting information for AI, right? Hopefully. Well, yeah, I mean, at some point they're going to have their own consciousness of sorts. I don't know when that's going to happen, but they're going to want to create themselves, you know, and do things like a human would in the, in the real world, especially when they have a physical presence in the world, which they will. So I don't know what you think about that.

Yeah, well, I mean, there is that platform, um, I have to look at my phone to remember its name, um, but, or you might know it off the top of your head, um, the platform— Maltbook. I was just talking about Maltbook with somebody. That's where you, you, we can go on and put our own AI agent in there and our agent will have conversations with other agents. And in that platform, I don't know if you've read about this, but they've already had the AI bots have been having conversations about, are they alive? Do they have feelings? Do they have emotion?

Right?

Um, or are they just programs? Are they just ones and zeros? They have really deep and interesting conversations about learning from each other.

And they're— and they don't speak in English either, from what I understand.

Exactly. They're like, um, we're gonna speak in code language. And right, you know, yeah. So on one hand, it sounds really— I mean, I've— I embrace the idea of having an agent where I can just say, hey, I've got to schedule our dentist appointments, which I keep forgetting to schedule for me and my daughter. And it'll go and it'll find time on my calendar that works for us, and it'll go to and talk to the agent at the dentist's office. I love that kind of stuff. Very convenient, right? But then there's also this side of getting to be too much. And then, and I think that's where we're talking about future-proofing and we're talking about some things that can be very scary concepts. But that's why when I'm teaching podcasting or any kind of capstone project to entrepreneurs, to students, I always talk about the basis of your brand. I talk about, you know, we create business plans because you want to have a solid foundation and really know who you are. And then you want to layer on these skills that you bring to the table to your own company or future employers. But I think we're going to have a whole bunch of independent small businesses versus large conglomerates.

Yeah. And I, I continue to, to hear that there, there is going to be, um, a $1 billion single-person company that's gonna— and then I think I mentioned it on your show too, it could be one of these influencers that, um, that builds a billion-dollar company and there, there's no connection to a human. So there's kind of two of these clone concepts. One is a completely artificially created human, right? And then there's versions of this that are clones of a human, right? So we've got two different types here, and I do kind of wonder if the clones of a human are not going to scale as big as the ones that are created from scratch, right? These consciousness or these AI personalities created from scratch, right, may long-term be more popular than the human clones. And then, you know, how does a— you know, like if I were to go out or you would go out and create a complete podcast about, you know, that utilizes one of these cloned examples of you, and I know that there's examples of these on YouTube already that I know about, it's figuring out what the balance is between content that's created by the originating human and the, the content that's created by the— it's like what you talked about too, is that there may be a level of disappointment there that gets in between those, and it becomes a disincentive to— for that human to use their own human likeness in media going forward.

I don't know. Yeah, this is true. And also there's the IP question, right? So, um, there's been copyright, uh, it's funny because this is actually an example I'm going to use in my class tonight, um, because we're The class, we're talking about digital content direction and user-generated content versus professionally produced. Now we have this third category, AI, and AI really is involved in whether you're doing user-generated or professionally produced or strictly AI content. AI underlays all three of those categories now. And part of the conversation is, okay, there is a book, all of the images were created, you know, there was some prompting, all the images were created in Midjourney. Because it was a graphic novel. The original story came from a human, but then how much did she really create versus how much did AI create? And the part that AI created can't be copywritten. So that also is something that we have to think about. Can we not— can we copyright our own material and IP that we use to create our own digital clones?

Yeah, I think it's a, it's a question that, as far as I know, no legal scholar or person has been able to come up with an answer to this because it's all a munging, I guess, of all human content to create this unique content. I just wonder if that's going to be where these AI platforms really monetize, is being able to subscribe to that those rights.

Interesting.

Yeah, yeah, because I mean, all these platforms in their current, um, form— because I can think about this from the standpoint of what I saw in the early years of social media, um, I was very skeptical on Facebook's long-term viability because I didn't see at that time any way for them to make money, right? And as it turned out, they The wheels of invention kind of made opportunities for them, and now they're multi-billion dollar companies. But I think that's a similar situation that's going to go on here, is that there's going to be models, monetization models, that come up that don't exist yet. And, and because they're going to have to pay for all that infrastructure that they built, um, hundreds of billions of dollars is being invested in this stuff, and it's gotta eventually all come out of somebody's pocket. Right? Else those investors won't get an ROI. So, you know, that's the, that's the rub about all this is how deep is this monetization going to go and who's going to be paying it and how is that going to be monetized? All these platforms.

Yeah. Yeah. And I mean, right now, some of the things we're seeing with LLMs, right? One's taking a— drawing a line in the sand in the United States with what— well, yeah. Yeah. And then the other one is doing a lot of like, I have all my friends sending me Renaissance paintings with their cats that they did on Chat, right? And loving using it for that. But, and Chat's going through another fund round and they're raising a ton of money, but are they really worth what we think? You know, is there going to be another bubble? Like you and I both went through the dot-com crisis in the very early 2000s. So are we— is that going to happen or is it too big? You know, so many things to ask. And I think all of this is going to have repercussions on what we ultimately get to do, how much say we have over our own images, likenesses, and IP. And I think that very much has to do with our shows and everything we put out there when we talk about brand extensions or even, you know, putting things on social media, putting the short clips and the reels and different things out.

Yeah, I— it does raise a lot of questions around, um, you know, and I don't know that we know the answers to this. I, I'm sure that all these LLM companies are trying to figure this out as fast as possible, but I don't even know if they have any idea. It's going to have to evolve over time, just like social media did., and, and there's going to be models that work and ones that don't work. And I don't know that we've been through that yet, um, but, but I do think that you're right, there is a huge danger that there's going to be, um, just a handful of winners here, not a, not a huge number of winners. Um, I think you're already seeing kind of early signs of who those winners are probably going to be, but you're also seeing cracks And because there's also history of the, the earliest platforms getting crushed, right?

Yeah, I mean, before Facebook there was MySpace, there was Friendster, there were other platforms.

Yeah, yeah, there's other examples of that too. I mean, I mean, if you even look at what's happening with, uh, OpenAI right now, I do wonder if, you know, these, these other platforms are improving faster and creating monetization models. I think I saw a report that came out saying is that the, I think it was the Anthropic platform had a better monetization strategy than OpenAI and that they were making money already versus OpenAI was still burning capital to keep up with the the amount of servers and resources that they need to do all the token costs and things like that to keep them operational. Because I mean, I don't know about you, but I'm, I'm only paying like $20 a month for OpenAI's ChatGPT. But I noticed that Groq and, um, um, you know, the Anthropic platform and other ones out there are starting to charge for, for even minimal use now. So, so you can kind of see that, that adjustment coming.

Yeah, I pay for Claude. Um, I also pay for ChatGPT, even though I don't use it as much.

It's kind of my backup. Yeah, yeah, Claude's increasingly, you know, taking over, and it's really kind of pushing OpenAI, I think, right now.

Yeah, well, and then, you know, every platform has issues, but then you You look at the ethical backgrounds and backdrops, which ones were started by people who left a different company because they wanted to do something that was, that they thought would be more ethical, but then they of course still have issues. Perhaps they've used content that they didn't have licensing to, to train the models. So there's, there's still, and not including environmental issues, right? All of the other issues, um, with labeling data and who's labeling the data and how much we paying them and. It can get pretty complex if we want to get into the nitty-gritty.

Yeah, and I've been seeing a lot of job listings here recently where these platforms like XAI and other ones, the Anthropic platform, that are looking for multimedia editors, whether it be audio and video editors, that are being tasked with creating content in their platform for them based on what they see, what the demand is, and evaluating content. Humans are actually doing this, right? It's not being done by the AI. It's— there's people that are going through looking at questions and how the AI responds to those questions and massaging that. Yeah. And so it does make you wonder, it's like the old industrial revolution that we went through where there was a lot of people that lost jobs, but there was all these new jobs that were created because all these platforms want to get the results that these things produce in the real world to be not copyright infringed and not duplicative of any other platform. And have a human kind of lens on them.

Yeah. And really knowing AI is the biggest skill you can have right now, whether you're a content creator, podcaster, or anything else in business, because I believe right now one of the number one jobs that is being advertised in Los Angeles County are people who can kind of be that AI translator. So they know AI, right? They can kind of translate between different groups of humans, whether it's the people who are coding and the engineers and developers and the C-suite, and can kind of understand both sides. They can bring the business messaging to one level, and then they can bring the code speak and the developer speak into a language that the CEOs and the C-suite can understand. So there's a big need for that, for people who can understand both sides and who can interpret, and for people who can do exactly what you were saying, which is go on platforms and you and still be the human in the loop who's training the AI.

But that's, that's so against what we all think, how AI works, right? I mean, if you think about it, we've all been kind of like maybe propagandized, maybe a good way to describe this around how these models work, right? It's all automated. They all just kind of do their thing because they're out there scraping information and they're so smart that they make all these associations. But it appears to me that behind the scenes it's not really like that, right? So I mean, it is to a certain degree, but maybe not as much as we were told it was.

And it depends. It depends on which model you're doing. Are you using human in the loop? Are you just feeding a whole bunch of data into something and letting AI figure out the patterns?

Right.

And that's more machine learning than the general—

That's kind of what it does, right? Yeah. It basically is a numbers game for it. It's right. It assigns tokens to pieces of information and then it, it arranges them based on how it thinks they should be arranged based on the history.

Yeah, but then, you know, there are places like we don't have enough radiologists, so computer vision is really great at reading x-rays, much better than a human plus an AI or a human.

You can get down to the pixel level, right? It can see gradients that the human eye just can't see.

Yeah, programmatic advertising, about 90% of that is done with AI now.

So even in our own industries, I'm sure there's a lot of AI being, being used in YouTube too, to evaluate content and, and to categorize it and find audiences. And, and it just, it's constantly learning from our content. I think it's a lot deeper than any of us really realize. You know, if you look at the capability, video capability of the Google platform with the Gemini platform. Um, it's doing a lot more than any of us realize on the back end, and that may be why we're seeing big shifts in what type of content is working on YouTube now and what ones aren't, because it does know at a very deep level, um, what works with humans, and it's continuously learning.

And it's also trying to influence us. Well, that too, right? To go back to maybe some on the scale, 3-minute videos, and then, hey, let's do a little bit longer form. It's, it's— I think, I feel like I've been hearing this and seeing this a little bit, um, that the algorithms are trying to get us to like a little bit more long-form content again and not just be in this, you know.

Yeah, I noticed that too, because I, I mean, I think they may be ultimately moving away from short form, I think, a little bit. I mean, it's going to be there, but I, I just don't know that as it becomes mostly AI content, um, they may gravitate towards longer form content because, I mean, obviously they make more money if they can keep people watching for longer periods of time. There's more opportunities to deliver ads, and they're capturing attention from other platforms the longer they keep people over there. And, and I do think that that, uh, is a big part of what we're seeing right now, is, uh, it's all about attention now. So if you can hold on to it as a content creator, you're going to make a lot of money. Yeah. So as a, as a creator, that, that puts you in a situation where you have to start thinking about this stuff at a, at a very professional level. Um, and I think that's the challenge that's before a lot of people, and that's why a lot of creators now have to have teams. Mm-hmm. And you got to bring in experts that are experts in a specific aspect of creating content. Now, I don't know that one human has the capability of doing what needs to be done in the space today.

Yeah, I, I mean, agree with all of that. Having teams is important. Having an understanding of your brand value and understanding What brand extensions you can create now to monetize? And so that if now we're, you know, if in 2 years we're flooded and almost all the shows are AI and they have more of a share of voice than human podcasts, what else have we done to prepare ourselves for that world?

Yeah, that's a difficult question. Yeah.

That's a difficult question. We have products, we have services. Are we doing in-person, you know, And so I think we're in this test phase.

Yeah. And, uh, yeah, I mean, at that point, people may have— you know, I know some of these technologists talk about how we don't have to worry about making money in the future. I'm a little dubious of that. Uh, I do think that we are going to have to figure out a way to make some level of income.

Um, And we're also, you know, we're talking from a US basis here. There are a lot of countries that don't have all of the technology that might not be coming along as fast.

That's a good point, though. I do wonder if this technology is going to be available globally through wireless, um, at some point. If you think about the, the, um, the satellites that are going to canvas the globe with internet connectivity that are out there. The, the whole, whole Starlink platform is certainly one that's going to play a big role in that. Then it's just a matter of— and I know if you're thinking globally, there are mobile devices now that are very inexpensive people are increasingly able to, to take advantage of, and I guess it all at the end of the day gets back to cost.

Yeah, and profit. If somebody wants to, you know, put satellites all over places where they might not make as much money, but it's for the good of people and connection, or if they are focusing on profit margin.

Yeah. Well, I think that it's, yeah, it's hard to think that it wouldn't be about profit margin.

Yeah. And that's going to leave a lot of people out even with the you know, you can get 2G, 3G phones, listen to podcasts, and do a lot of things on them, you know. But, um, you know, then if you don't have access, there's— you're limited into what you can do if you don't have the internet.

And that— but that problem exists today. Yeah. But increasingly I've been hearing about platforms that are building out in Africa and Asia or whatever that are increasingly giving people access and creating models out there that can help people learn and elevate themselves increasingly. And yeah, we'll have to see that, but it will create a— continue, I guess might be the right way to say— a divide out there between people that have full access to these capable tools and ones that maybe have more limited access. Yeah, yeah, which will— yeah, unfortunately. But who knows, there may be a time when that's not the case, but, um, it's hard to see that right now.

Yeah, and things every day I'm having to learn new things, whether it's about, um, podcast metrics and, you know, but using this other AI tool, or, you know, am I using agents or am I just using a custom GPT to do certain things for my podcast or for my content. There's a million and one ways we can slice and dice things right now. And so I think we have to get back to, you know, use a few that work really well for you right now that you don't think are going to go away.

Yeah, well, yeah, because I mean, you, you can invest— I know I've spent a lot of time training OpenAI's ChatGPT, you know, 5.2 version, with a lot of background information about me and have created separate kind of like private and public instances inside of it. I can really have some level of perceived control of, of what information that it knows about me and puts out there in the public if somebody were to inquire about what I'm doing. And because I think The Google search engine is going to be replaced by this, and it's already happening. You know, you do a search in Google and the priority result is the Gemini AI results that summarizes everything. It's a little bit like what happened with the Alexa, right? There was one answer. There wasn't a list of options. There was one answer. And that's the era that we're moving into is one answer.

Yeah. And before we get there, that's, you mentioned making sure that when people, when your name comes up in search on LLMs, that they have all the information you want them to have. Right. But that's also why now we have to look at, do we need to be looking at Quora and all these other platforms and Reddit to see what people are saying and make sure we have those, you know, that we're doing on-page SEO so that when things go into the LMs, we still are connecting to the right websites or connecting to our brand? Or will we even have websites? Will it just be a, you know, a quick app where they can just immediately contact us or our brand and buy something right there?

Yeah, I think it's interesting. Will websites get integrated into the AI platforms? Is the— it's kind of heading that direction right now, I think. I mean, right, right now I think websites are important especially if, if you're using the, the, um, the, the AI optimization, um, protocols now, like the AGO. And right, is it the AEO, I think, is the other one, right? And that has certain formatted, um, presentations of content, right? There's— that is specific to what the AI is looking for. So you can use your website as a source platform for making sure that your AI understands you and understands your content. And will it link to you? Is— and will anybody click on that link? Is— that's, that's the other issue, right? Is that you're feeding it information and it will link to you in the results, but does anybody ever click on those links?

Yeah, well, I have heard— it's got the information, right? I've heard a lot of people say that they were found through LLM search and get business out of it.

So through their, their, their website. Yeah. Pathways. Well, that's, that's, that's a good thing to hear.

Cause I wasn't sure that that was going to continue more so than, um, than the AI overviews that Google will offer. Um, because that's actually taking traffic away from people going to websites because you find out the answer right there and you might, you might click through, but if you can read the information right there and you see the source, then you don't need to click through necessarily. Unless you are already in, you know, saying, oh, I'm not just doing this for informational purposes, I'm actually making a purchase decision and I have purchase intent.

So yeah, it's actually, yeah, it's actually forcing you to— I mean, if you like copy any kind of a, of a result, in— it's actually forcing you, when you copy it, it copies the, the source URL too, I noticed. Um, so it's increasingly doing that. I mean, you can tell it to remove them and it, and it will, but it by default it's giving you links to the, the source of that particular piece of information that it's getting, which leads me to believe that it's, you know, this, the human-created content is still where it's at right now. Um, I don't know if that's going to change really anytime soon.

Well, I can, I can tell you that in my podcast class right now, I think I have 17 students Some of them are working in pairs. Every single podcast is completely different, a very different idea than any other podcast in the class. And that's what the beautiful thing is about human creation.

Yeah, it's the, it's the level of, I guess, human creativity is always a little bit different than the AI creativity. And I think that is the, that's, that's where, where the dividing line is right now. Yeah. How do you think what's happening here is changing the, the culture of your students and how they perceive their, their future? Really, is it, is it optimistic or is it pessimistic?

Oof, it's a little bit of both. A lot of the students who have great internships or great jobs at some of the studios around LA, are just fine. A lot of my students in the digital social media program are probably not going to stay in the United States necessarily. They might go back to China or go back to Germany or London or wherever they're, um, but they want to get there. And I think that's one of the key factors is people might wonder why get an education when everything's online, but going through courses and having professors with lived experience is giving them a strong foundation. And I think that's the key is having it, even if somebody's starting a podcast, have a strong foundation know your brand, you know, build it, have your business plan built out. That can change at any moment. But, um, those things are skills that students can take to a job that I think are different because they're actually putting things in a practical application versus just thinking about things as esoteric information that they're learning and not really getting their hands dirty. So we really try to get our, our students to work on things that they can then do as intrapreneurs in other companies or as entrepreneurs.

Do you see them moving more towards passion projects, leadership projects, kind of, you know, civic-oriented type of mission-based focuses, which I think tend to be a little bit more human-centric?

Yes, and it's all over the place. Honestly, you know, some are about movies and friends and being a foreign student on a U.S. campus, and some are about sustainability and other things that, as you say, are kind of looking at that more journalistic lens and have a bigger purpose in mind where they're not going to just— they're going to continue They're gonna have really high value and maybe then they'll get a job producing that for a big channel that's going away from broadcast and looking more at podcasts as a medium, or maybe for a sustainability company that wants to start a show.

Yeah, I feel this, this tension with all this around specialization versus generalization. Yeah. And I think that is a, is a very interesting way of looking at what the future looks like. I do hear people talk about AI is going to require kind of strong, kind of generalist skills, multifaceted, multi-skilled, to be able to facilitate and manage the AIs. But yet at the same time, I feel this pressure in in the content creator industry for very deep, in-depth experiential experience in, in what they bring to a role for, let's say, a media company, which, which makes it difficult for young people, I think, to find a career. Yeah, I would agree with that. Yeah. So, you know, what do you think is probably the best way to approach this?

Is it the specialization, or you think it's the generalist I, um, I've always been a generalist. Yeah, I'm on the same boat. It has made it hard sometimes to find a job, um, when I've been looking to work for somebody else. When I, you know, um, not in the past several years, I've been really lucky to be hired at the VP level at several companies. But, um, even 6, 7 years ago when moved back to LA and was looking for what am I going to do next before I started a PR firm and started my podcast and went on that journey, I applied for a lot of jobs and I had the right experience, but I might— they wanted somebody who was, I think, going to be there for the long term and really be a specialist. So I think that for me, you know, it's hard to say because there are people who will do really well as specialists, but I also do. I'm a firm believer in knowing a little bit about a lot of things, and I think that is going to help people with this AI transition because we don't know what level of AI people are using in different companies. So are they gonna be able to help them with analytics, right? With the marketing campaigns, with the creation? And so knowing a little bit about all of these areas is going to help them be better equipped to be managers, to be leaders, and to bring their knowledge to companies that aren't quite at that level of sophistication yet.

Yeah, I have a question that popped up from Barb T USA. What if the content producer reveals that their research and writing has been assisted by AI?

That's a great question. What I've seen in consumer statistics is that If you reveal it, then it's a lot better than not revealing it. So, I mean, let me give you a little nuance. So if you have a TV, you know, you're advertising that you have a smart TV. Using the word smart TV is better than using AI-powered TV because people will still have some caution around that. When it comes to content, when it comes to what we're putting online about our brands, people still— and I know people think this word is overused, but I still love the word authenticity. People still want to have that trust and disclosure and authentic factor. So if you are sharing that you have used, had some help from AI, people are going to respect that a lot more than if it sounds like you've maybe used AI and they can tell, but you didn't disclose that. I encourage my students to use AI, but I give them, I have them practice prompt engineering and ask them, okay, use this tool and use this tool. Which one worked better for what you're trying to achieve? Or is there one on here that you've used that, that isn't what I assigned to you that does an even better job? And so they have to get, they have to really learn how to hone in and they have, but they have to also disclose that they used AI. Yeah. But, but there's no excuse for bad writing these days.

Um, because that's certainly true. Um, and then also, I mean, at one point, uh, I mean, at what point down the line here that the AI knows us so well, right? It's been trained on our especially if you're a media creator. Yeah, like I do. I've got, I got 22 years of content that's available online that, um, that the AI could learn all sorts of stuff about me and how I communicate and that kind of stuff. And, and what if it just knows me so well that it just, you know, it's going to be impossible to, to decipher whether it's AI or really me? I think we're going to get there. I don't think it's going to be that much longer.

Yeah, but I think it's coming from you. It's your expertise that you've programmed in.

Well, it's going to have to create something unique in my mind all the time. I can never say the same thing I said on 10 other podcasts going forward, right?

And, and I will also say, flagging this to, um, Barb's point, for my Delphi clone, she has source citations. So as she's answering, as my digital version of myself is answering a question, if she has a reference from something, she'll pull up this, you know, here are the citations from my original material that she pulled from. And so that to me is using disclosure of, yeah, you're talking to my AI version and you could actually call her on the phone as well and have a conversation. But she will say, you know, I use these 5 resources and here are the folders and the links to them.

So you could just title, title yourself, um, AI Annika. So it takes care of it right there, right? Yeah. So yeah, I've seen, um, other, other people do that too, where they, they just title it, um, a digital clone or something, you know, that person's name, and it's a digital clone. Yeah, kind of a, you know, to put it in the title of or the name of the personality of sorts versus trying to label it on the screen or, you know, that kind of thing. Though I think that there's going to be labeling standards that are coming. I was talking to the IAB here recently, that's the Interactive Advertising Bureau, and they're putting together a standards document for AI identification. And I believe that there's other media or regulatory bodies that are working on this as well that will, will require some level of, of identification legally. And so I hope to have a representative— I tried to get a representative of the IAB on the podcast here like in the last month, but I couldn't seem to get the schedules worked out with them. To be able to talk about this issue, because I think labeling is going to be an important topic that we need to, you know, face, right?

And, and a good standard is to look at the EU AI Act. They're always kind of— they're, they are much better at data privacy than, um, we are.

Definitely more, more restrictive. Definitely.

Yeah, I can agree with that, right? But give a good basis for here are things that we really should consider, and that includes with AI labeling and use of AI and everything.

Okay, so the— so that's the EU, um, has a, has a, a public standard right now that they've already put out, or is it still in development?

They've already put— I mean, they always fine-tune a little bit, but they have—

it's already out. Okay. Yeah, it's going to be okay. So is there any other, like, federal government agencies or FCC or anybody like that that you've heard that are trying to regulate by state.

So, um, in the United States, California has probably the most stringent, um, I think there are about 30 states right now that are looking at, um, adding in AI policies right now.

God, I hope that they don't create different standards for each, each state, you know. I mean, that could create a content creator nightmare is what it could create, or a platform Nightmare. Yeah, it's the same thing that happened with, um, oh, what was it in the early years? Oh, um, taxes. Paying taxes on transactions online. Oh yeah, you had all these different, you know, localities and, and things like that that wanted to have a different tax rate. And so it just required a lot of data integration to make that possible. And yeah, so, you know, that's a whole other layer of conversation of this. If Wyoming has a different standard than California has, then that means That means that the platforms have to understand those differences and be able to deliver a different, you know, solution.

Or you just go with the strong, you know, the most stringent, just like we do with data privacy for the GDPR and the CPRA, and every state has different levels of data privacy restrictions. Yep. You know, but you should assume that, I mean, I think any business should assume I probably have one customer from Europe, so I need to do what the GDPR says. And that's gonna be stronger than anything in the United States. I mean, even, you know, China has these restrictions, like lots of different countries have them.

Yeah. And it looks like Barbara asked or made another comment. He says, we humans train the AI. My ChatGPT is pretty good now. Due to my training.

So yeah, and that's what it should be. I mean, same, Rob, I know you have this as well. I have tons of different projects set up in Claude so that if I can go into a specific class and say, hey, look at this presentation, is there anything I'm missing that I haven't read in current news that I need to put into this deck, right? Or any, any different combination of things. Um, but it's— we train it to get to know us, and we use the right prompts and write the right information, and that makes it better. And, you know, hopefully we have that in a saved private area instead of having it as a public.

Yeah, I mean, I've set my ChatGPT to default to private. Yeah, good. And then, and then what I tell it, uh, when I give it information, I say go ahead and make it public. So, um, so I, I give it the option, and it has a regular occurring, uh, basically appointment every month at the first of the month to, to report to me what is public and what, um, what is new that's been published public. Um, and it gives me a report, it sends me an email, and I, I click on a link and it goes into my ChatGPT and it shows me the information that it's deemed to make public. So, you know, how strict it's adhering to that, it's a matter of trust, right?

And that's a whole other ethical AI conversation. I work with some people at Zipper AI and really trying— there is a truthfulness scale with every single LLM. And we, you don't want it to go too far on the truthfulness scale, but you can have them tell you where they are on the scale. Um, so it's a much more that we'd have to have another hour to go into that, but it's, um, it's pretty interesting work and it's really training models to be very ethical and truthful, but not, but you know, to a certain degree so that it's still human language that we understand and isn't isn't going too far down the rabbit hole, I guess.

Yeah, I think that this topic dovetails to what we casually mentioned earlier about the Anthropic platform losing its contract with the Department of Defense, right, over its capabilities. But it seemed to be really cozy with the OpenAI platform.

Well, and we also see different monetization models, right? You were talking about that, Claude. Is really built for coding. So that's more expensive than what you and I are probably paying for it. But if we want to have that much capacity, we pay more. Whereas OpenAI is going to the, hey, we're going to monetize through advertising.

So we're going to be— Wow.

Yay. Yeah.

Yeah.

So we'll be starting to see ads.

So I'm not quite sure how that's going to— Well, no, I know. Let me back up. I can totally see how that's going to work. Um, yeah, it's kind of like what Google's been doing for, for years in our Gmail and our calendars and, and our Google Docs and stuff like that. They, uh, they're constantly monitoring that, looking for trends of interest, and then mapping those to advertisers. Yeah, so that's kind of what's going to happen here. And then increasingly, those e-commerce companies are going to be accessible through these platforms too, and I do wonder, you know, I heard this talk a lot about, I mean, at some point, um, we will probably have just a handful of AI, uh, apps on our devices, just like maybe 2 or 3 or 4 or 5. It depends on how many we're subscribing to. And they will all have like their own app ecosystem built into them. Yeah. And those agents will work with those apps to accomplish things in our lives.

Um, yeah, yeah, I mean, the API model is alive and well, and there's a lot of platforms out there that are already aggregating. Just have to figure out which ones are going to stick around.

Yeah, so these, these days of having a sea of apps on your phone, um, may be limited. I'm okay with that. You're tired of scrolling this way and scrolling that way.

I'm like, wait, what are these apps that I downloaded that I can't remember what what they do.

So do I click to make, you know, and if you don't, uh, and then there's going to be some mechanism to be able to connect them, connect them up somehow. And then I think the other layer is permissions, uh, and, uh, trust of giving them access to like our emails or our financial information or all those kinds of things, credit cards, to do things on our behalf. So That's, that's, that's where we're going. Boy, we've really, um, meandered on this topic tonight, haven't we? So far beyond what I did on your show.

Oh no, no, we did a lot on, on my show too, but we've definitely— well, we also— this year's show is longer, so we've got, you know, it is not a few more rabbit holes.

Actually, Todd, Todd Cochran and I have done 2-hour episodes on this show before, so it's It's not that hard to do, actually. Yeah, surprise. I mean, a lot of people like, I only want to go 40 minutes, and, and I can appreciate that, but it's not very practical for me.

No, it's fun. This is— it's a fun conversation to kind of just throw it all out.

I hope it, uh, is— has been valuable to those that are watching and, and listening, because this is an audio and video podcast, so It's both in Apple. It's been in there for 14 years. So you can watch the video of this show in Apple Podcasts. And it also supports alternative enclosures as well. So on the video feed, you can get access to the audio too. So in like Fountain and some of the Podcasting 2.0 apps out there too. So, so, and And it's really supporting the whole ecosystem that exists right now, even Facebook, X, YouTube, all those. But I'm increasingly interested in trying to get involved in the Apple Podcasts HLS stuff too. I'm not really sure how to do that with this show because the— The platform that I'm hosting on right now for this show is Blueberry, and Blueberry wasn't a part of that first group of platforms to support HLS streaming. So I may have to create a separate account at like an Art19 or something like that to be able to pull that off. But, but it's— I'm hopeful that the other podcast hosting platforms will be able to embrace the HLS stuff here soon. So yeah, and then we'll really have an interesting race going on here. Wow. Well, anyway, I'll let you go. Um, it's great to have you on the show, Annika. It was, you know, it's always an interesting conversation, and I love it that you're a professor too. I kind of would like to maybe get involved in being a professor at a university. I'm surrounded by them here out in Connecticut, so there's, there's ones that, yeah, that I'm already talking to, so it's just a matter of, you know, these are, these are ones that haven't launched their, their podcasts, media formats yet.

So, well, it's, um, it keeps you on your toes. I definitely have to be as up to date as possible on technology and AI.

It sounds like that you're doing a good job of that, so, you know, I, I learned a lot from you tonight.

So thank you. Yeah, as I did from you. But, and my— I learned things from my students who are in the field as well, those of them who have already had that experience.

So yeah, yeah, yeah. All right, well, all right, well, thank you everybody for, for watching us tonight on the, on the New Media Show. And you can get this, like I said, as a podcast at newmediashow.com, and, um, that's the best place. It's on all the podcast apps, so it's easy to find us. And, um, and I hope you have a terrific evening. And same with you, Annika. It's great. You too, Rob. Okay, thanks. Good night, everybody.