The Media Copilot

Perplexity is here to answer your questions. Answering questions requires good information, though, and providing answers is easier if the people with the information actually want to give it to you. That’s a super-simple way of describing why Perplexity — the “answer engine” AI startup that pairs generative summaries with search — this week launched the Perplexity Publishers’ Program. By sharing advertising revenue with partners, Perplexity hopes to create better incentives for them to allow access to their content. To unpack the new program and what it means to the media business, I spoke to Dmitry Shevelenko, the Chief Business Officer of Perplexity. If you're a cynic, you might think the move is purely defensive — that Perplexity is doing this so it doesn’t get sued. My takeaway, though, after talking to Shevelenko, is that Perplexity recognizes it needs good facts and good journalism to help fuel its so-called "answer engine, and that it wants to find a way to keep those coming. But yes, he and I do talk about those notorious accusations about plagiarism from Forbes and Wired, what's the right way to attribute original reporting in an AI summary, and yes — that OTHER AI search engine that was just announced by a competitor you might know called OpenAI. The Media Copilot is a podcast and newsletter that explores how generative AI is changing media, journalism, and the news. ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Subscribe to The Media Copilot newsletter.⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ ⁠⁠⁠⁠⁠⁠⁠Explore our courses⁠⁠⁠⁠⁠⁠⁠ on how to use AI tools, tailored for media, marketers, PR professionals, and other content creators. ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Follow us on X.⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Subscribe to the podcast on: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Apple⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Spotify⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ ⁠⁠⁠⁠⁠⁠⁠Other podcast apps⁠⁠⁠⁠⁠⁠⁠ Music: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Favorite⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ by Alexander Nakarada, licensed under Creative Commons BY Attribution 4.0 License © AnyWho Media 2024

AI companies hoovered up the entire internet before anyone questioned: Is that OK? This is the question that has led to a multitude of lawsuits (including, famously, The New York Times suing OpenAI), many deals between AI companies and various publishers, and an emerging consensus that content creators should be compensated for their data. That's all well and good in theory, but what about the "how?" That's what my guests this week, Olivia Joslin and Toshit Panigrahi, are here to talk about. Joslin and Panigrahi are the co-founders of TollBit, which is one of a handful of companies tackling a very hard problem in media today, and that's how to get people to pay for content. Well, not people — machines. The web scraping that AI companies do used to be fairly benign. It helps power things like Google search and has traditionally been used as a research tool. But now the big AI guys are just taking that information and summarizing it, without doing that much of the linking anymore. And that's kind of a problem. Publishers are very motivated now to find ways to be compensated when someone scrapes their content, and that's exactly what TollBit is trying to create: a marketplace where publishers and those who want their data can make a simple exchange: money for information. Of course things aren't always that simple, and I got into the complexities, the politics, and the realities of trying to get someone to pay for something that they've up until now been getting for free. It was a really illuminating conversation, and I hope you listen all the way to the end where we zero on what a healthy media future looks like. As ever, if you enjoyed this conversation, it would be great if you could follow the show on Substack, Apple Podcasts, Spotify, or any other podcast app, really. Also, I’d appreciate it if you’d leave a rating or review — it really does help the show. And if you’re on YouTube, please like the video and subscribe to the channel. Much obliged. The Media Copilot is a podcast and newsletter that explores how generative AI is changing media, journalism, and the news. ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Subscribe to The Media Copilot newsletter.⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ ⁠⁠⁠⁠⁠⁠⁠Explore our courses⁠⁠⁠⁠⁠⁠⁠ on how to use AI tools, tailored for media, marketers, PR professionals, and other content creators. ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Follow us on X.⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Subscribe to the podcast on: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Apple⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Spotify⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ ⁠⁠⁠⁠⁠⁠⁠Other podcast apps⁠⁠⁠⁠⁠⁠⁠ Music: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Favorite⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ by Alexander Nakarada, licensed under Creative Commons BY Attribution 4.0 License © AnyWho Media 2024

When you think about AI disrupting Hollywood, you typically think about CGI characters and creation tools like Sora and Pika that turn text prompts into impressively realistic visuals in seconds. But while technologies that create the things that meet our eyes understandably grab most of the headlines, there’s a quieter disruption happening at the very beginning of the moviemaking process. Dilip Rajan is the co-founder of Jumpcut, whose tagline is, “Automate grunt work so you can get back to storytelling.” It does that with a product called ScriptSense, which inserts AI into the tedious slog of script coverage — the act of reading and evaluating screenplays. If a script isn’t from a known writer, a studio doesn’t have a lot to go on, so they often outsource it to a network of contractors, which obviously takes time and depends a lot on the tastes of those contractors. Powered by large language models (LLMs), ScriptSense can read way more scripts than any human could, analyze them to a common standard, and then summarize what they're about and if they're any good. Subscribe now Of course, there's a little more to it than that, but you can see why any movie studio would be interested in a machine that can read screenplays at scale, especially when there are tens of thousands of scripts floating around Hollywood at any given time — a number that will surely increase now that ChatGPT is joining the writers room. I chatted with Dilip about where Jumpcut came from, how writers and agents should think about the disruption AI represents, and why Jumpcut's script processor might be just the tool Hollywood needs to get out of its current creativity rut. I thought it was fascinating conversation, especially when we went deep on how ScriptSense works — it's not just "Chat with PDF" on steroids. If you want to check out Lucihub, please email us a ⁠team@mediacopilot.ai⁠ with Lucihub in the subject and we'd be happy to send you a discount code. The Media Copilot is a podcast and newsletter that explores how generative AI is changing media, journalism, and the news. ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Subscribe to The Media Copilot newsletter.⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ ⁠⁠⁠⁠⁠⁠Explore our courses⁠⁠⁠⁠⁠⁠ on how to use AI tools, tailored for media, marketers, PR professionals, and other content creators. ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Follow us on X.⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Subscribe to the podcast on: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Apple⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Spotify⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ ⁠⁠⁠⁠⁠⁠Other podcast apps⁠⁠⁠⁠⁠⁠ Music: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Favorite⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ by Alexander Nakarada, licensed under Creative Commons BY Attribution 4.0 License © AnyWho Media 2024

When it comes to AI changing video, generative tools like Sora and Dream Machine have stolen a lot of the spotlight. with their ability to "imagine" video clips from text prompts. As cool as they are, most of them aren't available yet, and there's a big question of whether they actually save you any time. But while these magic clip generators create a lot of excitement, there's a quiet disruption taking place downstream. Because in the real world, when you undertake a video project (whether it's editorial or marketing), there’s a universal rule: You want it done well, you want it done fast, and you want it done cheap. And you can only have two of those things. Could AI put us on a path to getting all three? That's what I wanted to find out from Amer Tadayon. Amer is the CEO of Lucihub, a service that can create videos that would have taken days or weeks, and turn them around in hours — sometimes minutes. You still need to give it raw footage (it's not a purely generative tool like Sora), but it uses AI so that Lucihub’s human editors can cut videos incredibly quickly, and for a lot less than you'd pay an agency. For creative teams struggling for manpower and resources, it feels like a game-changer. I talked with Amer about how Lucihub is disrupting the conventional wisdom around video, as well as what that means for the big picture. Could AI make video great again? If you want to check out Lucihub, please email us a team@mediacopilot.ai with Lucihub in the subject and we'd be happy to send you a discount code. The Media Copilot is a podcast and newsletter that explores how generative AI is changing media, journalism, and the news. ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Subscribe to The Media Copilot newsletter.⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ ⁠⁠⁠⁠⁠Explore our courses⁠⁠⁠⁠⁠ on how to use AI tools, tailored for media, marketers, PR professionals, and other content creators. ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Follow us on X.⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Subscribe to the podcast on: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Apple⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Spotify⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ ⁠⁠⁠⁠⁠Other podcast apps⁠⁠⁠⁠⁠ Music: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Favorite⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ by Alexander Nakarada, licensed under Creative Commons BY Attribution 4.0 License © AnyWho Media 2024

Quick question: How do you know it’s really me in this podcast video? After all, with AI services like Synthesia and ElevenLabs ready to clone my likeness and voice in mere minutes, it’s more than a possibility these days. Besides obvious artifacts you might see in an AI-generated video, there’s the concept of content provenance, which in practice means a set of technical standards that imagery must adhere to so it can include metadata — nonvisible information that reveals whether an image was created by AI or not — and what modifications there were before it hit your eyeballs. The thing is, for content provenance to work, you need everyone in the chain — from camera to website — to pay attention to it. So it turns out infrastructure plays a big role, and that’s what I wanted to to talk to Livepeer CEO Doug Petkanics about on The Media Copilot podcast. Livepeer is a decentralized video platfom, which has an ace up its sleeve with respect to content provenance: the blockchain. By using the same tech that powers cryptocurrencies, Livepeer aims to create a cost-effective, scalable, and reliable way to both process video and label it properly. Its new AI subnet means the network now supports AI-generated content, including tools like OpenAI's video generator, Sora. Doug and I explored the critical issue of content authenticity in the age of AI. As deepfakes and AI-generated media become more prevalent, this is obviously a growing challenge. Doug explained how Livepeer's blockchain tech can help maintain a healthy media ecosystem by providing verifiable “attestations” of content creation and modification, helping ensure that audiences can trust the what they’re looking at. The Media Copilot is a podcast and newsletter that explores how generative AI is changing media, journalism, and the news. ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Subscribe to The Media Copilot newsletter.⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ ⁠⁠⁠⁠Explore our courses⁠⁠⁠⁠ on how to use AI tools, tailored for media, marketers, PR professionals, and other content creators. ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Follow us on X.⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Subscribe to the podcast on: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Apple⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Spotify⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ ⁠⁠⁠⁠Other podcast apps⁠⁠⁠⁠ Music: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Favorite⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ by Alexander Nakarada, licensed under Creative Commons BY Attribution 4.0 License © AnyWho Media 2024

Journalists are naturally skeptical people. They look critically at new things, especially when the incentives around them are complex, and that's certainly the case with AI. Given the early missteps of some sites publishing AI content and the existential threat the technology poses to distribution, it's only natural that a stigma around using AI has emerged among many reporters. That stigma is something Josh Brandau is wearing down, one newsroom at a time. Josh is the co-founder of Nota, a content platform for augmenting newsrooms with AI tools. I spoke to Josh for The Media Copilot podcast about the company and how it's grown since its launch in the summer of 2022 — well before ChatGPT and generative AI exploded into the mainstream. Josh and I discussed how Nota is helping newsrooms, especially small to midsize ones, giving them easy ways to leverage AI to create content more efficiently across multiple formats. But we also talked about how transformative AI is going to be, both for how journalists do their work and the industry as a whole. With everything happening with Google’s AI Search and ChatGPT’s new ability to really talk to you, that discussion is definitely more urgent than ever. If you enjoy this conversation, I’d encourage you to follow the show on Substack, Apple Podcasts, Spotify, or any other podcast app, really. Also, we’d appreciate it if you’d leave a rating or review — it really does help the show. And if you’re on YouTube, please like the video and subscribe to the channel. The Media Copilot is a podcast and newsletter that explores how generative AI is changing media, journalism, and the news. ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Subscribe to The Media Copilot newsletter.⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ ⁠⁠⁠Explore our courses⁠⁠⁠ on how to use AI tools, tailored for media, marketers, PR professionals, and other content creators. ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Follow us on X.⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Subscribe to the podcast on: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Apple⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Spotify⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ ⁠⁠⁠Other podcast apps⁠⁠⁠ Music: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Favorite⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ by Alexander Nakarada, licensed under Creative Commons BY Attribution 4.0 License © AnyWho Media 2024

In the latest episode of The Media Copilot podcast, I had the pleasure of talking with Jon Gillham, founder of Originality.ai, about the nuanced world of AI-generated content and its detection. Jon's company started from a simple need in his content marketing business: ensuring that content was authentically created by humans, not AI. As AI sophistication has grown, so has the necessity for robust detection tools. The field of AI detection is more complicated than you might think. Jon points out that while not all AI content is spam, almost all spam is now AI-generated. That leads us to an unpacking of Google’s dilemma — that targeting AI-generated content in search results might result a better experience for customers but its position as a major LLM developer inherently conflicts with that goal. Nonetheless, AI detection tools are essential for publishers trying to navigate the new digital landscape without compromising their search rankings or credibility. We also talk about the importance of transparency and authorship as AI becomes more ingrained in digital content creation. Projecting forward, you can begin to see a “hybrid” future where AI aids content creation under stringent guidelines to ensure quality and authenticity, and that’s OK! I’m really happy with how the conversation goes deep on the complexities and realities of having AI “out in the wild” in our information ecosystem, and how the interplay between AI technologies and content creators has evolved — and will continue to evolve. If you enjoy this conversation, I’d encourage you to follow the show on Substack, Apple Podcasts, Spotify, or any other podcast app, really. Also, we’d appreciate it if you’d leave a rating or review — it really does help the show. And if you’re on YouTube, please like the video and subscribe to the channel. The Media Copilot is a podcast and newsletter that explores how generative AI is changing media, journalism, and the news. ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Subscribe to the newsletter.⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ ⁠⁠Explore our courses⁠⁠ on how to use AI tools, tailored for media, marketers, PR professionals, and other content creators. ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Follow us on X.⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Subscribe to the podcast on: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Apple⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Spotify⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ ⁠⁠Other podcast apps⁠⁠ Music: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Favorite⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ by Alexander Nakarada, licensed under Creative Commons BY Attribution 4.0 License © AnyWho Media 2024

Media expert Ricky Sutton discusses AI's impact on the media landscape, including the transformation of news creation, distribution, and monetization. He explores the necessity for media brands to adapt to AI technologies to stay relevant in the evolving industry and shares insights on navigating the AI revolution in media.

This week on The Media Copilot podcast I’m thrilled to talk to Alex Fink. Alex is the founder and CEO of Otherweb, a news aggregator that uses AI to give readers a healthier news diet than your average social media feed. Instead of optimizing for outrage or clickbait, Otherweb favors “kale over cake” — an analogy we come back to a few times in the conversation. Alex has an interesting career. After working for a long time in computer vision, he decided the world had enough cameras and decided to focus instead on the decisions technology could help with rather than the tech itself. Otherweb isn’t his first rodeo — he’s been a founder twice before and a very astute observer of the media business. He’s full of great observations about the arguably corrupted incentives of ad-based media, which helps to guide Otherweb and the way it ranks and serves up content. You can check out Otherweb here: https://otherweb.com/ The Media Copilot is a podcast and newsletter that explores how generative AI is changing media, journalism, and the news. ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Subscribe to the newsletter.⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ ⁠Explore our courses⁠ on how to use AI tools, tailored for media, marketers, PR professionals, and other content creators. ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Follow us on X.⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Subscribe to the podcast on: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Apple⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Spotify⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ ⁠Other podcast apps⁠ Music: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Favorite⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ by Alexander Nakarada, licensed under Creative Commons BY Attribution 4.0 License © AnyWho Media 2024

Yes, we know generative AI is bad for writing articles whole-cloth. But what IS it good for when you want to apply AI in a newsroom? In this week's episode of The Media Copilot podcast, host Pete Pachal explores that question Joe Amditis, Associate Director of Operations at the Center for Cooperative Media. As part of his role, Joe researches how journalists can apply generative AI, both at the individual and organization levels, and has written guides on publicly available tools, including the how to create custom GPTs for reporting use cases. Joe's advice to those new to using AI for journalism? Experiment. Journalists need to use AI to understand its capabilities and limitations, and should focus on low-stakes tasks initially. Once they're comfortable, some of the most useful applications he's found are: Documentation and transcription Brainstorming ideas on why people should care about a story Filtering through data and documents to surface potential leads Generating stock images and graphics to accompany articles The key takeaway: Use AI tools pragmatically to gain efficiency in workflows, but do not lose sight of the human element and relationships at the core of journalism. As long as journalists don't lose sight of creating quality, valuable content for their communities, they'll be able to find ways AI can help move faster toward that goal. The Media Copilot is a podcast and newsletter that explores how generative AI is changing media, journalism, and the news. ⁠⁠⁠⁠⁠⁠⁠⁠Subscribe to the newsletter.⁠⁠⁠⁠⁠⁠⁠⁠ ⁠⁠⁠⁠⁠⁠⁠⁠Follow on X.⁠⁠⁠⁠⁠⁠⁠⁠ Subscribe to the podcast on: ⁠⁠⁠⁠⁠⁠⁠⁠Apple⁠⁠⁠⁠⁠⁠⁠⁠ ⁠⁠⁠⁠⁠⁠⁠⁠Spotify⁠⁠⁠⁠⁠⁠⁠⁠ Music: ⁠⁠⁠⁠⁠⁠⁠⁠Favorite⁠⁠⁠⁠⁠⁠⁠⁠ by Alexander Nakarada, licensed under Creative Commons BY Attribution 4.0 License © AnyWho Media 2024