Understanding emergent intelligence in work: Agentic, robotic and creative
9 Jul 2025 15:00h - 15:20h
Understanding emergent intelligence in work: Agentic, robotic and creative
Session at a glance
Summary
This discussion at the AI for Good conference explored the intersection of artificial intelligence and creativity, featuring perspectives from the arts, music industry, and technology sectors. Harry Yeff, who curates the AIVI initiative, moderated a panel with James McAulay from Eleven Labs, Chris Horton from Universal Music Group, and independent artist-director Michaela Ternasky-Holland. The conversation centered on how AI tools are transforming creative processes and the implications for artists and the broader creative industry.
The panelists discussed the concept of “bespoke media” and whether audiences desire infinite customization in creative content. Ternasky-Holland emphasized that context matters, noting that personalization can create powerful moments of connection for social impact work, while broader audiences may prefer shared cultural experiences. Horton from Universal Music Group highlighted the tension between customized experiences and the communal aspect of music, where shared songs create collective cultural moments.
The discussion revealed how AI is being used as an augmentation tool rather than a replacement for human creativity. Ternasky-Holland described her AI-assisted filmmaking process, where traditional animators collaborate with AI tools to tell more stories efficiently, while maintaining the collaborative essence of creative work. McAulay shared examples of innovative applications, including interactive AI agents based on film characters and installations that give voice to environmental data.
A significant portion addressed the ethical considerations around AI training data and artist compensation. Horton argued against the false dichotomy between content protection and AI success, emphasizing that artists deserve compensation when their work trains AI systems. The panelists concluded with optimistic predictions about AI serving as intelligent creative assistants while preserving the essential human elements of storytelling and curation that audiences value.
Keypoints
**Major Discussion Points:**
– **Bespoke vs. Universal Creative Experiences**: The panel explored whether audiences want infinite customization and personalized media, or if there’s value in shared cultural experiences where everyone engages with the same content (like singing the same song lyrics together at concerts).
– **AI as Creative Augmentation vs. Replacement**: The discussion emphasized viewing AI as a collaborative tool that enhances human creativity rather than replacing it, with examples of AI assisting with time-intensive tasks like animation in-betweening and music composition while preserving human creative direction.
– **Content Rights and Ethical AI Training**: A significant focus on the need for proper licensing and compensation when using creative works to train AI systems, challenging the narrative that content protection and AI development are mutually exclusive.
– **New Forms of Creative Collaboration**: The conversation highlighted emerging collaborative structures between humans and AI, including interactive experiences like AI agents based on film characters and giving voice to data (extinct animals, melting icebergs).
– **Future Visions for AI-Assisted Creativity**: Panelists shared predictions about AI becoming like “smart interns” that handle administrative and grunt work, allowing creators more time for actual creative tasks, while emphasizing that human curation and storytelling will become even more valuable.
**Overall Purpose:**
The discussion aimed to redefine creativity in the AI era, moving beyond traditional artistic boundaries to explore how AI tools can enhance problem-solving and creative collaboration across different media (voice, music, visual). The session sought to address both opportunities and challenges in AI-assisted creativity while fostering dialogue between artists, technologists, and industry representatives.
**Overall Tone:**
The tone was optimistic yet realistic, beginning with some acknowledgment of past “ambiguity and anxiety” around AI in creative spaces but evolving into a constructive exploration of possibilities. The conversation maintained a collaborative spirit throughout, with panelists building on each other’s ideas while honestly addressing challenges like casting difficulties and licensing concerns. The tone remained forward-looking and solution-oriented, emphasizing the importance of continued dialogue and ethical collaboration.
Speakers
– **Harry Yeff**: Artist and curator of AIVI initiative (with ITU and AI for Good); contributed to AI for Good since 2019; works with voice and creative technology
– **James McAulay**: Leads creator growth at Eleven Labs (world’s leading AI audio company); musician who writes piano music in evenings
– **Chris Horton**: SVP of Strategic Technology for Universal Music Group (world’s largest music company); focuses on technology policy and strategy, particularly AI
– **Michaela Ternasky Holland**: Independent artist and director; works in dance and performance; specializes in emerging technology for both social impact and narrative spaces
Additional speakers:
None identified beyond the provided speakers names list.
Full session report
# AI and Creativity: Redefining Artistic Collaboration in the Digital Age
## Executive Summary
This panel discussion at the AI for Good conference explored the evolving relationship between artificial intelligence and human creativity. Moderated by Harry Yeff, artist and curator of the AIVI initiative, the conversation featured James McAulay from Eleven Labs (creator growth), Chris Horton from Universal Music Group, and independent artist-director Michaela Ternasky-Holland. The discussion focused on AI as a collaborative tool, the balance between personalised and shared cultural experiences, content rights, and practical challenges facing creators working with AI technology.
## Key Themes and Perspectives
### AI as Creative Augmentation Tool
All panellists agreed that AI should serve as a collaborative tool rather than replace human creativity. Michaela Ternasky-Holland framed this perspective by comparing AI to established creative software: “It’d be like saying I’m a Premiere Pro powered studio or I’d be saying I’m a Photoshop powered studio. It’s like that is just a tool in your toolkit.”
She emphasised the importance of maintaining creative control: “There’s a power hierarchy that we as creatives have with the tool we’re using.” In her experience, AI assists with time-intensive processes like animation and score composition, but the creative process still involves struggling with AI systems, which reinforces that they are tools rather than replacements.
Chris Horton envisioned AI functioning like “a collaborative smart intern that handles grunt work, giving artists more time for creativity.” James McAulay highlighted how AI enables new interactive experiences, such as conversational agents based on film characters.
### The Personalisation Versus Shared Culture Debate
A significant portion of the discussion examined whether audiences want personalised content or value shared cultural experiences. Michaela advocated for strategic personalisation, particularly in social impact work: “What I do think is really powerful about customisation and personalisation… is that it can really kind of give somebody a very clear moment where they feel seen or they feel heard.”
However, Chris raised concerns about excessive personalisation in music: “There’s tension between customised music experiences and the shared cultural moments of hearing the same song together.” He questioned what happens to collective cultural experiences “if what I hear when I’m listening privately is slightly different than what you hear.”
The panellists agreed that context determines whether personalised or universal content is more appropriate, depending on the intended outcomes and audience needs.
### Content Rights and AI Training
Chris Horton provided a strong industry perspective on content rights, arguing that “AI learning differs fundamentally from human learning.” He challenged the common narrative that AI learns like humans, explaining that while a child might need 500,000 hours to learn the alphabet, AI systems require vast amounts of copyrighted content for training.
“It’s a false dichotomy to claim you can’t have both content protection and AI success, as licensing is possible,” Horton stated, emphasising that responsible companies are working on proper licensing for AI training data. He drew parallels to the peer-to-peer era, suggesting that similar licensing solutions could work for AI.
### Practical Challenges and Industry Adoption
Michaela shared specific examples of industry resistance, including casting difficulties where voiceover actors wouldn’t work on projects involving AI assistance. She described the frustration of explaining that AI was merely assisting with production work, not replacing human performance.
James provided a concrete example of innovative AI applications, describing a film producer who created AI agents that allowed audiences to have conversations with characters from the film, demonstrating new forms of audience engagement.
### Future Visions and Applications
The panellists discussed various applications of AI in creative work. Michaela mentioned her work on projects involving gun violence, nuclear weapons, and U.S. politics, where AI assistance helped with production challenges. She expressed hope for AI agents that could help with administrative production work to support independent creators.
James predicted that human curation and the stories behind art would become more important as AI-generated content increases. Chris suggested that future collaboration would assume people work with AI assistants, creating new considerations for hiring and workflow.
Harry Yeff shared his own experience with AI projects, including work with the Voicing Nature’s Cube installation, demonstrating how AI can give voice to environmental data and create emotional connections with abstract concepts.
## Areas of Consensus
The discussion revealed strong agreement on several key points:
– AI should augment rather than replace human creativity
– Context determines the appropriate use of personalisation versus universal content
– New forms of interactive audience engagement are emerging through AI
– Proper licensing and compensation structures are necessary for ethical AI development
– Human curation and storytelling remain essential even as AI capabilities expand
## Ongoing Challenges
Several unresolved challenges emerged from the discussion:
– Industry acceptance of AI-assisted work, including casting and collaboration difficulties
– Balancing personalised experiences with shared cultural moments
– Developing ethical frameworks for AI training and content use
– Creating better support systems for independent creators
– Understanding what constitutes effective AI collaboration tools
## Conclusion
The conversation demonstrated a mature understanding of AI’s role in creativity, moving beyond replacement fears toward collaborative possibilities. As Harry Yeff noted in his closing remarks, the path forward requires continued dialogue and “difficult conversations,” but offers potential for meaningful creative collaboration.
The panellists emphasised that successful AI integration in creative work depends on maintaining human agency, ensuring fair compensation for content creators, and applying technology thoughtfully based on context and intended outcomes. The discussion suggested that the future of AI in creativity lies in augmentation that amplifies human creative potential while preserving authentic connections and shared cultural experiences.
Session transcript
Harry Yeff: Hello. So thanks to all attending. This session, just to set the scene a little bit, addresses a central concept to AI for Good. I first contributed to AI for Good in 2019, and there was a series of artists that did a performance in one of the main UN spaces, and it was the first artistic expression of its kind to happen in this space. And to speak honestly, I think there was an ambiguity and anxiety, this idea of installations and concepts and the sort of creative industries within the potential of AI. And it’s hard to define, but there was, and you have to take my word for this, there was a magic in the air that evening, and something happened there, and the involvement of the arts, the involvement of artists, but also all of the difficult questions that are happening in the creative world has become more and more a part of AI for Good. So now I curate an initiative called AIVI, which is with ITU and AI for Good, and we’re really interested in these stories, in these strategic questions that come when we’re thinking about how we’re collaborating. And I really believe that the intimate relationship with how creatives are working with this technology is a signaling. I think artists work so intimately with this technology that there are many predictions, insights, but now there are many amazing tools being built, ways of working, and of course very, very difficult questions being asked. So just starting with James, I would love you all to just quickly introduce yourselves in this direction.
James McAulay: Thanks, Harry Yeff. My name’s James. I lead creator growth at Eleven Labs, and Eleven Labs is the world’s leading AI audio company. So we started with an amazing text-to-speech tool that allows agents to speak and express themselves in the same way as we do, as humans. And increasingly we are now offering these conversational AI agents, and there are some very exciting creative use cases of AI agents that I can tell you about later on. In my evenings, I am a musician and I write piano music and, yeah, do that.
Chris Horton: Hi everybody. My name is Chris Horton. I’m SVP of strategic technology for Universal Music Group, and Universal Music is the world’s largest music company. If you have a favorite artist, there’s a one-in-three chance that they’re distributed by Universal. And what I do for UMG is technology policy and strategy, and my focus in the last few years has largely been AI.
Michaela Ternasky Holland: Hi all. My name is Michaela Ternasky-Holland, and I’m an independent artist and director. I have worked a lot in dance and performance before I started diving into the world of emerging technology, and I’ve had the pleasure to be able to work in both the social impact spaces and the narrative spaces, and I’m so excited to be here today.
Harry Yeff: So all of us in this room, we have quite a unique opportunity because we have voice, we have music, we have vision, and I think seeing the new partnerships and the way that these different spaces are collaborating is an essential set of questions. So my aim for this session is to help us maybe redefine our thinking of creativity, just the literal description of it, and it doesn’t just belong to the arts or lateral thinking. There’s a desperate need for problem solving, and this is creative thinking. So one question I have immediately, Chris and Michaela, there’s a real debate around bespoke media, and I think like myself as an artist, you’re starting to see this fractaling out of tools where you can more and more make something bespoke. You can adapt to every nuance that someone brings to an experience. Do people want infinite creativity? Do they want the abundance of being able to have bespoke media? And if not, are there any reasons why?
Michaela Ternasky Holland: I think it all depends on context. Some of the work that I do is really trying to create thought-provoking moments for people. It’s trying to take people out of their traditional comfort zones and putting them in a space of wanting to learn about something or putting them in a space of not realizing that they had been so closed off to an idea, whether that’s around gun violence or nuclear weapons threat or even like right now I’m working on a project about U.S. politics. And so what I do think is really powerful about customization and personalization, which is something that generative does give us the ability to do, is that it can really kind of give somebody a very clear moment where they feel seen or they feel heard. And from that moment, we can then expand what they thought they knew. And I think that is a very powerful mechanism for change. The other hand, though, you know, if I’m trying to create something that is more for a larger mainstream audience, I want a large amount of people to see it, I’m not thinking just about the hyper-personalization. At that point, I’m thinking about can I tell a specific story that has a big universal kind of way for people to fall in love with it, whether that’s a traditional animated film, using AI-assisted technology, whether that is a larger exhibition at a museum where I know many people from many different backgrounds will come into the museum and want to enjoy new emerging technology. So for me, I think it’s really about context and it’s about what is the shared experience trying to do for the audience.
Chris Horton: And at UMG, we’re I think trying to do two main things. One is enhance the artist’s creativity and then two, deepen their connection with the fan base. And one of the ways that they may deepen the connection is by allowing the fans to interact with the music, possibly customize it. And we’re in the early stages of figuring out who is comfortable with what. And so I think we’ll know more about that a little bit in the future. But one of the interesting things that has been pointed out to me by a friend was we also have this cultural context of music where we all hear the same song on the radio and we go to the shows and we’re all singing the lyrics together. And if what I hear when I’m listening privately is slightly different than what you hear, what is the impact of that on that shared cultural moment? And so it may be that these things coexist, that the song is still released as imagined by the artist, but there may be context where you can customize. And I think we’ll learn how that resonates with the fan base in the near future.
Harry Yeff: Well, I like the idea that sometimes limitation is what brings quality to the human experience. But on the concept of abundance, James, I mean, there’s so many different artists for all kinds of different reasons and creatives and other use cases using Eleven Labs. I am curious what stands out to you. Has there been any examples, use cases that you think might inspire other artists out there to guide them or just something that you found particularly interesting?
James McAulay: Yeah, I’m in a very lucky position where I see hundreds of these very cool use cases every single week. I mean, one that stood out to me recently was a very well-known film producer came to us talking about a film that they’re working on and exploring how the audience and how the fans could talk to those characters either before the film comes out or maybe after the film has come out, creating these like AI agents based on characters in the film that the fans can then continue the conversation with. This is similar to your point about how do you connect with the fan beyond that moment of performance. So yeah, I think that creating interactive experiences around art is very interesting. Another example that you might have seen outside is the Voicing Nature’s Cube, where if you put the headphones on you can have a conversation. Correct me if I’m wrong, Harry Yeff, because this is Harry Yeff’s pride and joy. You can have a conversation with an animal that’s become extinct. You can have a conversation with an iceberg that is melting. And so we’re giving a personality, we’re anthropomorphizing these inanimate objects and letting people connect with them on a vocal level. Is there anything you’d want to add to that?
Harry Yeff: I appreciate the plug. But no, this concept of giving data a voice, and I think it’s really interesting, the experiential nature of agentic. Just how the ability to create an environment guides someone through a specific story, but the way that you can choose causes. And something that I personally think that stands out with the opportunities in the tech is the ability to summarize sentiment and articulatory intelligence, to say things and not just be a research partner, but also to sculpt the poetry of a journey and guide someone through that, I think is a really interesting use case. But there’s whole new forms of collaboration. I mean, Michaela, what I think is interesting about your work is outside of the power you have as a creator, it seems that you bring teams together and you’re very verbose and upfront with, I may be wrong, there’s a touch of pride in the fact that you have brought AI process to support these teams and the way that you create work. So do you see this more of an augmented process? would you differentiate that from the narrative of just pure automation, the replacement of creatives where you seem to be using it to enhance a directorial project?
Michaela Ternasky Holland: Yeah, I think, you know, I had a conversation earlier today with a fellow filmmaker who’s been using AI technology and, you know, a lot of this narrative around AI filmmaking or AI powered studios, I think is the wrong way of talking about the technology. It’s actually, it’d be like saying I’m a Premiere Pro powered studio or I’d be saying I’m a Photoshop powered studio. It’s like that is just a tool in your toolkit, but the reality is is like for me when I’m creating stories, when I’m creating animated worlds, the best work comes when I’m not by myself alone on a computer. For me, that is not the way that I want to make or create work because a big part about being a director, a big part about making work for me is bringing a team of people together so we can reach an audience. And so I’ve been just so amazed to see traditional animators that I’ve worked with in some of my past work come to the table saying we want to explore, we want to make this happen. It will allow us to tell more stories that we can tell. And some really concrete examples of that is, you know, we were trying to put voiceover actors into our film and no one wanted to sign on to the film. Backstage took down the castings. Another voice, one, two, three, took down my casting because I was very clear that it’s being made with the support of generative AI tools. And so then we fell back to 11 labs because we couldn’t cast voiceover actors. And so, you know, being able to use a tool like 11 labs, being able to use even tools like cling or VO because our animators are incredible but they only have so much time. And specifically for me, I’m only making animated films. And animation is such an expensive process and it’s such a time-consuming process. And so to be able to let my creators and my makers and my art directors and my concept artists just make work and then have the AI do things like in-between animation, which is a very time-intensive process, or to allow my score composer to compose score for six or seven minutes and then have the AI be kind of an agent that helps him with the score. So my whole team is just being, I think, empowered and assisted by the AI. And so we often say it’s AI-assisted work or we say that it’s AI-assisted work that has AI assistance to it. And I think that’s really important because it changes the power hierarchy that we as creatives have with the tool we’re using.
Harry Yeff: Well I think the common zeitgeist in narrative currently is the forms of collaboration, these different structures. I wouldn’t say that they’re so front and center. I don’t think it’s very clear to all artists how augmentation can happen and what is collaboration. Of course, what is ethical collaboration? What makes people feel safe and connected to the ways in which they’re creating, which I think is education. And I think this is most common, in my opinion, in music. I think the scale of music, the size of music, and the human experience of music, how profound it is to the human experience, there’s of course so much energy around the conversation. So to Chris, what is the line between creation and exploitation?
Chris Horton: Yeah, I think that it’s very easy for companies to say that AIs learn just like people. But I think if you dig in, that’s not really true. It doesn’t take us 500,000 hours to learn the alphabet. We don’t learn and then stop and then execute after that. We are constantly learning. And we don’t throw our knowledge into a blender and read a piece and then learn how to. We learn entirely different than AI. I think it’s just a shorthand. So I think that we need to recognize the people who make the data that goes in to train these things that has value. The content that is being used to train the AIs has value. And one of the messages that we want to communicate is that it’s a false dichotomy to say that you can either have content protection, copyright, or AI success. That’s not true. I mean, that’s kind of what we heard many years ago in the P2P era, that it was just too hard to get licenses done. And clearly, that wasn’t true because the world’s largest companies, many of now who are building AI systems, managed to license the entire world’s collection of music. So we want to make sure that artists have the ability to get compensated for their creations that are going into feeding these tools. And these are, in many cases, multi-billion dollar companies. And it’s just a matter of not wanting to pay as opposed to they can’t pay or they can’t do the work. And there are responsible companies, some who are sitting on stage who are talking about getting licenses. So we know it’s possible.
Harry Yeff: Yeah, I think it’s so needed, this panel, case in point, for all sides of the power structure to collaborate. I think the concepts of artists in research, the concepts of individual artists working, just the power of the developers of entities, the power of institutions. I think we’re going to develop more structures to be able to think like, what do artists want? What does the average human being desire when it comes to creativity? And what are really powerful tools? So I have to ask, what are your predictions? I mean, dare I say, what are your hopes for where this might go? If there was something small that you would hope for or dream of in terms of how this might fuel the creative experience, what would that be? And I’ll start with Mikayla.
Michaela Ternasky Holland: Oh man, I was hoping I went last. I will say that I am probably more burnt out trying to figure out these AI assisted types of productions, whether that’s in these interactive installations I’m doing, or if it’s in these animated films I’m doing. And so really, like in 2030, the dream would be it’s not just AI’s focus and engineering focus to give us the magic button in creativity. Like, what do you want to just make a thing? I actually really like the creative process of struggling with these systems and struggling with these tools because it really helps me realize this is not a replacement system. Like, if anyone goes out and play with them, you’ll realize that it’s a very difficult process to have control over these items. But what I hope is that we get more of maybe these agentic items that help us with our day to day, like, admin production work, because I’m so overwhelmed as an independent creative and artist and producer who’s trying to do it all and support a team. And I wish that I could not just have maybe that intern or that production assistant who’s amazing, but have something that helps empower them to be even more sustainable. That is also something that is powering and sustaining my creative team as well.
Chris Horton: Yeah, I think that we’ve heard from a number of musicians during this conference. I’ve been pleasantly surprised by the inclusion of music in AI for Good. Thank you, AI for Good. And what we’ve heard, I think, in many cases is that they want a collaborative, maybe an intern, a smart intern that’s confident. And so my crazy prediction, because why not, is that at some point in the future, it may be that when you’re working with someone, you expect that it’s them and their AI assistant. And so as I’m working, my AI is learning how I work and what I do and what I like and what I don’t like. And I can rely on it to do some of the busy grunt work that is the boring part and give me more time to be creative as an artist. And if I have to go in, I have a melody and I decide, you know what, instead of a B, I want that to be a C, and I have to go through and tweak every note. The AI assistant can do that. I don’t need to do that, right? Give me more time. So that has interesting implications about hiring. Now you’re hiring a person and their AI, and what does that mean about confidentiality? It’s a whole mess, but maybe we reach that point where the assumption is we all have AIs that work with us.
James McAulay: I think human curation and the human stories behind all of this art become even more important as the amount of art is just exploding exponentially. After this conference, you could go to a restaurant, you could pull out an app and try and get the optimum wine, and you could put in the data and try to get the optimum wine to go with your meal. Or you could speak to the wine expert and have this theatrical performance and be told about the story of this wine and where this wine came from. And I think as people, we want that. We want to connect with the people behind the stories. So yeah, for me, I think human curation becomes even more important in this AI era.
Harry Yeff: So my role is to continue to facilitate the conversation. I think my responsibility as an artist is to try to create these doors for these unique collaborations, to fight for explosive art, and there is going to be an abundance of media that is pronouncing itself. But I think not only will we have some of the most stunning creations that humanity has ever known, I think these tools, if used properly and if the right conversation is happening, we will be able to understand aspects of the human experience that we still have unanswered. So we must collaborate, we must work together, we must ask the difficult questions, and together I think there are answers to these questions. And the artist and the creative is very, very much a part of that journey. So thank you so much. Thank you.
Harry Yeff
Speech speed
157 words per minute
Speech length
1131 words
Speech time
430 seconds
Artists work intimately with AI technology providing predictions, insights, and raising important questions about collaboration
Explanation
Harry Yeff argues that artists have a unique and intimate relationship with AI technology that serves as a signal for broader implications. He believes that because artists work so closely with this technology, they are able to provide valuable predictions and insights about its potential and challenges.
Evidence
References his experience curating AIVI initiative with ITU and AI for Good, and mentions that artists are now building amazing tools and asking difficult questions through their work with AI
Major discussion point
AI and Creative Collaboration
Topics
Sociocultural | Economic
Belief that proper collaboration and difficult conversations will lead to stunning creations and deeper understanding of human experience
Explanation
Harry Yeff expresses optimism that through collaborative efforts and addressing challenging questions, AI tools can enable humanity to create exceptional art. He emphasizes that this collaborative approach is essential for progress in creative AI applications.
Evidence
States that ‘we must collaborate, we must work together, we must ask the difficult questions’ and that artists and creatives are very much part of this journey
Major discussion point
Future Vision for AI in Creativity
Topics
Sociocultural | Interdisciplinary approaches
Vision that AI tools will help humanity understand previously unanswered aspects of the human experience
Explanation
Harry Yeff believes that AI tools, when used properly with appropriate conversations, will enable humanity to gain insights into aspects of human experience that remain unexplored. This represents his broader vision for AI’s potential beyond just creative applications.
Evidence
States that ‘if used properly and if the right conversation is happening, we will be able to understand aspects of the human experience that we still have unanswered’
Major discussion point
Future Vision for AI in Creativity
Topics
Sociocultural | Interdisciplinary approaches
Michaela Ternasky Holland
Speech speed
201 words per minute
Speech length
1028 words
Speech time
305 seconds
AI should be viewed as a tool in the toolkit rather than defining the creative process, similar to Premiere Pro or Photoshop
Explanation
Michaela argues that the narrative around ‘AI filmmaking’ or ‘AI powered studios’ is incorrect framing. She believes AI should be understood as just another tool in a creator’s toolkit, comparable to established software like Premiere Pro or Photoshop, rather than something that defines or powers the entire creative process.
Evidence
Compares AI to saying ‘I’m a Premiere Pro powered studio’ or ‘I’m a Photoshop powered studio’ to illustrate how AI should be viewed as just another tool
Major discussion point
AI and Creative Collaboration
Topics
Sociocultural | Economic
Agreed with
– Chris Horton
Agreed on
AI as augmentation rather than replacement of human creativity
Personalization through AI can create powerful moments where audiences feel seen and heard, enabling expanded understanding
Explanation
Michaela believes that AI’s ability to customize and personalize content can create meaningful connections with audiences by making them feel acknowledged. From this foundation of feeling seen, creators can then expand what audiences thought they knew about various topics.
Evidence
Mentions her work on projects about gun violence, nuclear weapons threat, and U.S. politics as examples of using personalization to create thought-provoking moments
Major discussion point
Personalization vs. Shared Cultural Experience
Topics
Sociocultural | Content policy
Disagreed with
– Chris Horton
– James McAulay
Disagreed on
Value of personalization versus shared cultural experiences
Context determines whether bespoke, personalized content or universal storytelling is more appropriate
Explanation
Michaela argues that the choice between personalized AI-generated content and universal storytelling depends entirely on the context and goals of the project. Different approaches serve different purposes depending on the intended audience and desired impact.
Evidence
Contrasts her thought-provoking personalized work with mainstream projects like traditional animated films or museum exhibitions that aim for broad universal appeal
Major discussion point
Personalization vs. Shared Cultural Experience
Topics
Sociocultural | Content policy
Agreed with
– Chris Horton
Agreed on
Context determines appropriate use of AI personalization
AI assists teams by handling time-intensive processes like in-between animation and score composition, empowering creators
Explanation
Michaela describes how AI tools help her creative team by taking over labor-intensive tasks such as in-between animation work and assisting with musical score composition. This allows human creators to focus on higher-level creative decisions while AI handles the time-consuming technical work.
Evidence
Provides specific examples of using AI for in-between animation (a time-intensive process) and having AI assist score composers, mentions that animation is expensive and time-consuming
Major discussion point
AI as Augmentation Rather Than Replacement
Topics
Economic | Future of work
AI tools help overcome practical challenges like casting difficulties while supporting traditional creative teams
Explanation
Michaela explains how AI tools can solve practical production problems, such as when traditional approaches fail. She demonstrates how AI can provide solutions while still maintaining collaborative relationships with traditional creative professionals who want to explore new technologies.
Evidence
Describes how Backstage and Voice123 took down her casting calls because she mentioned using generative AI, forcing her to use ElevenLabs for voiceover work, and mentions traditional animators wanting to explore AI to tell more stories
Major discussion point
AI as Augmentation Rather Than Replacement
Topics
Economic | Future of work
Disagreed with
– Chris Horton
Disagreed on
Approach to AI compensation and licensing
The creative process involves struggling with AI systems, which reinforces that they are not replacement systems
Explanation
Michaela argues that working with AI tools is actually quite difficult and requires significant effort to maintain control over the output. This struggle demonstrates that AI is not simply replacing human creativity but requires active human engagement and skill.
Evidence
States that she likes ‘the creative process of struggling with these systems and struggling with these tools’ and that ‘if anyone goes out and play with them, you’ll realize that it’s a very difficult process to have control over these items’
Major discussion point
AI as Augmentation Rather Than Replacement
Topics
Economic | Future of work
Agreed with
– Chris Horton
Agreed on
AI as augmentation rather than replacement of human creativity
Hope for AI agents that help with administrative production work to support independent creators and their teams
Explanation
Michaela expresses her desire for AI systems that can assist with the overwhelming administrative and production tasks that independent creators face. She envisions AI not just as a creative tool but as support for the business and logistical aspects of creative work.
Evidence
Describes being ‘overwhelmed as an independent creative and artist and producer who’s trying to do it all and support a team’ and wanting AI that helps empower production assistants and sustain creative teams
Major discussion point
Future Vision for AI in Creativity
Topics
Economic | Future of work
Chris Horton
Speech speed
188 words per minute
Speech length
787 words
Speech time
250 seconds
AI can enhance artist creativity and deepen connections with fan bases through interactive experiences
Explanation
Chris Horton outlines Universal Music Group’s two main goals for AI: enhancing artists’ creative capabilities and strengthening the relationship between artists and their fans. He suggests that AI can enable new forms of fan interaction, including allowing fans to customize music experiences.
Evidence
Mentions that UMG is exploring ways for fans to interact with and possibly customize music, though they are still in early stages of determining artist comfort levels
Major discussion point
AI and Creative Collaboration
Topics
Economic | Digital business models
Agreed with
– James McAulay
Agreed on
AI enables new forms of fan and audience engagement
There’s tension between customized music experiences and the shared cultural moments of hearing the same song together
Explanation
Chris highlights a potential cultural concern about AI personalization in music, noting that shared musical experiences create important cultural bonds. He questions what happens to communal experiences like singing along at concerts if everyone hears slightly different versions of songs in private.
Evidence
References the cultural context of everyone hearing the same song on radio and singing lyrics together at shows, questioning the impact if private listening becomes personalized
Major discussion point
Personalization vs. Shared Cultural Experience
Topics
Sociocultural | Cultural diversity
Agreed with
– Michaela Ternasky Holland
Agreed on
Context determines appropriate use of AI personalization
Disagreed with
– Michaela Ternasky Holland
– James McAulay
Disagreed on
Value of personalization versus shared cultural experiences
AI learning differs fundamentally from human learning, and the content used to train AI systems has value that should be compensated
Explanation
Chris argues against the common claim that AI learns like humans, pointing out key differences in learning processes. He emphasizes that the content used to train AI systems has inherent value and that creators should be compensated for their contributions to these training datasets.
Evidence
Points out that humans don’t need 500,000 hours to learn the alphabet, don’t learn then stop and execute, are constantly learning, and don’t throw knowledge into a blender
Major discussion point
Content Rights and Fair Compensation
Topics
Legal and regulatory | Intellectual property rights
It’s a false dichotomy to claim you can’t have both content protection and AI success, as licensing is possible
Explanation
Chris refutes the argument that copyright protection and AI development are mutually exclusive, drawing parallels to past technology transitions. He argues that licensing solutions are feasible and that claims of impossibility are often about unwillingness to pay rather than technical barriers.
Evidence
References the P2P era when similar claims were made, noting that the world’s largest companies (many now building AI) successfully licensed the entire world’s music collection
Major discussion point
Content Rights and Fair Compensation
Topics
Legal and regulatory | Intellectual property rights
Disagreed with
– Michaela Ternasky Holland
Disagreed on
Approach to AI compensation and licensing
Responsible companies are already working on getting proper licenses for AI training data
Explanation
Chris points out that some companies are already taking the responsible approach by securing proper licenses for the content they use to train AI systems. This demonstrates that ethical AI development with proper compensation is not only possible but already happening.
Evidence
References that there are responsible companies, including some represented on the panel, who are talking about getting licenses
Major discussion point
Content Rights and Fair Compensation
Topics
Legal and regulatory | Intellectual property rights
AI should function as a collaborative smart intern that handles grunt work, giving artists more time for creativity
Explanation
Chris envisions AI as an intelligent assistant that learns individual work patterns and preferences, taking over mundane tasks to free up creative time. He suggests this could become a standard expectation in creative work, where people work alongside their AI assistants.
Evidence
Gives example of AI handling technical tasks like changing a B note to C and tweaking every related note, allowing artists to focus on creative decisions rather than tedious execution
Major discussion point
AI as Augmentation Rather Than Replacement
Topics
Economic | Future of work
Agreed with
– Michaela Ternasky Holland
Agreed on
AI as augmentation rather than replacement of human creativity
Prediction that future collaboration will assume people work with their AI assistants, creating new implications for hiring and confidentiality
Explanation
Chris predicts a future where hiring someone means hiring them along with their AI assistant, as the AI learns their work patterns and becomes integral to their productivity. This raises new questions about workplace dynamics, confidentiality, and the nature of employment.
Evidence
Describes AI learning how individuals work, what they like and don’t like, and mentions this creates ‘interesting implications about hiring’ and ‘what does that mean about confidentiality’
Major discussion point
Future Vision for AI in Creativity
Topics
Economic | Future of work
James McAulay
Speech speed
162 words per minute
Speech length
441 words
Speech time
162 seconds
AI enables creation of interactive experiences like conversational agents based on film characters and anthropomorphized objects
Explanation
James describes how AI voice technology is being used to create new forms of interactive entertainment and educational experiences. These include allowing audiences to have conversations with film characters and giving voices to inanimate objects or abstract concepts to create emotional connections.
Evidence
Mentions a film producer exploring AI agents based on movie characters for fan interaction, and references Harry Yeff’s ‘Voicing Nature’s Cube’ where people can talk to extinct animals and melting icebergs
Major discussion point
AI and Creative Collaboration
Topics
Sociocultural | Digital identities
Agreed with
– Chris Horton
Agreed on
AI enables new forms of fan and audience engagement
Human curation and stories behind art become more important as AI-generated content explodes exponentially
Explanation
James argues that as AI makes it easier to create vast amounts of content, the human element becomes more valuable and necessary. He believes people will increasingly seek out the personal stories and human expertise behind creative works rather than just consuming algorithmically optimized content.
Evidence
Uses analogy of choosing between an app that provides optimal wine recommendations versus speaking to a wine expert who provides theatrical performance and stories about the wine’s origin
Major discussion point
Personalization vs. Shared Cultural Experience
Topics
Sociocultural | Cultural diversity
Disagreed with
– Michaela Ternasky Holland
– Chris Horton
Disagreed on
Value of personalization versus shared cultural experiences
Agreements
Agreement points
AI as augmentation rather than replacement of human creativity
Speakers
– Michaela Ternasky Holland
– Chris Horton
Arguments
AI should be viewed as a tool in the toolkit rather than defining the creative process, similar to Premiere Pro or Photoshop
The creative process involves struggling with AI systems, which reinforces that they are not replacement systems
AI should function as a collaborative smart intern that handles grunt work, giving artists more time for creativity
Summary
Both speakers emphasize that AI should augment human creativity rather than replace it, with AI handling technical or administrative tasks while humans focus on creative decisions
Topics
Economic | Future of work
AI enables new forms of fan and audience engagement
Speakers
– James McAulay
– Chris Horton
Arguments
AI enables creation of interactive experiences like conversational agents based on film characters and anthropomorphized objects
AI can enhance artist creativity and deepen connections with fan bases through interactive experiences
Summary
Both speakers see AI as creating new opportunities for interactive experiences that strengthen connections between creators and their audiences
Topics
Sociocultural | Digital identities
Context determines appropriate use of AI personalization
Speakers
– Michaela Ternasky Holland
– Chris Horton
Arguments
Context determines whether bespoke, personalized content or universal storytelling is more appropriate
There’s tension between customized music experiences and the shared cultural moments of hearing the same song together
Summary
Both speakers recognize that the choice between personalized and universal content depends on context and intended outcomes, acknowledging both benefits and potential drawbacks of personalization
Topics
Sociocultural | Content policy
Similar viewpoints
All three speakers view AI as a collaborative tool that enhances human creativity and enables new forms of artistic expression when used thoughtfully
Speakers
– Harry Yeff
– Michaela Ternasky Holland
– Chris Horton
Arguments
Artists work intimately with AI technology providing predictions, insights, and raising important questions about collaboration
AI assists teams by handling time-intensive processes like in-between animation and score composition, empowering creators
AI should function as a collaborative smart intern that handles grunt work, giving artists more time for creativity
Topics
Sociocultural | Economic
Both speakers envision AI assistants becoming integral to creative work, handling administrative and technical tasks to support human creativity
Speakers
– Michaela Ternasky Holland
– Chris Horton
Arguments
Hope for AI agents that help with administrative production work to support independent creators and their teams
Prediction that future collaboration will assume people work with their AI assistants, creating new implications for hiring and confidentiality
Topics
Economic | Future of work
Both speakers emphasize the continued importance of human elements and shared cultural experiences even as AI capabilities expand
Speakers
– James McAulay
– Chris Horton
Arguments
Human curation and stories behind art become more important as AI-generated content explodes exponentially
There’s tension between customized music experiences and the shared cultural moments of hearing the same song together
Topics
Sociocultural | Cultural diversity
Unexpected consensus
AI as a difficult tool requiring human skill and struggle
Speakers
– Michaela Ternasky Holland
Arguments
The creative process involves struggling with AI systems, which reinforces that they are not replacement systems
Explanation
Unexpectedly, rather than portraying AI as making creativity easier, Michaela emphasizes that working with AI is actually difficult and requires significant human effort and skill, challenging common narratives about AI automation
Topics
Economic | Future of work
Industry representative advocating for content creator rights
Speakers
– Chris Horton
Arguments
AI learning differs fundamentally from human learning, and the content used to train AI systems has value that should be compensated
It’s a false dichotomy to claim you can’t have both content protection and AI success, as licensing is possible
Explanation
Unexpectedly, the representative from a major music corporation strongly advocates for artist compensation and content protection, challenging the narrative that industry and creator interests are opposed
Topics
Legal and regulatory | Intellectual property rights
Overall assessment
Summary
The speakers demonstrate strong consensus on AI as an augmentative tool rather than replacement, the importance of human curation and collaboration, and the need for context-appropriate applications of AI personalization
Consensus level
High level of consensus with complementary perspectives rather than conflicting viewpoints. This suggests a mature understanding of AI’s role in creativity that balances technological capabilities with human values and practical considerations. The agreement across different stakeholder perspectives (independent artist, major corporation, and technology company) indicates potential for collaborative solutions in the creative AI space.
Differences
Different viewpoints
Approach to AI compensation and licensing
Speakers
– Chris Horton
– Michaela Ternasky Holland
Arguments
It’s a false dichotomy to claim you can’t have both content protection and AI success, as licensing is possible
AI tools help overcome practical challenges like casting difficulties while supporting traditional creative teams
Summary
Chris emphasizes the need for formal licensing and compensation structures for AI training data, while Michaela focuses on practical AI adoption to solve immediate production challenges, suggesting different priorities in addressing AI ethics
Topics
Legal and regulatory | Economic
Value of personalization versus shared cultural experiences
Speakers
– Michaela Ternasky Holland
– Chris Horton
– James McAulay
Arguments
Personalization through AI can create powerful moments where audiences feel seen and heard, enabling expanded understanding
There’s tension between customized music experiences and the shared cultural moments of hearing the same song together
Human curation and stories behind art become more important as AI-generated content explodes exponentially
Summary
Michaela advocates for personalization as a tool for social impact and connection, Chris warns about losing shared cultural moments through customization, and James emphasizes human curation over algorithmic personalization
Topics
Sociocultural | Cultural diversity
Unexpected differences
Role of struggle and difficulty in AI creative processes
Speakers
– Michaela Ternasky Holland
– Chris Horton
Arguments
The creative process involves struggling with AI systems, which reinforces that they are not replacement systems
AI should function as a collaborative smart intern that handles grunt work, giving artists more time for creativity
Explanation
Unexpectedly, Michaela values the difficulty of working with AI as part of the creative process, while Chris envisions AI as seamlessly handling mundane tasks. This represents a fundamental disagreement about whether AI should be challenging to use or effortlessly helpful
Topics
Economic | Future of work
Overall assessment
Summary
The speakers show moderate disagreement on implementation approaches rather than fundamental goals, with tensions around personalization versus shared culture, formal licensing versus practical adoption, and the role of difficulty in AI creative processes
Disagreement level
Moderate disagreement with collaborative potential – while speakers have different priorities and approaches, they share common ground on AI as augmentation rather than replacement, suggesting productive dialogue is possible despite differing perspectives on execution
Partial agreements
Partial agreements
Similar viewpoints
All three speakers view AI as a collaborative tool that enhances human creativity and enables new forms of artistic expression when used thoughtfully
Speakers
– Harry Yeff
– Michaela Ternasky Holland
– Chris Horton
Arguments
Artists work intimately with AI technology providing predictions, insights, and raising important questions about collaboration
AI assists teams by handling time-intensive processes like in-between animation and score composition, empowering creators
AI should function as a collaborative smart intern that handles grunt work, giving artists more time for creativity
Topics
Sociocultural | Economic
Both speakers envision AI assistants becoming integral to creative work, handling administrative and technical tasks to support human creativity
Speakers
– Michaela Ternasky Holland
– Chris Horton
Arguments
Hope for AI agents that help with administrative production work to support independent creators and their teams
Prediction that future collaboration will assume people work with their AI assistants, creating new implications for hiring and confidentiality
Topics
Economic | Future of work
Both speakers emphasize the continued importance of human elements and shared cultural experiences even as AI capabilities expand
Speakers
– James McAulay
– Chris Horton
Arguments
Human curation and stories behind art become more important as AI-generated content explodes exponentially
There’s tension between customized music experiences and the shared cultural moments of hearing the same song together
Topics
Sociocultural | Cultural diversity
Takeaways
Key takeaways
AI should be viewed as a collaborative tool that augments human creativity rather than replacing it, similar to other creative software like Photoshop or Premiere Pro
The context determines whether personalized/bespoke content or universal storytelling is more appropriate – personalization can create powerful moments of connection while shared cultural experiences maintain collective bonds
AI enables new forms of interactive storytelling and audience engagement, such as conversational agents based on characters or anthropomorphized data
Human curation and the stories behind art become increasingly important as AI-generated content proliferates exponentially
AI learning fundamentally differs from human learning, and content creators deserve compensation for their work being used to train AI systems
Proper licensing and fair compensation for AI training data is both possible and necessary, as demonstrated by successful music licensing models
The future likely involves people working alongside their AI assistants as collaborative partners, with implications for hiring practices and confidentiality
Resolutions and action items
Continue facilitating conversations between different stakeholders in the AI and creative industries
Fight for ‘explosive art’ and create opportunities for unique collaborations between artists, technologists, and institutions
Ask difficult questions about ethical AI collaboration and work together to find answers
Develop better structures to understand what artists want and what constitutes powerful creative tools
Unresolved issues
How to balance personalized AI-generated content with shared cultural experiences in music and other media
What constitutes ethical collaboration between AI systems and human creators
How to make artists feel safe and connected in their creative processes with AI
The implications of hiring ‘a person and their AI’ including confidentiality concerns
How to effectively educate artists about augmentation and collaboration possibilities with AI
The ongoing challenge of casting and industry acceptance of AI-assisted creative work
How to develop better administrative AI agents to support independent creators with production work
Suggested compromises
Coexistence model where songs are released as originally imagined by artists but also allow for customization in certain contexts
AI-assisted rather than AI-powered approach to creative work, maintaining human creative control while leveraging AI for time-intensive tasks
Recognition that responsible AI companies should work on proper licensing while acknowledging that some are already doing so
Acceptance that struggling with AI systems is part of the creative process and helps maintain the human element in creation
Thought provoking comments
I think it all depends on context… what I do think is really powerful about customization and personalization, which is something that generative does give us the ability to do, is that it can really kind of give somebody a very clear moment where they feel seen or they feel heard. And from that moment, we can then expand what they thought they knew.
Speaker
Michaela Ternasky Holland
Reason
This comment reframes the debate about bespoke media from a technical capability question to a strategic tool for social impact. She introduces the concept that personalization isn’t just about preference but about creating empathy and understanding as a gateway to broader learning and change.
Impact
This shifted the conversation from abstract concerns about infinite creativity to concrete applications. It established a framework that other panelists built upon, moving the discussion toward practical considerations of how AI tools serve different purposes in different contexts.
And if what I hear when I’m listening privately is slightly different than what you hear, what is the impact of that on that shared cultural moment? And so it may be that these things coexist, that the song is still released as imagined by the artist, but there may be context where you can customize.
Speaker
Chris Horton
Reason
This comment introduces a profound philosophical question about the nature of shared cultural experiences in an age of personalization. It challenges the assumption that customization is inherently good by highlighting what might be lost – the collective experience of culture.
Impact
This comment elevated the discussion from technical possibilities to cultural implications, introducing the concept that limitation might actually enhance human experience. It provided a counterbalance to the enthusiasm for personalization and influenced the moderator’s follow-up about how ‘limitation brings quality to the human experience.’
It’s actually, it’d be like saying I’m a Premiere Pro powered studio or I’d be saying I’m a Photoshop powered studio. It’s like that is just a tool in your toolkit… And so we often say it’s AI-assisted work or we say that it’s AI-assisted work that has AI assistance to it. And I think that’s really important because it changes the power hierarchy that we as creatives have with the tool we’re using.
Speaker
Michaela Ternasky Holland
Reason
This analogy fundamentally reframes how we should think about AI in creative work – not as a replacement or even a collaborator, but as a sophisticated tool like existing creative software. The insight about ‘power hierarchy’ is particularly profound, suggesting that language shapes our relationship with technology.
Impact
This comment directly challenged the prevailing narrative of AI replacement and shifted the conversation toward a more nuanced understanding of human-AI collaboration. It provided a practical framework that other panelists could relate to and influenced the moderator’s question about ‘augmented process’ versus ‘pure automation.’
I think it’s very easy for companies to say that AIs learn just like people. But I think if you dig in, that’s not really true… So I think that we need to recognize the people who make the data that goes in to train these things that has value.
Speaker
Chris Horton
Reason
This comment cuts through a common industry talking point with technical precision, exposing the false equivalency between human and AI learning. It then connects this technical insight to ethical and economic implications about compensation and value creation.
Impact
This was a pivotal moment that introduced the ethical and economic dimensions of AI development. It shifted the conversation from creative possibilities to questions of fairness and sustainability, directly addressing the ‘creation vs exploitation’ question and grounding the discussion in real-world power dynamics.
I think human curation and the human stories behind all of this art become even more important as the amount of art is just exploding exponentially… We want to connect with the people behind the stories.
Speaker
James McAulay
Reason
This insight suggests that AI abundance doesn’t diminish human value but actually increases the importance of human connection and storytelling. The wine expert analogy effectively illustrates how human expertise becomes more valuable, not less, in an automated world.
Impact
This comment provided an optimistic counterpoint to concerns about AI replacing human creativity. It suggested that rather than competing with AI, humans should focus on what they uniquely provide – story, context, and connection. This helped conclude the discussion on a forward-looking note about human-AI collaboration.
Overall assessment
These key comments transformed what could have been a surface-level discussion about AI tools into a nuanced exploration of creativity, culture, and human values. Michaela’s contextual framework and tool analogy provided practical grounding, while Chris’s insights about shared cultural experiences and AI learning challenged common assumptions. James’s perspective on human curation offered a hopeful vision for the future. Together, these comments created a progression from technical capabilities to cultural implications to ethical considerations to future possibilities, demonstrating how thoughtful commentary can elevate a panel discussion beyond promotional talking points to genuine intellectual exploration.
Follow-up questions
What is the impact on shared cultural moments when personalized music experiences differ from the original shared version?
Speaker
Chris Horton
Explanation
This explores how customization of music might affect the collective cultural experience of singing the same lyrics together at concerts or sharing common musical references
How will artists and fans respond to customizable music experiences in terms of comfort levels and adoption?
Speaker
Chris Horton
Explanation
UMG is in early stages of determining artist and fan comfort with music customization, requiring further research to understand market acceptance
What are the implications of hiring practices when employees work with AI assistants?
Speaker
Chris Horton
Explanation
This raises questions about employment structures, confidentiality, and workplace dynamics when AI assistants become integral to individual work processes
How can sustainable production workflows be developed for independent creatives using AI tools?
Speaker
Michaela Ternasky Holland
Explanation
There’s a need to research better administrative and production support systems that can help independent artists manage the complexity of AI-assisted creative work
What constitutes ethical collaboration between humans and AI in creative processes?
Speaker
Harry Yeff
Explanation
This addresses the need to define boundaries and best practices for human-AI collaboration that make creators feel safe and connected to their work
How can proper licensing and compensation structures be established for content used in AI training?
Speaker
Chris Horton
Explanation
This involves developing fair systems to compensate artists whose work is used to train AI systems, moving beyond the false dichotomy of content protection versus AI success
What aspects of the human experience might be better understood through proper use of AI creative tools?
Speaker
Harry Yeff
Explanation
This suggests research into how AI-assisted creativity might reveal new insights about human nature and experience that remain unanswered
Disclaimer: This is not an official session record. DiploAI generates these resources from audiovisual recordings, and they are presented as-is, including potential errors. Due to logistical challenges, such as discrepancies in audio/video or transcripts, names may be misspelled. We strive for accuracy to the best of our ability.
Related event
