Dr Leon Furze

In the past 18 months, multimodal generative AI has become increasingly ubiquitous, influencing the way many create and interact with content across various domains, including text, image, audio, and video. While generative AI doesn’t “democratise creativity“, it will impact upon many creative and academic areas. As these technologies continue to advance and permeate our daily lives, it is important to develop frameworks that help us understand how people engage with and are affected by generative AI. By examining the roles individuals play in relation to these technologies, we can encourage more responsible and mindful use of generative AI tools.

Many frameworks have been proposed to conceptualise the ways in which people interact with digital technologies. One notable example from within my domain – English and literacy – is Bill Green’s 3D model, which considers three key dimensions: operational, cultural, and critical. The operational dimension focuses on the technical skills and competencies required to use technology effectively. The cultural dimension explores the social and contextual factors that shape technology use and the meanings derived from it. Finally, the critical dimension encourages individuals to question and analyse the power structures, biases, and implications associated with digital technologies.

Recently, Lucinda McKnight, an educator and researcher, and my PhD supervisor, has revisited Green’s 3D model in light of the rapid development of generative AI. In her blog post titled Putting criticality first in the age of AI, McKnight argues for a reordering of the three dimensions, placing the critical dimension at the forefront. She emphasises the importance of critically evaluating whether to use generative AI at all, considering the various risks and potential consequences associated with these technologies.

McKnight highlights several key concerns surrounding generative AI, including legal issues, personal and privacy risks, the perpetuation of biases, threats to democracy, and environmental costs associated with developing and maintaining large language models. By prioritising the critical dimension, McKnight encourages educators and students to thoroughly examine these risks and weigh them against the potential benefits before engaging with generative AI tools.

In the context of generative AI, McKnight proposes the following definitions for the three dimensions of Green’s model:

• Critical: Understanding the power dynamics and potential consequences related to generative AI, such as issues of access, cost, and environmental impact.
• Cultural: Exploring how generative AI is used to create and shape meaning across diverse contexts, such as in art, literature, and communication.
• Operational: Developing the technical skills and competencies needed to effectively use generative AI tools, such as crafting clear and instructive prompts.

By foregrounding the critical dimension, McKnight’s updated framework provides a valuable starting point for individuals and organisations seeking to navigate the complex landscape of multimodal generative AI responsibly and thoughtfully.

Critic, Creator, Consumer

Building upon the foundation laid by Green’s 3D model and McKnight’s recent update, my “Critic, Creator, Consumer” framework offers a complementary perspective on how individuals engage with multimodal generative AI. This framework identifies three primary roles:

• Critic: Individuals who critically analyse and evaluate generative AI technologies, their implications, and the content produced using these tools. In the context of GenAI, a critic might be an outspoken opponent, for example an individual who decides to avoid the use of GenAI on ethical grounds. A ‘conscientious objector’ to GenAI might challenge the technology due to its environmental impact, the concentration of power in the industry, or the possible threat to creative pursuits.
• Creator: Those who actively use generative AI to produce content across various modalities, such as text, image, audio, and video. With no ‘critical’ aspect, these creators might be seen producing huge volumes of work – text, images, music and so on – or they might be considered an “AI artist”, using the tools as their chosen medium.
• Consumer: People who engage with and consume content created using generative AI tools. Consumers may knowingly or unknowingly consume AI-generated content, for example unthinkingly sharing AI-generated deepfakes, or deliberately reading content generated by AI such as LLM-created summaries and transcripts.

The Venn diagram structure (who in education doesn’t love a Venn diagram?) highlights the fact that these roles are not mutually exclusive; individuals may occupy multiple roles simultaneously or move between them depending on the context.

Exploring the intersections

All of the most interesting things in life happen at intersections and in liminal spaces, and it’s no different with this framework. The individual sections of the Venn – critic, consumer, creator – are fairly self explanatory, but at the intersections there is much more nuance. Realistically, people don’t fall into clear cut categories, so these overlapping areas also offer more interesting ways of viewing how we interact with these technologies.

Critic-Consumer

This intersection represents individuals who consume generative AI content while critically evaluating its quality, authenticity, and potential biases. By applying critical thinking skills, these individuals can make informed decisions about the content they engage with and share.

It’s absolutely necessary to be a critical consumer of AI, lest we be manipulated by individuals and organisations using the technology to deliberately mislead or spread misinformation. Recent advances in deepfake technology like Microsoft’s VASA-1 research and the release of applications such as Synthesia’s updated avatars make it even more likely that you will encounter a convincing deepfake sooner rather than later.

Convincing image, audio, and video deepfake technology means we will need a renewed focus on media and critical literacies. Dr Jasper Roe and Dr Mike Perkins, co-authors of the AI Assessment Scale, recently published a pre-print of a research agenda for deepfakes in education which is well worth a read.

To illustrate, here are a few examples of different approaches or roles which might adopt the critic-consumer perspective:

• A media-literate social media user who pauses to critically evaluate posts that might contain deepfakes before sharing them, in order to avoid spreading misinformation.
• A teacher whose role is to educate students in critical literacy skills, such as how to identify potential deepfakes and think critically about the AI-generated content they encounter.
• A journalist who carefully fact-checks any AI-generated images, videos or quotes before including them in a news story, to ensure accuracy and maintain public trust.
• A political campaign staffer who is on the lookout for deepfake videos that could be used to smear their candidate, and is ready to debunk them if they appear.
• An official investigating a case who must use digital forensic tools to discern which evidence gathered from social media or surveillance footage is authentic vs potentially fabricated by AI.
• A parent who wants to help their child develop critical thinking skills to navigate the AI-generated content they will increasingly encounter online as they grow up.

Creator-Critic

Creators who also occupy the critic role are more likely to consider the ethical implications and potential consequences of their work. They may actively seek to mitigate biases, ensure responsible use of data, and create content that aligns with their values. These creators might also use the technology in ways which are deliberately provocative, in order to highlight the ethical concerns.

From artists using cats and bees to critique AI systems, to art projects like Kate Crawford’s Anatomy of an AI System, artists, researchers, and industry experts alike are using art as a medium for critiquing and interrogating artificial intelligence.

Here are a few possibilities for this intersection:

• An artist who uses AI image generation tools in their creative process but critically reflects on the implications of AI on art and authorship.
• A game developer leveraging AI to create more immersive worlds while considering the ethical challenges of AI-driven content and player interactions.
• A filmmaker utilizing AI-powered video editing tools who remains mindful of their responsibility to depict events and people authentically.
• An author writing a science fiction novel about AI who researches the technology thoroughly to ensure their depiction is grounded in reality.
• A journalist using AI writing tools to draft articles more efficiently while upholding journalistic integrity and fact-checking standards.
• A marketer employing AI to personalize ad content while critically evaluating the line between persuasion and manipulation.

Creator-Consumer

Individuals who both create and consume generative AI content have a unique perspective on the capabilities and limitations of these technologies. They may draw inspiration from the work of others while also contributing to the growing body of generative AI-powered content. However, lacking the critical aspect they may also fall prey to “hype” and unconsciously be contributing to some of the issues this technology represents.

It can be exciting and fun to “play” with new technologies, and generative AI is no different. As soon as applications like Udio and Suno are released, for example, hundreds of thousands of users dive in a begin to create music, images, or whatever else the platform is designed for. Unfortunately this can result in a tsunami of low-quality synthetic media, or what I call Digital Plastic. Digital plastic, like its real world counterpart, threatens to clog the waterways of the internet and reduce online spaces to islands of generated content.

My framework, much like Green’s 3D model, highlights that without the critical dimension we run the risk of being used by the technology, rather than using it in ways which are creative but thoughtful.

Here are a few examples:

• A digital artist who uses AI tools to generate images for their projects while also following and being inspired by other AI artists, but with little concern for the intellectual property or copyright concerns raised by AI.
• A filmmaker who employs AI video editing software and keenly studies how AI is being used in films they watch, who is less interested in how video generation models are constructed.
• A game streamer who leverages AI to create unique gameplay challenges and learns strategies by watching other AI-assisted streams, while being unaware of the broader impact of GenAI on the industry.
• A writer who uses GenAI writing tools and reads works by other authors experimenting with AI to inform their own craft, but who does not fully understand the implications of unlicensed work in datasets.
• A musician who generates novel, but ultimately derivative, sounds using AI and seeks inspiration from AI-influenced albums and artists they enjoy.
• An entrepreneur who develops AI-powered products and services while being an avid user of other AI tools in their work and life, who disregards ethical concerns with GenAI because “everyone is using it”.

Critic-Creator-Consumer

The intersection of all three roles represents an ideal scenario in which individuals can critically engage with generative AI, create content responsibly, and consume content mindfully. By occupying all three roles, individuals can contribute to a more balanced and thoughtful approach to generative AI use, and remain wary of hype and the influence of the technology companies responsible for developing GenAI.

Unlike the critics and conscientious objectors, I have never advocated people cease using Generative AI. Despite the ethical considerations – and there are many – the technology is not going to go away if we simply avoid it. In order to avoid the passive consumption or unthinking creation of Digital Plastic, we need to add the balance of critique.

This, clearly, is the role of education in digital technologies – not only to teach students how to use the technology, but also why, when, and occasionally why not. To do that successfully, educators also need to locate themselves at the intersection of all three areas of the framework.

• An AI ethicist who develops guidelines for responsible AI use, critiques existing applications, and uses AI tools in their research.
• A tech journalist who reports on AI advancements, offers informed critiques, and uses GenAI writing tools in their work.
• An educator teaching AI literacy, creating AI-powered learning tools, and using AI to personalise their teaching through differentiated lesson resources.
• A policymaker drafting AI regulations, evaluating AI’s societal impact, and using AI tools to analyse data and draft legislation, or to synthesise policies from other jurisdictions.
• A digital artist creating AI art, critiquing the AI art scene, and drawing inspiration from the AI-generated works they consume as well as using a blend of traditional and AI methods.
• An AI startup founder developing innovative products, critically examining the implications of their work, and using AI daily to optimise their business operations, while aiming to build fairly trained models which use licensed data.

Implications and applications

The “Critic, Creator, Consumer” framework can be applied across various contexts to guide responsible engagement with generative AI:

• Education: Educators can use the framework to design curricula that celebrate critical thinking, creative expression, and informed consumption of generative AI content. By encouraging students to occupy all three roles, educators can help develop a generation of responsible and empowered generative AI users. This can be achieved across a range of curriculum areas – I’m not advocating for separate courses on “AI literacy”.
• Industry: Companies and organisations can apply the framework to develop best practices and guidelines for generative AI use. By promoting a balance between the three roles, businesses can create a culture of responsible innovation and mindful consumption. Learning and development courses focused on GenAI should include aspects of all three elements of the framework, as well as a clear articulating of which tasks and roles fall to the intersections.
• Personal use: Individuals can use the framework to reflect on their own engagement with generative AI and make informed decisions about their role in the ecosystem. By striving to occupy all three roles, individuals can contribute to a more balanced and thoughtful approach to generative AI use in their personal lives. For creators, this may encourage a brief pause or slow-down before immediately turning to GenAI, or it may steer certain tasks away from AI entirely.

Conclusion

The “Critic, Creator, Consumer” framework provides a valuable lens for understanding the complex ways in which individuals interact with multimodal generative AI. By considering the intersections between these three roles, we can hopefully develop a more nuanced appreciation of the opportunities and challenges associated with these technologies.

The development of multimodal generative artificial intelligence is not slowing down: in fact, these technologies are increasing in their multimodality and ubiquity. Generative AI will also inevitably “recede into the woodwork” at some point, following the trajectory of other technological systems like electricity, telephones, and the internet. Once it has disappeared from public consciousness (when was the last time you thought about “going on the electricity grid”?), it will be all the more difficult to critique. By equally valuing our roles as critics, creators, and consumers, we can work towards a future in which generative AI is used responsibly, ethically, by ourselves and our students.

Frameworks are useful for helping us to understand complex issues. But of course, any framework is a simplification, and I’m not meaning to suggest that every user (or non-user) of Generative AI will fit neatly into these categories. But it can be helpful to both interrogate your own use of GenAI, and to help others understand where they fit. There will also be times at which you move from one section to another, consciously or unconsciously. For example, I might be creating content and be totally swept up in an uncritical use of an impressive new application, only to later step back and view what I have created, and the application itself, through a more critical lens.

Play around with the framework and think about how it applies to your role, particularly if, like me, you work in education. How might it be useful in articulating the ways we use Generative AI with your students?

If you’d like to discuss generative AI, consulting, advisory services, or professional development, please get in touch via the form below:

Cite this article: Furze, L. (2024, April 29). Critic, creator, consumer: A framework for using (or being used by) GenAI. Leon Furze. https://leonfurze.com/2024/04/30/critic-creator-consumer-a-framework-for-using-or-being-used-by-genai/