Interview: Andy Hood on how AI represents the “next step in creative technology”

Hood began his media career at AKQA in 1999 during the early days of interactive media, focusing on interactive technologies like Flash. His tenure at AKQA, which spanned over two decades, allowed him to delve into various creative technologies, from virtual reality (VR) and augmented reality (AR) to artificial intelligence (AI).

Read more IBC2024 Accelerator Project: AI Media Production Lab

At WPP, Hood’s responsibilities include helping teams and clients understand the strategic value of new technologies. This involves evaluating which technologies are likely to take off, understanding how they work together, and determining their potential impact on consumer behaviour.

Current perspective on AI

“Generative AI is radical in terms of the process, but it is a creative tool and kind of the next step in creative technology”

Reflecting on his time at AKQA, Hood shares his early experiences with AI through working with IBM Watson. “In my AKQA time back in 2016-2017, we were working with IBM Watson when it was first commercialised and doing some interesting things in terms of conversation, analysis, and search and really getting a good picture of how AI could be used within a business, or in UI/UX and solving problems,” he recounts. This early exposure laid the foundation for understanding AI’s potential in business and creative fields.

Hood points out that AI encompasses a wide range of applications beyond the popular generative models like ChatGPT and DALL-E, highlighting that AI includes applications in machine learning, audience analysis, conversational interfaces, and more.

He acknowledges the transformative impact of generative AI tools such as ChatGPT and MidJourney, which have brought AI into mainstream consciousness. “Generative AI is radical in terms of the process, but it is a creative tool and kind of the next step in creative technology,” he says. However, he cautions that AI should be seen as a tool that enhances human creativity rather than replacing it. “I think what gets missed a lot of the time is in articles with an image of something and a caption, ‘this was done by AI.’ The reality is that it was done by a person who was using AI.”

Hood notes that AI technologies have become more accessible over time, democratising their use. “There were things that we were able to prototype when Watson was commercialised that a couple of years previously, we would have had to be talking to IBM to do,” he recalls. This increased accessibility has allowed a wider range of developers and creative professionals to experiment with AI.

Hood is keenly aware of the ethical and regulatory challenges associated with AI. He underscores the importance of using AI responsibly and ensuring compliance with ethical standards highlighting WPP’s commitment to ethical AI use.

He believes that generative AI serves as a radical step in creative technology but ultimately remains a tool for creatives to use. “It’s about using those models which are copyright-free or trained using the right sources in the right ways.”

Collaboration: AI and the creative industries

One of Hood’s central themes is the collaboration between AI and humans. He argues against the fear-driven narrative that AI will replace human creativity. Instead, he believes that AI can be a tool used to enhance and amplify human creativity. He draws parallels between the advent of Photoshop and the current AI revolution. Just as designers did not lose their jobs to Photoshop, but instead became more effective by mastering it, today’s creatives can leverage AI to enhance their work. He points out that AI can be used to help generate initial ideas or iterations quickly, which can then be refined and perfected by human creatives. This synergy allows for more efficient workflows and opens up new possibilities in creative expression.

Emphasizing the collaborative nature of the technology, Hood points out that the best creative work involving AI is done by professionals who know how to use the tools effectively, explaining that AI-generated outputs often go through multiple iterations and refinements guided by human expertise. “The best work with these tools is being done by professional creators,” he says.

Practical examples of effective AI use

According to Hood, AI can’t replace the work of an experienced and talented human being. “If your Chief Strategy Officer created it, you’d probably be really upset. But if an intern produced it, you’d probably say, ‘Actually, that’s not a bad start,’” he says. This approach allows teams to build on a foundation provided by AI, saving time and enabling them to focus on higher-level creative tasks.

Watch now AI - Episode 3: Gen AI in practice - deep dive on deep fakes

Hood highlights the importance of domain expertise in getting the best results from AI tools. “I saw a prompt for an image which was a teenage model wearing a polo neck jumper, and I could have written the first 10 or 12 lines of the description of the image,” he says. “The next 10 or 15 lines was full of things that I wouldn’t even have thought of. The person had said that it needed to be shot on a specific kind of Hasselblad camera with a specific kind of Carl Zeiss lens and a specific kind of Rico filter. I wouldn’t have done that because I don’t know those things, because I’m not a photographer. That highlights where casual knowledge ends and professional knowledge begins.”

This level of detail in the image prompt showcases the depth of AI’s capability to incorporate professional knowledge into its outputs. “How you get the best out of these things is when you know what you’re aiming for and you know what the output needs to be,” says Hood. “And I think it’s that level of experience and expertise that comes from people who are creative professionals.”

Challenges and ethical considerations

The use of AI by creators comes with its set of challenges, both practical and ethical. Hood points out that there is a significant difference between a project that relies solely on AI for its output and one where AI is merely a tool in the creative process. “I think what AI does really, really well is it gets you off to a fast start,” Hood says. However, he warns against viewing the initial AI-generated outputs as the final product. The true value of AI, according to Hood, lies in its ability to augment human creativity, not replace it.

One practical challenge Hood mentions is the need for continuous learning and adaptation. Creatives must stay on top of the latest tools and integrate them into their workflows. “You don’t lose your job to the technology. You lose your job to the person who comes in, who can do it. You want to be that person,” he says. He encourages creatives to experiment with AI tools, find good guides and tutorials, and integrate these tools into their day-to-day work to become more fluent and confident in using them.

On the ethical front, Hood acknowledges the concerns surrounding the use of AI, particularly in terms of copyright and intellectual property. He cites the example of a graphic novel created with MidJourney, where the author’s copyright was initially granted but later challenged. “In my previous experiences, I’d never really particularly worked with a big legal team, but at WPP, there are several technological developments that have happened in the last few years where you need to have a focus on regulatory and ethical and standards,” Hood explains.

Read more Training models responsibly: AI and creator rights

To address these concerns, WPP has established a framework that ensures all AI use is ethical and compliant with regulations. This includes using models trained on copyright-free data and implementing robust guardrails to prevent misuse. “Having a clear vision of what you stand for, what your ethical position is, and having good people with a great understanding of what the regulatory systems are and what to look for is crucial,” he says.

WPP has created their own enterprise AI platform, called WPP Open. This is made up of a set of integrated tools that perform different functions: Creative Studio, Production Studio, Commerce Studio and others. The training of the AI models in this platform is done using what WPP calls ‘brains,’ developed by Satalia, to focus the AI on brands, products, audiences, and media platforms.

IBC Keynote preview

In his upcoming keynote at IBC, Hood plans to delve deeper into the theme of human-machine collaboration. He will discuss the initial scepticism and fear surrounding AI in creative industries and how these can be mitigated by viewing AI as a collaborative partner rather than a competitor. His talk will emphasise the importance of understanding and mastering AI tools to remain competitive in the creative field.

Hood will also share insights from WPP’s internal initiatives, such as the Creative Technology Apprenticeship program. This program identifies talented individuals who might not traditionally see themselves as part of the creative technology space and provides them with the training and tools to excel. This initiative underscores Hood’s belief in nurturing human talent to maximise the potential of technological advancements.

Moreover, Hood will touch upon the future of creative technology, highlighting ongoing projects at WPP where AI is integrated into bespoke creative and production tools. These tools are built on foundational models but are customised to meet specific needs, ensuring that WPP’s creative outputs are both innovative and ethically sound.

Hear more from Andy Hood and his thoughts on human-AI collaboration in the Keynote Fireside Chat: “At the intersection of AI and creativity: Why the future lies in Human-Machine Collaboration” at the IBC2024 Conference on Saturday, 14 September.

Register now for IBC2024