AI is rapidly reshaping AV, and perhaps nowhere more so than in the field of content creation. Anna Mitchell gathers experts in the sector to give a snapshot of where we are now, and the future possibilities on offer.
Like most industries, AV stands on the cusp of a technological revolution as AI promises rapid advancements with the potential to transform every aspect of the sector and how professionals approach their work.
But separating a buzzword from those technologies and developments that can genuinely shape the future is no easy task. At one of end of the spectrum we have the promise of enhanced profitability, streamlined operations and unprecedented creativity. They’re possibilities that are so compelling they can engender strong emotions that range from excitement to fear. At the other, we’re seeing an abundance of products with AI claims that sometimes don’t stand up; terms such as AI-washing have quickly been added to our collective vocabularies. Among all these hopes and fears, delights and disappointments, are cautionary voices telling us that AI as a concept, having been first conceived in the 1950s, is navigating the inevitable peaks and troughs of any technology hype cycle. Has a fabled technology finally arrived, or are we on the cusp of another AI winter?
“I think we’re entering a new era,” says Julian Phillips, senior vice president and managing director of XTG, AVI-SPL’s global business unit focused on delivering experience technology solutions. “AI is going to inform pretty much everything we do, and I think we have to understand what we’re letting ourselves in for; the rate of change that's going to happen is now going to be exponential.”
Technologies have been created that have produced seismic changes in the past with Phillips citing the internet and smartphones as examples. But that change, he says, has been gradual.
“With AI things change every month, every week. In the AV industry we're pretty good at innovation and change but I'm not sure we're necessarily ready for the rate of change that AI is bringing.”
AI and AV connect on levels beyond their shared initial; according to Phillips they’re uniquely linked. “Most of the innovation in AI now is happening with convolutional neural networks and the ability to ingest huge amounts of video, audio and imagery to do these amazing things. To not only be able to process it and analyse it, but also to create it. So we're now getting into the generative, creative aspects of what AI does. We are a creative industry. We rely very heavily on rich content to drive most of the value that we provide our customers. And that's why I'm very excited about what AI can provide; it's going to be a catalyst for huge growth for the entire AV industry.”
A new era of creativity and efficiency
AI's role in content creation is one of the most exciting recent developments in the AV industry. Generative AI, in particular, is revolutionising how content is produced, offering endless possibilities for creating dynamic, real-time experiences.
In the past, content creation was a time-consuming and costly process, often requiring extensive manual effort. Today, AI-driven tools can generate content in real-time, drawing from diverse data sources to create a constantly changing visual and auditory landscape.
One company at the forefront of that is Dubai-based design studio Vivid Studios. The company is applying AI to projects and product development in myriad ways and Moe Tarakomyi, the company’s chief creative officer, is excited about the many possibilities on offer. He’s brimming with ideas - some realised, others in the pipeline – about how AI can transform his company’s offering and the wider industry, ranging from automation of operations and management to creating support tools where users can chat with virtual agents that can provide maintenance and support, technology control and troubleshooting.
Tarakomyi is unequivocal about whether AI is set to transform the sector: “We’re seeing the next revolution, the next big thing in AV.”
Quoting Mark Zuckerberg, he adds: “We will eventually have more AI agents than humans. That means that everything can be optimised. Everything that is repetitive, doesn’t require human creativity, or emotions will eventually be automated using an AI model.”
Tarakomyi says that translating the huge potential of AI into the specialised field of pro AV will need to be powered by a whole community. When it comes to the part Vivid Studios can play within that, he credits the work they’ve been able to do within Zenon restaurant in Dubai (reported on in Inavate July 2024) and other immersive spaces they’ve delivered. “These customers are giving us the sandbox to go out and play with these things. That only happened because they trust us,” he says.
When it came to Zenon, the aims of the restaurant aligned with Vivid Studio’s own ambitions to harness the huge power of AI resulting in a fruitful collaboration. Vivid Studios is continuing to not only further those developments, but also democratise them.
AI generated content in Zenon restaurant in Dubai created by AI using Vivid Studios’ latest workflows.
“What we’re trying to do now is create tools that allow anyone with any level of expertise to communicate with the machine and ask for what they want,” says Tarakomyi. “We’re building a very natural way for people to communicate with hugely complex systems so creative studios can use these tools without needing to go back and learn the basics of programming. To support this, we’ve partnered with Nilor Corp, combining their expertise in AI and technology development with our creative vision.
“Eventually what we want to offer to the pro AV world is a device that would be installed in your immersive space with a simple set up at the beginning: select number of screens, mask them, prepare it. Then you basically have a creative studio in your server rack, you use your phone to communicate with it and, using natural language, ask for the content to be generated.”
Taking the concept even further Vivid Studios is offering the control to diners in one Dubai venue so they have the ability to dictate content mapped on to objects and sculptures in the restaurant.
Alongside these developments Vivid Studios is continuing to develop its own brand of servers. “During the day these can act as a media server; as soon as a venue is closed, they become a CCG (creative content generator) and start to render content,” explains Tarakomyi. “It’s a very nice way to get the most out of your investment in hardware.”
The landscape is starting to move to a place where AI workflows can understand requirements to generate content for any canvass; from multi-screen set ups, to complex projection mapping and down to a single 16:9 screen. It goes without saying that that’s transformative for AV integrators but there’s more. Tarakomyi runs through a list of examples where Vivid Studios has used these workflows to generate content for social media advertising for its clients and even presentations to support its own sales activities. In one example, Vivid Studios created a three-minute walkthrough of a space that didn’t exist, wasn’t designed and they only had the concept for. It doesn’t take long to see how AI is poised to fuel diversification in the services an AV company can offer to increase revenue, as well as boosting its own sales and marketing efforts. Meanwhile productivity throughout the whole lifecycle of a project stands to gain; Vivid Studios has an AI assistant present in every meeting that can be interrogated in various ways to deliver everything from mood boards to meeting summaries.
Any technology that promises to automate, streamline and take on tasks that had traditionally been carried out by humans will be met with some resistance or fear.
“Artists with no background in generative AI hated what we’re proposing to start with,” confirms Tarakomyi. “What they used to do in three weeks, could be done in three hours. But when they found out it’s all the labour work and automating some of the boring tasks, as well as the fact you still need a human to stay in charge of the creativity and put things together, they started to find it a useful tool.”
Rather than competing with human creativity, Tarakomyi says AI can be “a superpower” to make more people more creative.
That’s almost exactly the same terms that Dutch artist Jeroen van der Most, who has been using AI in his work since 2010, uses to describe the impact of AI on his creative process. “AI is really about supercharging my work,” he says.
“AI is continually evolving, back then in 2010 it was machine learning and three or four years later we saw the start of what I would call creative AI with things like Google DeepDream, which was really cool but already now is considered retro. Now we see every couple of years AI taking new steps; we’re currently in this trajectory where AI increasingly enables everyone to create amazing imagery or amazing movies or music.”
Van der Most pinpoints 2022 as around the time when many of the technical barriers to using AI dropped away, with big changes bringing it to mass audiences in recent years. These changes, says van der Most, have allowed him to work more quickly and efficiently so he is able to create much bigger projects in timescales that would have previously been unfeasible.
He says however the power is a double-edged sword. “The real danger, and it’s probably already happened to a certain extent, is it leads to a massive flood of mediocre content.”
Pinpointing a single strand of AI that has changed his process more than any other van der Most says: “Many of my most well-known projects are based on large language models, so basically ChatGPT and its alternatives that are available now.”
‘It feels like Prada but is based on broccoli and potatoes,’ says Jeroen van der Most explaining his recent artwork that captures vegetables with a camera and uses AI to build images and movies around them in the style of a fashion campaign.
Letters from Nature for example is a work that uses large language models to write letters on behalf of nature to politicians about the climate crisis and it’s been exhibited in various forms, from large-scale projections onto buildings, down to printed paper.
One of the unique things about working with AI as a content generator is its ability to surprise… and not always in good ways. “Three or four years ago we had to build in all sorts of checks before you could start running a piece live to avoid inappropriate or potentially offensive content,” says van der Most. However, as the technology has matured van der Most says it’s now quite unusual for it to create something inappropriate, although it’s something he still considers. He is always at pains to be very clear that the content in his projects is created by AI and not the views or output of an artist or the organisation hosting a work.
One more technical aspect of showing an AI generated artwork live is the simple fact that the systems to run it must either be local (with all the considerations required regarding transport of equipment and technical set up), or will be cloud based and then stable, powerful internet connection becomes crucial. It is something that van der Most says he does consider when creating an artwork but that he’ll also leave freedom to be surprised, often finding the ideas of how to show a work from the content that is generated.
Balancing that ability for AI to surprise while maintaining a coherent artistic vision is something that van der Most considers in his work. An interesting part of this is how he describes combining different strands of AI; for example a large language model and an image generator component. This brings us right back to the conversation with Tarakomyi who talked about getting various AI models to communicate with each other to perform a wider task. There are large language models that offer natural ways to give commands, image generators to create content and other models offering checks and balances. And then there’s the specialist AV part: a model that understands the technical requirements of different screen configurations for example.
Both Tarakomyi and van der Most effectively show how these models come together to create something that is greater than the sum of its parts. The AV industry can harness the huge collective development that’s going into widely used models, but as Tarakomyi advocates, it still needs to come together as a community to deliver specialist AV systems.
As AI continues to evolve, its potential in the AV industry seems limitless. Experts like Phillips believe that we are only scratching the surface of what AI can achieve.
“AI will make us more personally productive but will also be a driver of business opportunity,” he says, moving on to demonstrate how generative AI content can achieve this. “The AV industry wouldn’t exist if it wasn’t for the fact that we’re here to help people tell their story and engage their customers and their employees with amazing, immersive, rich content.
“The problem however has been that the creation process is time consuming, costly and requires a huge amount of talent. If you go to the lobbies of major enterprise organisations you see a big videowall and you might see the same content that was playing six months previously due to the cost and time involved in creating new content.
“In the next year you are going to see AI generation tools that are not just editing amazing stock footage, but it’s actually generating the video content itself to an amazing quality that is going to change the trajectory of the AV industry. If we have more content at a lower cost, that is more personalised and customised, we’re going to deploy way more AV than we’ve ever done in the past.”
The integration of AI into the AV industry is not just a trend, it's a transformation. We’ve focused on content creation but across resource management, meetings, system design, programming and beyond, AI is driving innovation, improving efficiency, and opening up new possibilities for AV professionals. As we look to the future, it's clear that those who embrace AI will be at the forefront of this exciting new era.
Van der Most questions whether we’ll even be talking about AI in years to come. “At some point AI will be everywhere, all software will be AI and then the term becomes meaningless.”
Top image credit: Amanita Silvicora\Shutterstock.com