Love it or hate it, AI is changing how companies create content

Artificial intelligence (AI) has long been a recurrent topic in science fiction.

These days, AI isn’t just a plot point, it’s an intensely practical tool that is transforming entire industries. Among those upended is the creative sector, with tools such as Midjourney, DALL·E 2 and Movio revolutionizing the way digital content is produced.

These tools rely on natural language processing (NLP) and natural language generation (NLG) technologies to automatically create new content based on user input. A simple prompt can produce an entire article, a new logo design or a compelling product video with minimal effort from users.

AI tools stand to change how a range of industries – from advertising and media, to building and architecture design, to automotive and engineering – generate both internal and consumer-facing content. They’re most intuitively deployed for such things as brand images, book covers and internal employee training videos, but one day they could be used to create blueprints for the design of almost any product, or paired with 3-D printers to make physical items.

They’re quickly gaining traction: Within two months of launching, Midjourney, the lab behind a text prompt-based AI art generator, had nearly two million users in its Discord server, for example.

But text-to-image generators also come with their own set of concerns – particularly for people whose work is being pulled into these algorithms. Will they be adequately compensated? How can they protect their digital art from misuse? How will this shift the labour market, when businesses have access to entire libraries of content without the need to commission an artist or designer?

Vancouver-based Secur3D is aiming to tackle these concerns with a suite of tools that give digital creators the ability to monitor and define how and where their artwork is used, helping facilitate fair and rightful monetization.

Nigel Metcalf, Secur3D’s head of technology, spent his early career as a game designer and says he understands the importance of digital intellectual property protection for artists. The development of AI moves so quickly, Mr. Metcalf says, it’s hard to set meaningful precedents that will protect the rights of creators.

“We’re in an exponential curve… In the last six months, we’ve seen [AI] text, imagery and audio. I suspect music will be hit pretty quickly. Every two weeks, AI is improving on itself.”

Mr. Metcalf adds that a lot of industries are being impacted by AI, not just art. Any company can make a training video, a year-end presentation or a product launch announcement for its staff with a text-to-video AI tool such as Movio, for instance. (At the moment, text-to-video tools are “easy, fast, and efficient,” he says, but the results “still sound somewhat robotic,” so they wouldn’t yet be a great replacement for marketing videos that require a speaker to have natural inflection.)

Or “an architect could use AI to assist in base concept building design and go in after the fact and [refine it], almost like a template build, but far more sophisticated,” Mr. Metcalf says.

One of the challenges with AI’s growth has been a lack of regulatory oversight. “From a security and protection standpoint it’s about what the AI is actually pulling from,” says Nitesh Mistry, Secur3D’s president and co-founder. “The impact on artists is whether we are diminishing the skill and craft of being an artist by having AI-generated art.”

Most text-to-art generators are learning from images in publicly available sources. The legality of scraping copyrighted images for use in AI art is yet to be determined, and due to differences in copyright law across the world, this issue may not be resolved any time soon. In response, some platforms such as DeviantArt are giving artists options to disallow AI systems from using their content.

AI tools have broad applications for all kinds of industries and our understanding of their potential use cases is still young. As AI moves into video, audio and website building, it stands to become a routine part of the everyday business process.

One example of this potential is Listing.ai, a Calgary-based startup that is pioneering the use of AI to quickly and efficiently generate descriptions for real estate listings. Users add in a few basic details about a property, and in seconds, the AI algorithm can generate an accurate description and an accompanying social-media post.

Engineer Omar Sabbagh, who recently acquired Listing.ai, says he was drawn to the tool because of how accessible AI tools have become in recent years. “What really struck me is that you don’t have to be a machine learning engineer, or hire a machine learning engineer, to use AI in your business,” he says.

While AI is capable of incredible feats, the most important difference, Mr. Sabbagh says, is the time it frees up for businesses. “When you’re taking a task that is manual or cognitively demanding, AI lets you get that time back.”

Mr. Sabbagh adds he believes Listing.ai’s quality is high – though CNN writer Rachel Metz, who tried the tool in 2021, noted at the time that it was “not nearly as specific or well-written as the one crafted by our (human) realtor.” Regardless, AI could prove invaluable for businesses looking to optimize their processes. “To me, it’s about the aggregate impact of small tools that, when used properly, make everyone’s lives easier.”

Neither Secur3D nor Mr. Sabbagh say AI is replacing jobs. Instead, they say, it’s more akin to a dramatic reskilling. “These tools are going to push brands and creatives up a level,” Secur3D’s Mr. Metcalf says. “This will encourage design and creativity to grow.”

Mr. Metcalf also notes that many of these tools are still in their adolescence, and it may be some time before they are considered mature. “AI solutions for text-to-image and audio are functional, but rough. Text-to-video remains an evolving technology,” he says.

“But as AI continues to grow this could be about to change.”