.jpeg)
AI is revolutionising the food and beverage (F&B) industry, echoing a trend seen in other sectors globally. Initially, many F&B companies were reluctant to adopt AI, cautious of associating their brand image with something artificial. However, this perception is shifting as companies discover the innovative and valuable opportunities AI provides. Leading brands like Beck's, Heinz, Coca-Cola, and Bodyarmor are already experimenting with AI in various capacities, incorporating it into marketing campaigns and product development.
But it's not just the major players that can benefit from AI. With the right guidance and knowledge, F&B companies of all sizes can harness the power of AI. Food images are an area where AI excels. Early research has found that individuals for the most part prefer AI-generated food images over real food images. Generated images enhance the appeal of depicted foods by leveraging key features such as symmetry, shape, glossiness, and overall lighting and colour. Subtle tweaks in the positioning of foods enhance their appeal – for instance, avoiding positioning food items directly towards the viewer. AI-generated foods tend to look more energy-dense and topping-rich, appealing to our evolutionary wiring¹.
The use of generated food images makes many business owners uneasy. Hesitance to use AI is often rooted in the idea that it's just... not real. There's a truth to learn from that. AI only delivers value to a business when the images created are an accurate representation of the products served. Accuracy is the foundation of any successful AI integration. Always keep in mind, that AI is a tool not an objective.
An easy way to illustrate both the current limitations and potential of AI-generated images is by starting a hypothetical business. Imagine you are opening a burger restaurant and need some impactful promotional material. You decide to explore AI images and you start, of course, generating an image of a burger.
The image above is completely AI-generated, and while this image is detailed and nearly perfect, by adjusting the prompt, you can modify every aspect of the image, including the background, viewing angle, lighting, image style, texture, and much more. For this example, you are happy with the style and look of the burger, but you know that your burger is served with tomato only on top of the patty. AI tools can now also modify specific regions of a picture without having to regenerate the entire picture, so after adjusting, you get an image that more closely matches your actual product. This flexibility allows you to save time and money.
Using this example as a jumping off point, imagine the countless applications this technology has within F&B. Whether it's hyper-realistic product shots, creative and fantastical imagery, virtual food styling, appetizing menu presentations, food previews, dynamic menu updates, eye-catching food photography, or recipe visualization, the possibilities continue to expand.
As impressive as this range of technologies is, F&B photography is in no danger of becoming obsolete. Quality, engaging photography remains vital to telling a brand story. However, photographers are already adapting and thriving by incorporating new technologies into their work. This enables their clients to create stunning ads that rival those of industry giants.
Say you want to create promotional material for your pizza that appeals to a younger audience, possibly through a cartoon-like aesthetic. Here, you might want to generate a picture similar to the one below:
What if you're about to unveil a revolutionary new taco creation? You want to capture people's attention and build excitement. This could be accomplished with one of the pictures below.
Now, a more challenging task: recreating an unrealistic advertisement, like the Jalapeno Burger advertisement from McDonald's (click here to see the original). AI image generators struggle to perfectly replicate highly specific or stylised ads. This is because AI creates from a library of training data – images in this case. It performs better with common images, such as standard burgers, because there are so many of them in its database.
To achieve the desired image, careful prompt engineering and high-quality source images are essential. However, in this example, the result falls short of replicating the original McDonald's ad. This highlights the current limitations of image generation technology, which may improve soon.
But for now, AI will still allow you to create eye-catching advertisements that take inspiration from the best practices of the industry. Below is a great example of how AI can recreate an image that would otherwise be very hard to replicate.
AI image generation is set to revolutionise food advertising by making it more accessible and cost-effective. While traditional food and beverage photography will remain important, AI offers new creative possibilities that level the playing field for businesses of all sizes. Resistance or fear towards this new technology is understandable. However, companies are gradually recognizing the immense potential that AI brings, and it won't be long before AI is an industry mainstay.
We believe that this technology will expand the industry, allowing more companies to create better advertisements. At Conceptional, we integrate AI into our processes for increased speed and flexibility. We can advise you on best practices and approaches towards AI. If you're interested in incorporating AI into your F&B strategy, contact us. Let's shape the future of food together.
'AI-generated food images look tastier than real ones'
Publication Date: 15 March 2024
Website: AI-generated food images look tastier than real ones | University of Oxford
Author: Giovanbattista Califano
Co-Author: Professor Charles Spence
Published in: Food Quality and Preference
.png)
