Food delivery apps are treading with caution amidst the ongoing discussions around Grok 2’s realistic and unfiltered image-generation feature, which may be used to mislead people.
Zomato CEO Deepinder Goyal recently announced a ban on AI-generated images of restaurant menus to build consumer trust in the platform. The initiative was met with a positive response from consumers online.
“This is great but you must already know that restaurants invest heavily in post-production for product photos which are far from the real pictures of actual products. Photos have always been on the borderline of being creative and deceptive to lure customers into ordering,” said Kumar Gaurav, an engineer and foodie.
Swiggy Does it Differently
On the other hand, Zomato’s competitor, Swiggy, has a different approach when it comes to embracing AI. The food delivery giant has a more nuanced strategy on AI-generated imagery.
Swiggy launched a free AI-fuelled photoshoot feature that helps restaurants click realistic images of the food items resembling a professional photoshoot. The app also equips one to change the background and enhance the quality of the images making them appear aesthetically appealing.
This hybrid model combines the benefits of AI-driven enhancements with the assurance of real-world representation.
Nag Bhushan, the owner of Bakeway Bakery in Hyderabad, said, “Swiggy has introduced a valuable feature that assists me in adding an image to menu items that don’t have one. I have added around 50 images to my menu, and it was all free. Earlier, I had to spend a significant amount on photoshoots. Thanks to this new feature, I can save a lot of money while still adding images to increase orders.”
Similarly, Swiggy’s ‘WhatTo Eat’ feature, which is also powered by AI, allows users to explore options based on their mood and cravings.
Data from Grubhub reveals that restaurants with pictures for their menu items receive at least 70% more orders and 65% higher sales than those without them. AI food image generators can be a valuable asset for small and emerging restaurants struggling to create professional food photos for delivery app listings.
However, using AI for such platforms is often questioned. Rahul Birendra Singh, a theatre actor, also raised doubts about the credibility of studio-generated images, shedding light on how photoshopped images can look different from real images, misleading the consumer.
Hybrid Model is the Best Bet
While the technology promised a visual revolution, the potential for deception and misinformation cast a long shadow. To regain customer confidence, food delivery platforms are encouraging authentic food photos and providing assistance to restaurants.
AI can be a powerful tool for manipulation, but it can also be used to enhance the quality and consistency of food images. When used responsibly, AI can help to improve lighting, colour, and composition without compromising on authenticity.
While striking a balance between visually appealing and truthful food images is ongoing, combining AI enhancements with real-life photography could be the answer while keeping production cost at minimal like Swiggy’s AI-fuelled photoshoot feature.
Ultimately, the industry’s success depends on trust, and genuine food images are essential to build trust.
AI Images are Not Ideal
Meanwhile, a new research unveiled during the Global Nutrition and Hydration Week 2024 has revealed a concerning trend: AI-generated food images often exaggerate portion sizes and calorie content. For instance, AI can artificially inflate the amount of fries in a picture or add extra whipped cream to a dessert.
There’s a growing worry that widespread exposure to these idealised food images could promote unhealthy eating habits by normalising excessive portions.
This increasing reliance on AI-generated imagery has also sparked concerns about transparency in the food industry. As consumers demand authenticity, the use of AI-altered food photos raises questions about ethics, truthful advertising and accurate product representation.
However, Zomato is being cautious. The company is encouraging restaurants to use real food photography. Restaurant owners without professional food images can request a photo shoot at cost with no profit to Zomato through its support team.
This can help with providing an authentic picture of the food while adhering to aesthetics for attracting consumers to the platform but production cost on the restaurant’s part still remains a challenge in the scenario.
Not Just Pictures
Images are not the only place where food delivery apps are experimenting with AI. Zomato and Swiggy also use generative AI for consumer interaction on the platform. Zomato is heavily focused on developing generative AI technology.
Earlier in an interview with AIM, Zomato had confirmed that the company is heavily investing in AI/ML, data science and analytics to give a customised experience for its consumers and partners.
Furthermore, they have appointed Vaibhav Bhutani to lead Zomato’s generative AI initiatives. His team is exploring various applications, including a system that uses multiple AI agents to recommend food choices and improve user satisfaction.
For now, Zomato cannot completely do away with AI.
