The Future of Instagram: What AI Can Teach Us About Visual Content
There’s a new generation of artificial intelligence making sense of our online data and it’s actually pretty useful for everyday tasks.
If you’ve been following our blog we have heartily espoused the value of visual content — after all, nobody is maximizing the potential visual content better than Instagram, and their 800 million daily active users means there are a lot of eyes looking at those images. Imagine what you could do if you could pore through all that content and see what those 800 million users see every day. The ability to identify trends in content and popularity right as they’re happening would be an incredible boon if you’re even marginally interested in maximizing your social media presence.
However, trying to make sense of the millions of images uploaded everyday is a mammoth task. There are a number of social listening tools and platforms that currently use technology to make sense of conversations (in text format) on Twitter, Facebook and others. What happens to all this rich data we can find on Instagram? This is where the AI comes in and is genuinely solving challenges.
In this exercise Pulsar Platform tested a next-gen AI module that has been trained only on foodstuffs. This type of AI—as it trained on a specific subject—is called a Vertical AI. Through deep learning, it has become brilliant at extracting food ingredients from images.
Okay, so this scenario isn’t life or death, but Vegan cooking can be challenging, especially if you’re not an aspiring chef. Not only that, the potential of this kind of technology in the future is what we’re really interested in. We have said before that keeping up with social media trends is going to be one of your keys to master social media, and this looks like an amazing way to do that. For now, let’s take a look at what the technology can do with food:
AI Case Study: Creating The Ultimate Vegan Dish from Online Visual Content
For this exercise they set out to use the power of Vertical AI to run an analysis on over 15,000 vegan dishes to identify the most common ingredients used. From this information, they identified the ultimate vegan 3-course meal recipe.
To begin, they set up search using Pulsar’s social listening tool named TRAC to pull in publicly posted Instagram posts, published globally over a course of a handful of days, which were captioned as #VeganFood or #VeganRecipes. This allowed them to analyze nearly 15,000 individual posts, worldwide.
Traditionally, if they ran this analysis through a general image recognition AI module, it would produce a plethora of image tags which are not associated with the search, such as Plate, Glass, Elephant, Man, Fruit.
However, as they chose the new Vertical AI image recognition module which has been trained solely to recognize foodstuffs, they get a much more accurate analysis of the image and what it contains.
Ready, steady, cook
The AI module picked up over 200 different ingredients and dishes. After hand-picking and eliminating a few false positives (some vegan foods really look similar to their non-vegan counterparts), they selected the top 20 ingredients. The winners are:
To everyone’s dismay, Avocado didn’t top the chart, and surprisingly Chocolate was used in 36% of the dishes. Now, using this set of ingredients, they came up with a 3-course vegan feast to delight the most hardcore of vegan foodies:
- Bruschetta starter: Tomatoes, garlic, basil and olive oil topped on a slice of crunchy bread
- Fresh Gnocchi: Potatoes for the Gnocchi, roasted corn, caramelised onions and pepper with chopped broccoli stems and sprig of basil and parsley on top
- Avo Chocolate Smooth dessert: Chocolate, Avocado, Ice and Bananas for a healthy smoothie
They had to leave out Apple and Rice in this menu, but we think that through the power of Instagram data and new Vertical AI, we have a winning Vegan meal here.
So, what does this all boil down to for you? What does this mean for your social media experience?
We know that online conversations have become increasingly visual at a staggering rate. Both brands and users are going to become more and more inclined to engage, understand, and participate in the conversation. But how do you do that when it’s so difficult to analyze exactly what makes visual content stimulating for users? That’s what this new tool can possibly do for us in the future—tell us what conversations are happening, and the context behind them. That’s the key to any kind of social media success: being able to get in on the conversation.