Back in the day when you asked kids what they aspired to be in the future, you would usually hear answers like fireman, doctor or singer. Nowadays they also want to be a YouTuber or influencer. In the recent years the job ‘influencer’ has evolved and taken over our social media feeds, especially in the fashion industry. Imagine being paid for trips abroad, taking pictures of your daily life or attending parties and meeting celebrities. Fashion influencers have become a vital part in the marketing strategy of many small and large fashion brands. Companies often offer influencers to advertise their products and pay a fee per post.
I also follow some fashion influencers on Instagram, mainly to gain some outfit inspirations. One day a post from @seraphinnevallora showed up on my feed and I realized it was an account with all AI generated pictures. Upon further research, I realized this was not the only ‘influencer’ account using generative AI. Another example is @feli.airt with over 320K followers. All these accounts have a similar aesthetic and models, perhaps due to them using the same generator. I tried using the same generative AI called Midjourney with Feng My Shui. You just need to write a prompt which is a text phrase that the AI will read to generate an image. The results are not exactly what I wanted – I noticed that you need to write very specifically and choosing the write words is key. Also, I used a free version and the paid version of Midjourney would probably generate better results. Personally, lightning and color proved to be especially challenging. According to the creators of @seraphinnevallora, one picture can take up hours as they need to adjust to the customers’ preferences. Apart from fashions posts, they also receive offers from real estate businesses to show off their buildings. This development lead me to think models and influencers may not be needed anymore as AI can also create pictures for marketing purposes. Remarkably, a common factor between the pictures is that almost no pictures show the lower body part. This was also the case for the generated pictures from Feng My Shui, even if I typed it into the prompt. The technology is still developing and I am curious to see how it will improve. Below you can find the pictures from various accounts on Instagram and the ones I generated. Can you tell which one is made with a free generator and paid version of Midjourney?
Interesting blogpost! The development of digital technology in the influencer industry, particularly in fashion, is indeed fascinating. I decided to check out the Instagram accounts, and I must say the images are stunning. I’ve never came across these types of Instagram, but it is surprising to see how many followers they have! The use of generative AI opens new avenues for creating content. Nonetheless, I don’t believe they will replace models and influencers, as these AI-influencers’ lack the ability to engage with their followers. This personal connection is often a significant part of what makes influencer marketing so effective.
To answer your question: I think the pictures in the top-left and bottom-right positions were created using the free generator because these images don’t display the lower part of the body (as you mentioned in your blogpost), unlike the other two picture where more of the body is visible. Furthermore, in my opinion, the free generated images are simpler comparted to the other two. I’m curious to hear about the right answer!
Funny that you think the bottom left is not made with the free generator. The top-right is indeed taken from an Instagram account, but the bottom-left is also one that I made. I intentionally tried replicating the bottom-right picture to see if I could make anything comparable, but if you look closely the hands are not very detailed