Why did I do it? Why did I pay $8 for 50 different abstract, impressionistic, mind-blowing portraits of myself? Am I such a narcissist? Are we all?
Our Instagram and other social feeds are currently inundated with these high-resolution interpretations of our faces and various laymen, also known as Magical Avatars. They all come courtesy of the Lensa app, an AI image generation engine that creates some wild and often fantasy-driven interpretations of the images we feed it.
There have been other similar platforms like Reface that let you put your face in famous movie clips. Everyone was busy for a while. I turned myself into Tom Cruise (opens in new tab).
The app was free (with a lot of ads, if I remember correctly) and eventually we all decided it might not be such a good idea to upload our images to a random app developer.
Lensa fills that space, but with a product that you almost have to pay for, first with an expensive subscription price and then more casually with a pay-per-AI image batch offering.
What’s shocking to me is just the sheer number of people paying to have their images digested and then spit out as eye-catching AI artwork. The FOMO is so strong here that everyone (even this now embarrassed reporter) succumbs. My adult son spent nearly $15 for a set of images of his girlfriend.
I’m not proud
Why did I do it? Two reasons. I got tired of seeing other people’s Lensa Magic avatars on my feed and I really needed something new for Instagram too. I’m kind of like this about my social feeds, I always try to keep the pipe full for reasons I can barely explain here.
Adding my AI images to Instagram’s growing legion of majestic Lensa-generated portraits wasn’t difficult. The app is available for free on iOS and Android. Opening it will give you a $49.99 subscription offer that you can “cancel at any time.” Personally, I hate these kinds of apps, apps that tease you with an incredibly cool feature but demand exorbitant cash payments upfront (I might be cheap, too).
Like many other apps, the sub-offer is kind of a front. If you ignore it, the app immediately dumps you into the pay-per-play section, where you can purchase anywhere from 50 ($7.99) to 100 ($14.99) AI-generated Magic Avatars without committing to anything. to subscribe.
Now I can’t remember the last time I paid $8 for the privilege of using a single app feature, but that was the siren song of those beautiful portraits. I rationalized the cost in my head: “Come on, you’d pay that much for two Krispy Kreme donuts and a water,” something I bought last week.
After that, the app asks you to upload between 10 and 20 images of yourself in different poses and with different expressions. I decided to peruse my iPhone’s “Selfies” folder, where I found a bunch of suitable photos. I uploaded a batch, but Lensa rejected some of them as unusable (she didn’t say why), so I selected a few more. In hindsight, I should have chosen far fewer profile pictures and images of me making funny faces. As with most AI imaging systems, you get what you put into Lensa.
Overall, I’m happy with the results (you can see some examples above), but I felt a little uncomfortable and a little cheated. I should have asked some questions, many in fact, before uploading my photos.
A day after Lensa delivered 50 4K AI images of mine (they stay in the app, but you can download your favorites in standard or 4K resolution), I sent Lensa developer Prisma Labs a list of questions:
- Do you keep photos uploaded to you? If so, are they encrypted?
- When did you launch the Magic Avatar section?
- How exactly is AI used to generate the images?
- How many people have uploaded their images?
- Why preload the subscription fee when people can only pay $7.99 for 50 images?
- How do you address concerns about people uploading images without their consent?
- Have you heard of some being used to generate adult content?
In some ways, you could take one or more of these questions and apply them to a virtual AI image-generating system. While all of these AI systems are exciting, they feel like black boxes. For some, we just add text prompts, but we still don’t understand how they generate art.
For example, the heavy lifting behind Lensa’s Magic Avatars is handled by the free, open-source image generation platform Stable Diffusion – one of many accused of appropriating artists’ work (opens in new tab) to train its AI model.
But that’s just one of my concerns. Another is what happens to the images we love to provide to it? For example, how do we know that images we upload are not used to further train the AI? I don’t see an option to waive that option.
There are also no prompts warning against uploading someone’s image without permission. I’ve heard reports of some people using the system to generate pornographic images of unsuspecting people (opens in new tab) without their permission.
Where do our original images go when Lensa’s work is done? With no information promising that Lensa will delete them, I have to assume they keep them all. We can only hope that they are at least encrypted.
Prisma Labs responds
I do have good news. While Prisma Labs didn’t include much of what I consider necessary information in its app and didn’t immediately answer my questions, it did point me to a extensive FAQs (opens in new tab) that solves some of these concerns.
When asked what Lensa does with our images, Prisma Labs writes that they are removed from Prisma Labs’ servers once processing is complete. However, the FAQ also contains this language:
“We keep avatars for as long as necessary to provide the service to our users. Please note that we are currently working on a new feature that will allow users to permanently delete their avatars from our servers.”
In other words, your original photos are gone, but those images Lensa took? Prisma Labs still has them. By the way, there is no mention of data encryption.
The FAQ does not address image ownership considerations: the image you upload must be yours or an image to which you have access rights. It does deal with “sexualized” images, but mostly the AI’s tendency to sexualize female characters. More importantly, Prisma Labs adds, “Intentionally using the app to create explicit content is absolutely prohibited by our Terms of Use and requires Lensa users to comply with that requirement.”
It’s not clear how Prisma Labs enforces that particular rule and, even less helpfully, Prisma Labs adds this comment: “Unfortunately, all of these efforts have not yet made AI absolutely safe from biased content and explicit imagery. Therefore, we argue that the product is not intended for use by minors and warn users of the potential content risks.”
I can’t rightfully tell you not to try the app, especially after I went ahead and did it. Before doing so, however, make sure you’re comfortable with Prisma Lab’s answers (and lack of answers) to any of these questions.