Marketers beware: AI-generated art might be fun, but it presents serious ethical risks
Every technology is a double-edged sword; what matters is how humans choose to wield it. Here are some takeaway lessons from the recent controversies surrounding Lensa AI, an app which creates digital portraits using artificial intelligence.
AI-generated portraits of the author, created using Lensa AI. / Webb Wright
If you’re even remotely active on Instagram or Twitter, you’ve probably been seeing some exceptionally flattering digital selfies popping up on your feed over the past week or so.
Created by an app called Lensa AI, these images (and many others like them) signal a radical new vision for the future of artistic expression; some might call it beautiful, others dystopian. Regardless of how you feel about AI-generated artwork, one thing appears to be virtually certain: AI is advancing rapidly, and it’s poised to revolutionize just about everything — including marketing.
Rise of the (intelligent) machines
Computers have a long track record of outperforming humans in domains which were once thought to be uniquely and inviolably … well, human. Chess, for example, was once believed by many to be a uniquely human pursuit, requiring not only the rote memorization of a combination of moves but also an ability to plan ahead, spontaneously adapt to an opponent’s moves and occasionally bluff or take leaps of faith. Then IBM’s Deep Blue — an AI program — beat grandmaster Gary Kasparov in 1997, forever dispelling the notion that humans had some kind of monopoly over chess. The same thing happened in 2015 when Google’s AlphaGo program beat a human in the strategic board game Go.
Fine, so AI became the undisputed masters of board games. But surely, many people told themselves, machines could never exceed human talent in the realm of the arts. True art, it was widely believed, must be imbued with some kind of intangible human spark — welling up from somewhere deep in the emotional brain — which can never be imitated by a machine. Not very well, anyway.
As it turns out, that might also be wishful thinking.
In recent years, AI-generated artwork has become very, very impressive; so impressive that you probably couldn’t tell the difference between an AI-generated painting created in half an hour, say, and a painting that took a gifted human painter months to create; so impressive that an AI-generated painting — a breathtaking piece which looks like something out of one of Neal Stephenson’s most vivid dreams — recently won first place in an annual art competition hosted by the Colorado State Fair. After it won, the human artist who submitted the piece told The New York Times: “Art is dead, dude. It’s over. A.I. won. Humans lost.”
Regrettably, not even writers appear to be safe from the impending AI revolution. If you need proof, just Google “AI Harry Potter chapter” and prepare to be amazed (and laugh quite a bit).
AI is advancing rapidly, and marketers would be well-advised to pay close attention. Soon, AI-generated avatars — like those created by Lensa AI — could replace human actors in many ads. And who knows, there may come a day in the not-so-distant future when intelligent machines are able to come up with better ideas for marketing campaigns than their human counterparts. Given the impressive feats that AI has been able to accomplish in the short time it’s been around, it doesn’t seem all that unlikely.
Don’t worry, marketers don’t need to worry quite yet about machines creeping up and stealing their jobs. At the same time, it should be acknowledged that AI is here to stay. The time has come, therefore, for marketers to begin thinking about how this emergent technology may soon begin to impact their creative strategies.
Lessons learned from the AI-generated selfie trend
To that end, it can be useful to look a bit more closely at the intriguing case of Lensa AI.
Launched in 2018, the app truly burst onto the cultural scene in fairly phenomenal fashion late last month, with the launch of “Magic avatars,” a feature which uses a minimum of 10 user-submitted selfies (which can be either images or videos) — and a text-to-image model called Stable Diffusion — to create fantastical, AI-generated depictions of the user, with aesthetics that range from fairy tale, to sci-fi and virtually everything in between. Lensa AI was reportedly the most popular app in the “Photo & video” category on Apple’s App Store last week. And according to data from Statista, the app jumped from around 342,000 downloads in October to more than 5.8mn downloads thus far in December.
Sounds harmless enough. In recent days, however, a number of ethical concerns have been raised about the app. For one thing, it appears to be somewhat biased towards a conventional — and in some cases sexualized — notion of physical beauty: Some users have reported that the AI-generated avatars that they received through the app were scantily clad, unrealistically thin and so forth. Tech Crunch also reported last week that “it’s possible — and way too easy — to use [Lensa AI] to generate non-consensual soft porn” using images of real people.
Artists (human artists, that is) have also been voicing concerns about the ethics of using Lansa AI. Some claim that the app uses artistic styles that are essentially plagiarized from flesh-and-blood artists, while others worry about the effects that the app’s ability to produce such a vast amount of digital art for such a cheap price (it creates 200 avatars for just $7.99) will have upon their job security.
Suggested newsletters for you
Where do we go next?
These are still early days for AI. As the technology advances, ethical questions are going to continue to pile up. (Is it morally permissible to replace all human cashiers with intelligent machines? Should autonomous vehicles be programmed to drive itself off a cliff to avoid hitting pedestrians? Should we continue to enhance the abilities of AI-powered art platforms like Lensa AI even if it means that people might have nonconsensual nude portraits of them created and distributed online?)
AI poses some enormous opportunities to marketers, but it also presents certain dangers. As with any powerful new technology, marketers ought to keep a close eye on developments in the AI space, but they should be wary about jumping onto sudden trends — such as AI-generated selfies — without thinking through and carefully analyzing the long-term implications and risks.
As we’re presently seeing in the crypto space, not all hot technological trends are destined for eternal glory. Sometimes, it’s best to sit back and observe how things play out before you jump on board.
Lensa AI did not reply to a request for comment.
For more on the latest happening in tech, sign up for The Drum’s Inside the Metaverse weekly newsletter here.