Artificial Creativity
Most of the anxiety about AI and misinformation points at the deepfake: the flawless synthetic forgery that fools everyone. Eduard-Claudiu Gross studies something cruder, more common, and arguably more corrosive, the obvious fake that was never meant to fool you in the first place.
Eduard Gross is a communication scholar and researcher at Lucian Blaga University of Sibiu in Romania, with a PhD from Babeș-Bolyai University. His work centres on disinformation, deepfakes, and how people actually interact with generative AI. In our recent conversation on the LiveInnovation channel, he laid out why the real threat of AI abundance isn’t a single convincing lie, it’s the slow erosion of our willingness to believe anything at all. And, in a turn that connects directly to my own research, why that same flood of synthetic content is quietly raising the value of imperfect, human-made work.
What Are Cheap Fakes?
Cheap fakes are crude, low-quality, obviously manipulated images or videos — the opposite of sophisticated AI deepfakes. Their purpose is not to make you believe the fake, but to make you distrust everything, including authentic content.
Gross’s clearest example is the video that circulated at the start of the Russian-Ukrainian war, showing President Volodymyr Zelenskyy telling his troops to surrender. It was so poorly made that almost nobody believed it. That, he argues, is the point. The creators of a cheap fake know it won’t convince discerning viewers. What it does instead is pollute the information ecosystem.
“The purpose of cheap fakes is not to make you believe them, but it’s to make you more skeptical about true content you are being exposed at.”
The mechanism has a name: the liar’s dividend. Once feeds are saturated with crude fabrications, people grow fatigued and paranoid. When genuine, documented footage of a real event appears, they wave it off as “probably fake too.” The lie doesn’t win by being believed. It wins by making the truth unfalsifiable.
Why Debunking Often Backfires
The instinctive response to a flood of fakes is to fact-check them. Gross’s research complicates that reflex.
In a survey of over 400 consumers of debunked content, he found that around 80% already held at least a bachelor’s degree. Fact-checking, in other words, largely preaches to the choir. It’s consumed by the highly educated who were unlikely to be fooled in the first place, while rarely reaching the people who actually fall for disinformation.
There’s a subtler problem, too. Continually issuing formal debunks for absurd claims keeps those claims alive in the media cycle and adds a layer of legitimacy to them. Gross’s counterintuitive advice is that some false narratives are better left alone — let the subject “die alone” rather than granting it another round of coverage. This runs directly against the mainstream assumption that more fact-checking is always the cure.
Behind all of it sits a stubborn piece of human psychology: people resist facts that threaten their sense of who they are. As Gross puts it, they “try to keep their vision of life intact” and dislike contradicting their own beliefs. He speaks from experience — he began this research after noticing his own early arrogance, trying to force-feed data to relatives during the pandemic and watching it change nobody’s mind.
AI Slop and the Normalization of the Fake
Cheap fakes are one symptom of a larger condition Gross calls AI slop: the continuous stream of low-quality, AI-generated content flooding every platform.
The examples are already familiar — the synthetic photo of Pope Francis in a white puffer jacket, the “All Eyes on Rafah” graphic, AI-generated war imagery quietly appearing on stock photography sites. Individually, each is minor. Cumulatively, Gross warns, they train us into a dangerous passivity: people gradually accept AI slop as just another kind of content and stop critically assessing what they see. The deeper risk is historical. Image generators hand bad actors the power to manufacture photographic “evidence” for any fabricated event, and synthetic images are already slipping into the archival databases journalists and historians rely on.
The Abundance Paradox: Synthetic Perfection Raises the Value of Human Imperfection
Here is where Gross’s argument turns, and where it meets the central thesis of my own work on creative abundance.
As AI makes flawless text, images, and audio cheap and ubiquitous, synthetic perfection stops being impressive. It becomes wallpaper. And in reaction, audiences are drawn back toward the things AI can’t convincingly fake: raw human imperfection, physical labour, tactile craft. Gross points to the surge of public engagement around deliberately basic human activities — knitting a sweater, throwing pottery, being visibly imperfect at a craft.
“It enhances the value of human creativity, and it’s something I noticed in the engagement of people doing basic stuff such as knitting a sweater or making pottery or simply being imperfect at their crafts.”
The flaws are the proof of authenticity. And underneath the craft sits something machines fundamentally lack — lived experience. Gross puts it plainly: “Melancholy is a really important aspect in creativity and our feelings; you cannot write about something if you are not living it.”
He tells one story that captures the whole tension. He discovered a song, enjoyed it thoroughly, and then learned it was AI-generated — and immediately began to dislike it. As a scholar who studies exactly this bias, he still fell for it: “I became a victim of my own critique.” Knowing the machinery didn’t inoculate him against the reaction. The moment authorship was revealed, the emotional bond broke.
Media Literacy Alone Won’t Save Us
If there’s a sobering counterweight in the conversation, it’s Gross’s realism about solutions. In Romania, he describes media literacy efforts as fragmented, underfunded, and undermined by weak teacher training. During fieldwork, researchers trying to teach educators to spot fake news found history teachers who sincerely believed in ancient aliens. When the people meant to teach critical thinking are themselves susceptible to conspiracy, “just add media literacy” stops sounding like a plan.
What This Means for Creators and Communicators
Several practical principles fall out of the conversation:
Stop over-debunking. Don’t hand fringe, ridiculous claims additional screen time. Some dead-end stories are best left to die rather than kept alive with another formal rebuttal.
Lean into imperfection and physicality. Rather than competing with AI’s polished output, emphasise your human voice, your visible process, and your flaws. In an age of synthetic perfection, those are the features, not the bugs.
Verify historical images. Journalists and anyone communicating about the past need strict checks on archival imagery, because synthetic stock photos are increasingly contaminating the record.
Watch the Full Conversation
Gross’s perspective is bracing precisely because it refuses both easy panic and easy optimism. The danger of AI abundance isn’t only the convincing lie, it’s the exhaustion that makes us stop trusting the truth. But the same flood that erodes trust is also sending people back toward what is unmistakably, imperfectly human.
The full interview goes deeper into cheap fakes, the failures of debunking, and the psychology of why we believe what we believe. Watch it on the LiveInnovation YouTube channel.
