Imagine scrolling through TikTok and stumbling upon a video that transports you to the trenches of World War I or the bustling streets of Pompeii just hours before disaster strikes.
It’s a fascinating way to experience history—at least, that’s what hundreds of thousands of people following AI-generated history-themed accounts believe.
Two of the most popular accounts, Time Traveller Pov and POV Lab, boast nearly a million followers combined, offering visually immersive but historically questionable recreations of the past.
These clips, often going viral, have sparked excitement among viewers but also concern among historians who worry about the spread of misinformation.
Questionable Accuracy: Historians Speak Out
While these videos captivate audiences, experts argue that they often get key details wrong.
A widely shared clip from Time Traveller Pov offers a supposed tour of an Allied trench in 1917, showing soldiers reading letters from home and standing at an observation post.
However, military historian Robin Schaefer dismissed the depiction as completely inaccurate, criticizing the unrealistic uniforms and overall inauthentic portrayal.
Similarly, an AI-generated reenactment of daily life in Pompeii before Mount Vesuvius erupted in 79 AD has raised eyebrows.
Archaeologist Dr. Hannah Platts pointed out multiple errors, from incorrect types of bread loaves to modern-looking wine glasses.
She emphasized the importance of handling historical representation responsibly, warning that AI should not be used to distort history.
The Entertainment vs. Education Debate
Daniel, the 26-year-old creator behind Time Traveller Pov, defends his work, claiming his videos are meant to be cinematic and engaging rather than strict documentaries.
He argues that sparking curiosity about history is a win, even if the details aren’t perfect.
“All my content is labeled as AI-generated, so people know it’s not real,” he explains.
“I add comical elements to remind viewers not to take it too seriously.”
However, historians remain unconvinced.
Professor Barbara Keys from Durham University criticized an AI-generated video depicting the Chernobyl disaster in 1986, stating that the portrayal was inaccurate, even down to the look of the power plant.
She expressed concerns that AI fails to provide proper sources, making it difficult to verify the authenticity of its content.
AI and Historical Representation: A Distorted Reality?
Other videos have also drawn scrutiny.
A clip showing life during the Black Death presents sick individuals looking oddly well-groomed, while another reenacting the Hiroshima bombing in 1945 offers a first-person perspective of someone witnessing the explosion before fleeing the city.
Experts argue that these portrayals lack historical depth and often resemble video game aesthetics rather than well-researched reconstructions.
Robin Schaefer strongly criticized the AI-generated World War I trench video, calling it nonsensical and devoid of historical accuracy.
He went further to state that these AI videos hold no real value, arguing that their lack of stylistic consistency and factual grounding makes them unreliable representations of the past.
The Risks of AI Rewriting History
One of the most criticized videos depicts the morning before Pompeii’s destruction.
It shows people enjoying a luxurious breakfast, high-status men dining at a grand table, and citizens wearing what appears to be modern clothing.
Dr. Platts pointed out that the video ignores well-documented archaeological evidence, such as actual carbonized bread loaves found in Pompeii.
She warned that such AI-generated content risks rewriting history in a way that is not properly researched, leading to an oversimplified, dramatized, and sometimes completely inaccurate version of the past.
Where Do We Go from Here?
AI-generated history videos are undoubtedly entertaining, but their growing popularity raises critical questions about the balance between engagement and accuracy.
While creators defend them as a fun way to spark interest in history, experts caution against their potential to spread misinformation.
As AI continues to evolve, the debate will likely intensify: Should there be stricter guidelines for historical AI content?
And more importantly, how do we ensure that future generations don’t mistake AI-generated fiction for historical fact?