AI-Generated “Ancient Surveillance” Videos Blur the Line Between Fact and Fiction
A Chinese commentary describes a trend of using text-to-video tools to depict historical events as mock surveillance footage, including scenes from the Hongmen Banquet and Wu Song fighting a tiger. Some videos use captions suggesting that authentic recordings had been discovered, while reposters may omit labels identifying the footage as AI-generated. The commentary also cites AI face-swap scams involving relatives or friends and an unauthorized livestream using a public figure’s likeness to sell products, but the supplied source does not independently verify those cases.
The report says synthetic videos can circulate without AI labels and be mistaken for real recordings. It also warns that face-swapped content can be used for impersonation and fraud, making labeling and verification important when AI-generated media is shared.
最近,小雷刷到一条AI玩梗不能混淆视听的热搜,看完内容感觉这世界越来越魔幻了。
先是一条标题叫《公元前206年,鸿门宴现场监控流出》的视频。点进去是黑白画面,右上角跳着时间码,镜头对着宴会厅,项羽刘邦对坐饮酒,项庄舞剑、樊哙闯帐,一举一动自然得不像话。
图源:微博截图
古代还有监控?本以为这就是博主的创意整活,结果越刷越离谱。玄武门之变有监控实拍、武松打虎有现场录像、连杨贵妃在华清池的日常,都拍得清清楚楚。
不得不说,现在AI创作的门槛已经低到离谱了。以前想整个活,得会剪辑、会P图、找半天素材,才能出一个作品。现在输入一句话,AI马上给你生成一条视频,要啥风格有啥风格。