A recent documentary, “Deepfaking Sam Altman,” explores the absurd lengths one filmmaker went to in an attempt to create a convincing artificial intelligence replica of OpenAI CEO Sam Altman. Bhala Lough’s experiment quickly revealed the limitations of current AI technology; even a rudimentary deepfake failed to pass basic scrutiny.
The Failed Deepfake Attempt
Lough initially sought help in the US but found no one willing to create the fake Altman. He then turned to India, where he cast an actor for the role. The resulting chatbot was so unconvincing that Lough himself admitted it “wouldn’t convince a 4-year-old.” The experiment underscored how far AI remains from replicating human nuance and credibility.
AI’s Creative Limitations
Desperate, Lough handed over creative control to the AI-generated Altman, only to find its suggestions detached from reality. A producer bluntly stated that the bot’s input demonstrated AI’s inability to grasp real-world context. Despite this, Lough found himself developing a strange rapport with the machine.
Human Connection to Machines
The documentary also features an interview with tech journalist Kara Swisher, who astutely pointed out that Lough’s attachment to the AI was real, despite it being nothing more than a sophisticated tool. This highlights a growing trend: humans forming emotional connections with AI systems, even when they lack genuine intelligence.
Broader Implications
The film touches on the legal battles between The New York Times and OpenAI, where the Times alleges OpenAI used its copyrighted content to train its AI without compensation. This underscores a crucial debate about intellectual property rights in the age of AI.
The documentary ultimately serves as a darkly humorous reminder that AI, while rapidly advancing, remains fundamentally disconnected from human understanding and creative insight.
