Reid Hoffman cloned himself with AI — here's what he learned from his deepfake twin

Reid Hoffman, co-founder of LinkedIn and a prominent figure in Silicon Valley investing, has ventured into the realm of artificial intelligence by creating a digital clone of himself. This innovative project, unveiled during a recent episode of the 'American Optimist' podcast, utilizes a custom GPT model, voice synthesis from ElevenLabs, and video generation tools. Hoffman aims to demonstrate that technology often labeled as 'deepfake' can be repurposed for constructive and beneficial uses. In his exploration, Hoffman introduced 'Reid AI,' a virtual version of himself capable of mimicking his speech and even communicating in languages he doesn't speak. Remarkably, during a speech at the University of Perugia in Italy, his AI counterpart was able to relay the same message in Hindi, Chinese, Japanese, and Italian. "I've never heard my voice speaking Hindi or Japanese," Hoffman remarked, expressing his amazement at the technology's capabilities. While Hoffman considers his work to be a form of experimental exploration, he acknowledges the potential risks of deepfake technology, particularly regarding misuse. He refrains from releasing his AI twin into the public sphere due to concerns about hackers potentially manipulating it to spread misinformation. The negative implications of deepfakes are not lost on him; previous instances include AI-generated impersonations of public figures, such as former President Joe Biden, that have circulated misleading messages. For the time being, Hoffman's AI remains a private tool, primarily serving as a means for translation and testing innovative communication methods across different cultures. He emphasizes the importance of transparency in his project, ensuring that its synthetic nature is always disclosed and its applications are carefully monitored. Hoffman's conclusion is a thought-provoking one: not all deepfake technology is detrimental. With ethical design and intentional implementation, it could serve as a bridge for connection across linguistic and cultural divides. "Our goal as techno-optimists is to shape the technology in a way that allows for positive applications," he stated. Hoffman’s initiative is not an isolated case; others have also started to experiment with AI personas. For instance, Steven Bartlett, host of 'The Diary of a CEO,' recently launched a podcast featuring an AI-generated version of himself narrating stories about influential business leaders.