The Next Frontier: Digital Clones Speaking Alien Languages and posing Deepfake Threat
A TECH billionaire has unveiled a hyper-realistic AI clone that can speak every language – including alien ones – and even picks at its virtual nose. The resemblance is uncanny – can you spot the difference?
Unveiling the Reid AI Clone
Billionaire Reid Hoffman, who co-founded LinkedIn, built a custom GPT trained on 20 years worth of his books, speeches, interviews, and podcasts.
With all of this data, the AI model is able to mimic Hoffman’s speech patterns and physical mannerisms, such as hand gestures and looking upward while ‘thinking’.
Innovative AI Applications
The clone, called Reid AI, uses both speech and video deepfake technology – the same kind that can be manipulated by hackers.
Reid AI is able to condense a 333-page book into a single sentence for four different types of people: the smartest person in the world, a five-year-old, Star Trek fans, and Seinfeld fans.
AI’s Transformative Potential
Hoffman has likened the dawn of AI to Europe’s industrial revolution beginning in 1760, calling it the “steam engine of the mind”.
The grand hope for AI is for it to relieve tough, and potentially life-threatening, duties off humans.
Digital Twins and Enhanced Capabilities
For Hoffman, a digital twin means he can take a step back from some of his responsibilities and have some fun.
“I’m not above being lazy, having you do work instead of me means that I can go and have fun and do other things,” he tells Reid AI in the interview.
Future Possibilities and Concerns
These new abilities, Hoffman speaks of, include being able to communicate in Klingon – as well as other languages – and write poetry.
It also means the businessman can offload tasks onto his AI clone that he “wouldn’t have had the time for”.
Deepfake Risks and Ethical Considerations
Hoffman partnered with generative AI video company Hour One and AI audio startup 11ElevenLabs to build Reid AI.
11ElevenLabs is reportedly the product that was picked up by scammers to create deepfake robocalls of US President Joe Biden earlier this year.
The Path Ahead
Fears over deepfake have escalated this year, as 2024 is considered a record year for leadership elections worldwide.
Hoffman skirts around the obvious ethical concerns surrounding deepfakes and digital twins in his interview.
Final Thoughts
He reckons that, despite fears, ‘we’ should continue strides in developing the eerie technology – and the regulations around it.
“It’s very important that we set the rules of the road with digital twins,” he said to Reid AI, adding that “in this case it’s easy because we’re doing this collaboratively.”