The fake AI video of Elon Musk, which was used to solicit investment is a classic example of bad use of artificial intelligence (AI) video technology by bad actors. Sadly, this has also raised concerns about what can be done to harm people’s reputation.
Lately, I’ve been impressed by the same technology if put into good use. One great example in this regard has been shared by the founder of the LinkedIN platform, Reid Hoffman, who has also been an investor in OpenAI (creator of ChatGPT). He has been experimenting with positive uses of AI. At some point he co-wrote a book with the assistance of AI.
However, what has caught my attention is the fact that Hoffman recently created a version of himself as an AI avatar. The coolest thing about this experiment is that Reid was interviewed by an AI version of himself.
I must say that it did a good job based on its understand of the person it was interviewing. It was also interesting to note the AI has Reid’s mannerisms to an extent, that while watching the interview it was difficult to distinguish between the real Reid Hoffman and the AI Reid Hoffman.
What is more important about this experiment by Reid is the fact that it shows us the positive aspects of fake AI videos. It shows that we can create a version of ourselves as AIs. In my view this matters for many reasons. In my line of work, I can foresee how useful it would be for me to assign my own AI avatar to interview someone on my behalf when I need to take care of something else. I can also see how useful it would be for me to assign my AI avatar to some meetings that I cannot attend. The possibilities are endless.
My sense is that AI videos can enable us to achieve more than what we’ve imagined. I also believe that our own versions of AI can enable us to counter fake AI versions that could be created by other people. I think we are at a point where we need to start creating AI avatars and versions of ourselves as a countermeasure against fake versions created by others.
We no longer have a choice in this regard. During the social media era, one way of discouraging fake social media accounts was to create one of your own. When you create an AI version of yourself and mark it as your AI avatar, this will be a form of your identity protection.
Here’s how Reid created an AI version of himself: the video avatar was generated by Hour One. The voice cloning was done by 11ElevenLabs.The persona or the way Reid AI formulates responses was generated from a custom chatbot built on GPT-4 that draws from his own books, speeches, articles, and podcasts that he had produced over the past 20 years. This may also include older statements and thinking shared by Reid online.
I challenge you to create your own AI avatar. I will do the same and share results in this column by the end of May. As many workers around the world are preparing to celebrate Workers day on May 1, it’s time to find ways of working together with AI.
Based on what I’ve seen, it can assist us to become more productive at work. If we succeed with this experiment, we may begin a process of turning around the bad use of AI to good use for our own benefit. Please try it and let me know online via X and on LinkedIN.
Wesley Diphoko has been working at the intersection of technology and media as the editor-in-chief of FastCompany (SA) magazine. You can follow him on X via: @WesleyDiphoko
BUSINESS REPORT