The deepfake era. Part 1, I guess.
Deepfakes are getting horror movie scary 👻. From faking your voice and even your tone of voice, to faking your face and your expressions.
Something we've been seeing a lot of lately are AI clones, not just for video, but for audio too. Without going too far, Spotify and Youtube are dubbing creators' voices so that their content can be translated into any language, trying to make it look like the creator is speaking in whatever language.
I have mixed feelings about this. On the one hand, I see the potential for creators to use this feature to get their content out to a wider audience. I also see the potential for non-native English-speaking creators to be able to make English content without fear of being bullied by Tiktok trolls... But on the other hand, I also see how this could mean that smaller creators who have their niche making whatever specific content in their native language might become irrelevant because mainstream creators are now able to make content in any language. Or maybe it doesn't matter, maybe you don't want to follow mainstream content. Time will tell.
While Spotify's new feature is incredibly good (listen for yourself at this link), Youtube's is still a long way from being hyper-realistic. But let's not get carried away, Spotify has only released 3 episodes from 3 creators, including Lex Fridman, so it's not like we can all just listen to every podcast available.
Anyway, Lex Fridman has not only had his podcast dubbed into several languages thanks to AI, he has also recorded a new podcast in the metaverse. Yes, his 3rd interview with Mark Zuckerberg wasn't recorded in Lex's studio, but in Mark's virtual world via those new photo-realistic Kodak avatars we saw a couple of years ago.
