As you sip your morning coffee or sink your teeth into your lunch sandwich, check out the video I did using HeyGen for the visuals and a locally created voice clone leveraging RVC Voice2Voice. The rather famous voice is free to download from „the Internet“. I know it’s nowhere near perfect, but what we hear and see is already enough to prove my 4 points:
- It’s all locally feasible – today: I am able to create fake content on my own gaming PC with an NVIDIA GTX1070 leveraging generative AI without any external platform. And for those who say „Yes, but you used HeyGen!“: Convenience – I could have created the moving lips using my locally installed version of Automatic 1111 with the „SadTalker“ plugin, but I was a little concerned about the negative talk surrounding the extension (supposed viruses in the extension, etc.)
- Open Source catches up: Audio generation using tortoise TTS, RVC and other audio tools like the „Ultimate Vocal Remover“ give powerful tools completely for free into the hands of the masses. This will level the playing field in generative AI in every part of the content ecosystem. If this pace is kept up, I am doubtful commercial offerings will stay relevant in the long term.
- Society has to be educated or else… …we will look back at Cambridge Analytical and judge this scandal as a mere breeze compared to the generative AI storm we’re in.
- Do not underestimate current tech and execution: The shortcomings of the video mentioned are temporal, just as wrongly drawn hands were the laughingstock 4 months ago. You’ll be surprised how fast this video will not be distinguishable from the real one.
What’s the take away?
Keep trying, testing, and getting your feet wet. Do not solely rely on folks (especially here on LinkedIn) who want to sell you dreams they support by stuff they read somewhere else. Get your hands dirty yourself, and if you cannot afford this, get your employees to do just that. Like organize a PromptAthon, for example, like we are doing next month; invite your colleagues to participate, especially those who say they do not know anything about „This strange generative AI stuff.“ Those are the ones we have to reach before they get sucked up in fear-mongering about AI and their future.
Technology is not a force of nature; it’s what we make of it. It’s in our hands.
What did you learn from the developments in this space so far?