“Digital Doppelganger: AI Manipulated Hitler Speech Creates a Buzz on X”

“An AI-Altered Hitler Speech Is Going Viral On X”

“To me, it seems clear that this type of deep AI phenomena—whether it’s recreations of well-known figures or the digital resurrection of loved ones—needs guidelines, not just on the technical side, but also ethically, legally, and societally.”

Oh sure, open the Pandora’s Box of AI, replicate Adolf Hitler’s voice – what could possibly go wrong? Meet Candide, the avant-garde AI lab that decided to recreate Hitler’s controversial speech using deep learning, and stirred quite the hornet’s nest. While the potential use and benefits of AI technology cannot be understated, there is an equally loud and important discussion required on the ethical implications of such exploits.

Apparently there’s a lot of grey area when it comes to AI ethics. Should there be a consent clause from the person being digitally resurrected? Is it okay to profit off of the ‘artificial’ voice of a deceased individual? Oh, and let’s not forget the crucial part of this uncanny technology: Can the ‘artificial’ altered voices be used in misinformation or disinformation campaigns? The world may never know!

But wait, there’s more! The AI-algorithm used by Candide is called Tacotron 2, and apparently it’s able to mimic any voice. That means you could technically have any historic figure, celebrity, or frankly, anyone, narrating your bedtime stories – as long as you can feed the AI with sufficient audio samples. Does that sound empowering or terrifying? Depends on how it’s used, right?

Now let’s boy-scout our way back to the main hitch. The guidelines (if there are any) barely cover the surface of this technological wild-west. It’s like trying to contain a river with a sieve. However, discussions and raising awareness around the gaffes and perils of AI exploitation is a start. For instance, Facebook’s Vice President of Artificial Intelligence and other scholars are already calling for a Dickensian “Ghost in the Machine” legislation to ensure ethical applications of AI.

One can only hope that we create a robust ethical outline for the herculean task of AI application before a deep learning algorithm recreates Don Quixote to spearhead a revolution. Stepping on the landmine of ethical complications, Candide’s stunt has only succeeded in creating more questions than answers.

So if you’re questioning whether AI is a saviour or a frankenstein, well, the jury’s still out on that one. Jaw-dropping technological advancements are all the rage, but whether they are ethically and morally concrete is a different ball game entirely. Let’s hope we don’t need to learn everything the hard way.

Read the original article here: https://www.wired.com/story/ai-altered-hitler-speech-viral/