AI is about to become real dangerous.

A couple of weeks ago, we had fake AI-generated images of Donald Trump getting arrested and the Pope with more style than us.

Now we have AI-generated podcasts that are too real for comfort. YouTube channel The Joe Rogan AI Experience released fake interviews with Sam Altman (OpenAI’s founder) and Donald Trump.

All content was generated using AI, and here’s why it’s scary:

• The voice is almost indistinguishable from their real voices.
• It even incorporates their mannerisms, using “um” and “like.”

Disrupting content creation

AI has already disrupted how images are created. And voice is next. Several podcast tools have released voice-cloning features — which can be used for post-production editing. Then finally, video — where we’re already seeing early progress.

But the dangers are clear.

• Misinformation — it’ll become more and more difficult to tell what’s real, especially as AI improves.
• Security — where voice and facial recognition can be used to access devices.

Celebrity deepfakes are already being used in ads without permission — and it hasn’t even been 5 months since ChatGPT was released.

Dive in: AI is Growing Too Fast, And Some Technologists Want Big Tech to Chill.