It’s only natural for someone like Mark Zuckerberg to have some strong opinions about anything and everything that Elon Musk might be involved in. After all, Musk owns X, which is the direct competitor of Threads. Many would assume that Mark Zuckerberg wouldn’t think very highly of Grok, an AI chatbot released by Musk’s xAI, and it seemed like the tech CEO had released a video criticizing it. In spite of the fact that this is the case, this video was a super realistic deepfake, and it’s actually quite easy to see it for what it is.
In this video, Zuckerberg appears to state that Grok is substandard and that his own Llama 2 AI is superior in every single way, shape or form. With all of that having been said and now out of the way, it is important to note that there are grammatical errors and other obvious tells that make it extremely clear what is going on.
Whatever the case may be, the realism of this deepfake has many wondering if people will be able to tell the difference. You would only be looking for errors if you know what to look for in the first place, and some might not be tech savvy enough to figure these things out for themselves.
It bears mentioning that this video was released by a startup called Argil, and they managed to go viral with this stunt. According to one of the startups co-founders Brivael Le Pogam, the clip was meant to be an obvious joke, but it’s still lead to some questioning the ethics of such a practice with all things having been considered and taken into account.
There is a chance that such deep threats could end up becoming a massive threat to national security. Argil itself has created a deepfake of Barack Obama in the recent past as part of its marketing campaign, and it wouldn’t be a stretch to imagine that ones of Donald Trump or Joe Biden might pop up too. In an election year, such things would be much more harmful than might have been the case otherwise.
Image: DIW-Aigen
Read next: Facebook is Getting Overtaken by AI Spam
In this video, Zuckerberg appears to state that Grok is substandard and that his own Llama 2 AI is superior in every single way, shape or form. With all of that having been said and now out of the way, it is important to note that there are grammatical errors and other obvious tells that make it extremely clear what is going on.
These AI deepfakes will blow your mind. They are now so realistic that it's impossible to tell they're AI-generated.
— AshutoshShrivastava (@ai_for_success) March 18, 2024
Examples of Mark Zuckerberg, Marc Andreessen, Obama in 🧵 below.pic.twitter.com/PuBQ4CL2vG
Whatever the case may be, the realism of this deepfake has many wondering if people will be able to tell the difference. You would only be looking for errors if you know what to look for in the first place, and some might not be tech savvy enough to figure these things out for themselves.
It bears mentioning that this video was released by a startup called Argil, and they managed to go viral with this stunt. According to one of the startups co-founders Brivael Le Pogam, the clip was meant to be an obvious joke, but it’s still lead to some questioning the ethics of such a practice with all things having been considered and taken into account.
There is a chance that such deep threats could end up becoming a massive threat to national security. Argil itself has created a deepfake of Barack Obama in the recent past as part of its marketing campaign, and it wouldn’t be a stretch to imagine that ones of Donald Trump or Joe Biden might pop up too. In an election year, such things would be much more harmful than might have been the case otherwise.
Read next: Facebook is Getting Overtaken by AI Spam
