Biden, Musk and Trump star in this shockingly good AI parody

Earlier this month, a new movie trailer for a French film starring Elon Musk, Alexandria Ocasio-Cortez, President Joe Biden and former President Donald Trump was released on X, the platform formerly known as Twitter. Jack Dorsey plays a baguette-carrying baker, and Tim Cook also has a cameo as a ditzy trolley passenger.

This trailer is of course fake. The AI-generated parody, titled “La Baye Aréa” (The Bay Area), was created by a user named @trbdrk and quickly went viral, with many users commenting that they were initially fooled by the stunning AI effects .

AI-generated commercials and movie trailers are a popular draw for users tinkering with generative text-to-video technology. But compared to a year ago, the production quality has improved significantly in almost all aspects. This video is clearly a parody, but it wouldn’t be far-fetched to imagine users creating videos that recreate realistic circumstances, a skill that inspires new possibilities but could also have real-world consequences.

As a full-time creative who has been working with AI tools for a few years, I have learned how to detect artificial intelligence in the wild and what tools were likely used to create the content. Here’s how to make a video like “La Baye Aréa” and what to look out for when surfing the Internet.

AI tools to create a video like “La Baye Aréa”.

A video like La Baye Aréa could be produced using three generative AI tools that users can now access:

• Runway Gen-3: Runway generates videos from natural language and image prompts, and Gen-3 came out in June. Runway signed a deal with Lionsgate earlier this month.
• In between 6. Midjourney is a generative AI platform that converts natural language prompts into images. It can be helpful to first create a storyboard of images for a video and then pipe those images to something like Runway in a command prompt to better ensure you get the output you want.
• Udio: Unlike Midjourney (images) and Runway Gen-3 (video), Udio specializes in creating AI-generated music.

I was wondering how I would personally go about recreating a video like this. Here is my attempt to deconstruct the process:

1. Create a concept and storyboard with Midjourney.
2. Use Runway Gen-3 to convert these still images into video sequences.
3. Identify gaps and transitions in video content, then generate additional scenes with Runway Gen-3.
4. Use Udio to create the necessary soundtrack and sound effects.
5. Combine all the elements in a chosen video editing software, sync the sound with the video and produce the final product.

These steps may seem simple, but La Baye Aréa was probably more challenging. It’s also worth mentioning that the video doesn’t contain any dialogue, which would have increased the effort.

How to recognize AI-generated videos

Now that we’ve talked about the tools used to create the video and the possible production process, it’s time to train our eyes and ears to recognize AI-generated videos.

First, look for visual clues such as inconsistencies in facial expression, unnatural body movements, or artifacts and disturbances in the foreground, background, or transition. You can easily spot glitches when these characters blink their eyes or move their lips. By glitches I mean unnatural movements in the way they blink their eyes, or transitions between blinks, as well as in the way they move their body, which sometimes seem robotic and inconsistent. As AI video and audio improve over time, these glitches will become more subtle and difficult to detect.

Sometimes you can spot a mistake even if you can’t quite put it into words – this happens a lot. An article in the journal Human Movement Science found that natural human movements arise from a complex interplay of neuromotor control, biomechanics and adaptability. In layman’s terms, human movements are subtle and our eyes often see more than we can put into words.

Next, look for auditory clues such as mismatched lip sync, inconsistent background noise, or unnatural intonation in vocal patterns or accents. I find that intonation is often the most obvious clue to me. Intonation refers in particular to the rise and fall of pitch when speaking. It is a way of conveying meaning, such as distinguishing a question from a statement, and it exists in all spoken languages, not just English. Vocal patterns refer to a broader range of elements, which can include intonation, but also things like rhythm, pitch, and breathing pauses. Once you notice these obvious and sometimes subtle differences, you will easily recognize an AI-generated video.

Finally, there are contextual cues, the elements that go beyond auditory and visual inconsistencies. We often distinguish fake content based on our own experience, knowledge and logic. In La Baye Aréa the cast is absurd and clearly a parody. But also consider the age difference between a character and that person’s age in real life. One reason for this is that AI is often trained on photos or videos from years ago.

At some point, AI videos will become so good that we will need to check whether such actors have appeared in certain films and other productions. IMDB and Google Search are available for many of us to check, and it’s usually a good idea to gather information from multiple sources.

Why is it important to recognize AI-generated videos?

You may have heard of the term “responsible AI,” which refers to a set of principles that help guide the design, development, deployment, and use of AI. While “La Baye Aréa” was easily recognizable as an AI-generated video, the creator also clearly marked it as AI content. This recognition helps viewers avoid making assumptions and speculations.

But what if creators, organizations and political entities decide not to disclose AI content? It can be a real problem. As we have seen in recent years, undisclosed AI and fabricated content can lead to a loss of public trust, the spread of manipulative misinformation, and serious ethical and legal concerns.

The more we can educate ourselves about what AI is currently capable of, the better off we will be as a community of educated thinkers. Just as companies must practice responsible AI, we also have a responsibility to learn its capabilities and limitations.

Stay up to date on AI

The good news is that there are a variety of ways to stay connected and informed. Consider learning a few AI tools for your daily life. Use accessible, free tools like ChatGPT, Perplexity, Claude, and Google Gemini to ask questions and find answers. Additionally, if you’re a developer like me, learn what AI tools developers are already using.

I hope you find this article helpful as we continue to navigate the growing world of AI. Welcome to my YouTube channel if you want to learn more about these tools and services in the future.