The world of video and AI: Tools and impact

AI is everywhere, more and more everywhere and we have to admit it, the technological advances are huge. No need to deny it or to cry scandal, it's a fact! And obviously our sector is also affected.

Not so long ago, we wrote an article on this blog about ChatGPT ( here ) and the place of the human in the midst of all these AIs in the fields of creativity.

Today, we're back to talk more specifically about video because yes, AI will revolutionize both video creation and editing.

And there is no way amap communication can miss this turn, whether we like it or not!

Instead, we prefer to get hands-on and test these tools, understand as much as possible about how they work and quickly establish how useful they can be.

It is more than likely that we will end up implementing AI in our processes, so why not start looking into it right away?

To paint a picture of what AI can do in the video, we're going to show you a few tools.

Keep in mind that we are still in the early days of AI in video and these tools are not perfect, far from it. Sometimes there are bugs, inconsistencies, surprising effects that make you smile.

But still, the possibilities are vast, the advances are fast and if ChatGPT has taught us one thing, it is that AI learns fast.

AI tools for video

‍

1) Runway

To start with, here is a website that allows you to edit directly on a page, without installing anything. That alone is pretty cool.

An "AI magic tools" section offers you tools with, as its name indicates, AI. Tools that, for example, allow you to remove a subject in the shot. We import a rush, we pass a brushstroke on it and hop, the AI will calculate frame after frame and make disappear throughout our video the element that we have chosen

‍

‍

A new version of its "Text to Video" tool could generate video sequences in a few minutes. Of course, for the time being, these technologies remain very limited.

‍

2) Chatgpt

The famous one! This virtual assistant with AI, incredibly powerful and precise, we have already told you about it, but we would like to come back to it.

You have already understood that with ChatGPT, the more we refine our searches the more precise they will be.

But did you know that it was also possible to copy and paste an article and ask ChatGPT to summarize it in a few sentences and, why not, by switching it from English to French?

Did you know that you can ask ChatGPT for video ideas or even ask for a script? Of course it won't be perfect. But instead of resetting if its answer doesn't satisfy you, you can correct it, specify your request further, dig deeper. So, with a bit of test and learn, ChatGPT can become a real asset to help us create our videos.

‍

3) Adobe Podcast

In Beta version, Adobe Podcast now offers AI!

The tool is simple, you had a microphone problem and your audio is atrocious? Adobe Podcast's AI greatly improves the quality of your audio and makes it usable without any worries.

Magical!

‍

4) Fliki

‍

Fliki is a text to speech and video converter that helps you create content quickly and easily using synthesized voices. Presented like this, we must admit that it is amazing!

With Fliki, you can convert your blog posts or any other textual content into video, podcasts or audiobooks with voiceover.

Of course, the solution is quickly paid and with credits that are quickly gone.
Royalty-free images necessarily, ditto for the music that are quickly limited.

And your videos have no soul, let's not fool ourselves! But this tool is still amazing and its existence forces us to always question ourselves.

‍

5) D-ID

Another funny tool. This one allows with AI to create a realistic avatar (or not) and to make it talk, to make for example presentation videos... with a virtual presenter.

A paying service with often... surprising results!

‍

‍

Conclusion for AI and video?

By the fall of 2022, tech giants Meta and Google had both announced they were working on their own video generation AIs.

For Google, on the AI side, the Bard project is in development and you can register now on a queue to access what will be a conversational tool, a bit like ChatGPT, but trained on all web data indexed by Google (which is not nothing...)

As we know, the video sector is also being explored by the current top champion of generative AI, OpenAI, as revealed by its leader Sam Altman. But despite the speculations, its latest model GPT-4 is not capable of such creations.

So we find ourselves with more and more tools that sweep up more and more possibilities in the video industry.

What can be frightening is when these tools blur the boundaries between the role of the tool and that of the author. When the tool is no longer just used to execute an idea.

In these moments, we have two choices. We can worry about such autonomy, or we can rejoice in the invention of all these new methods of creation that will, inevitably, end up being democratized.

On the movie side, we have Steven Spielberg, who has said he feels "very uncomfortable" with AI creations. To quote him, "The soul is unimaginable and inexpressible. It cannot be created by any algorithm."

And that's what we're getting at. Yes, tools that can save time and even money are more and more numerous and efficient. Yes, they will be democratized and probably less expensive than other solutions.

But it is up to us to take these tools in hand and implement them in our processes, to free up time to put our creativity, our humanity, our "soul" where we can make all the difference.

And for the time being, and perhaps forever, AI, no matter how facilitating it may be, will never be able to surpass us on "that little something extra" that only humans can provide.

‍

‍

‍