How to make videos with artificial intelligence

For this video, I used ChatGPT Plus for the script, Midjourney for the images and Canva Pro for editing. The tutorial shows the interfaces available when I made it; some buttons and features may have changed.

Open video or resource

History of Artificial Intelligence for Teenagers

Writing the script with ChatGPT

First, choose the topic, platform and audience. I wanted to explain the history of artificial intelligence to teenagers in a video for Youtube Shorts. With those choices made, I wrote the following prompt in ChatGPT:

Prompt asking Chat GPT to write a story about artificial intelligence for a Youtube Short video.

With just three lines of instructions, ChatGPT generated a first draft of the script. It felt like magic! Before using it, check the historical facts and adjust the text to the video’s length.

Creating prompts with ChatGPT to create images with Midjourney

To prepare the images, I asked ChatGPT to write the prompts I would later use in Midjourney.

Instructions asking ChatGPT to prepare Midjourney prompts

First I have indicated the number of prompts necessary to create the video. Below, I have indicated the structure. In this case, since it is a history video, I have indicated that the style be hyper-realistic, photo realism and cinematography.

I also asked for the prompts in English, which I chose for this workflow. I then specified that each prompt should be a single sentence with its parts clearly separated. This was a formatting choice to organise the instructions, not a guarantee of better results.

Next, I asked it to add the aspect ratio at the end of each prompt. For YouTube Shorts, I chose 9:16, a vertical format suited to phones.

Finally, I pasted in the story ChatGPT had generated. Although I had requested English prompts, it returned nine in Spanish, so I asked it to translate them.

Image generation with Midjourney

To access Midjourney, I used Discord. In a channel where the bot was available, I entered /imagine and pasted the prompt prepared with ChatGPT.

A Midjourney prompt and four images of scientists in the 1950s

Midjourney generates a grid of four images. The numbers run from left to right: 1 and 2 on the top row, 3 and 4 on the bottom. The U buttons select an image from the grid; their effect on resolution depends on the model version.

The fourth image was closest to what I wanted, but I wanted to explore some changes. I clicked V4 to generate four variations.

A Midjourney prompt and four images of scientists in the 1950s

I chose the first variation and clicked U1 to view it separately. I then opened the image and saved it using the browser’s options.

Midjourney displays an image of two scientists in the 1950s

From this view, you can explore the available options for modifying or expanding the image. The arrows extend the scene in a chosen direction, such as to the left. Midjourney offers several alternatives; if none fits, you can generate another set.

4 images created by Midjourney that have created possible scenarios of the image above.

In my case, I like image U1. And this is how it looks big.

Midjourney image of four scientists in the 1950s discussing artificial intelligence

This is the process that I have followed with all the images that appear in the video.

Video editing with Canva Pro

I selected the mobile video format and uploaded the images downloaded from Midjourney to canva. I then added one page for each image.

Canva Pro editing for videos in mobile format

I then placed each line of the script over the corresponding image and adjusted its size and position. I used the Burst animation to add movement to the text and adjusted its speed to keep it readable.

Adding Burst animation to text in images

Once we have repeated this process with all the images, we must adjust their time depending on the text we have placed so that the viewer can read the text correctly.

Next, I have animated the images and added transitions between them in order to transmit movement. Finally, I searched through Canva for a melody to accompany the images. To adjust the melody to the video, I have used the Beat Sync function.

Adjusting music in Canva with Beat Sync

I chose Canva because I find it intuitive, but you can use whichever video editor you feel most comfortable with.

Now, we have to enjoy the video that we have generated through Chat GPT, Midjourney and Canva Pro. I leave you a summary of the tutorial in video format that lasts less than a minute.

Open video or resource

Video summary of How to make videos using artificial intelligence?

More artificial intelligence

Saturdays.AI makes AI more accessible through shared learning and projects with a social purpose. Our motto is AI4ALL IS AI4GOOD.

If like us you like technology, become a premium member of Saturdays and you will be able to take advantage of all the benefits of a leading community in Artificial Intelligence: https://saturdays.ai/become-a-member

On our website you have free online courses and content to help you, as well as a community of passionate people like you, we are waiting for you!