Putting Google’s Veo 3 Video Creation Tool To The Test

Lessons learned and potential unlocked with a game-changing tool

Aug 18, 2025

Directing AI video creation is like conducting a symphony—intentional, imaginative, and mindful of every detail coming together. Image made in ChatGPT.

By Mica T. Mulloy, M.Ed

Google’s Veo 3 AI video creation tool could go down as one of the major developments that significantly moved the Artificial Intelligence conversation forward.

There have been a handful of these moments since 2022. Obviously, the original ChatGPT drop was the first. ChatGPT4 was another. Deep thinking/reasoning tools were an impressive glimpse into the future of Large Language Models.

Now Veo 3 is making its capabilities known in everything from Instagram Reels to primetime commercials. There have been other video tools around, but this is a game-changer.

I wrote about my dual terrified and excited reaction to seeing Veo 3 videos: The Gospel According to AI: New Video Tools Take Huge Step Forward

I got limited access to Veo 3 through Gemini and my Google AI Pro account, and I was anxious to dig in. I figured creating an introduction video for my high school “AI Frontiers” class was a great (and meta) way to start.

How I Made It

I had just finished with some revisions to my course syllabus, so I had big picture, backwards design course goals in mind. That’s what I wanted to convey. I considered using AI to draft my prompts, but I knew this would be a learning by doing and failing experience. So I studied prompts from other Veo 3 clips and started writing.

I can make three 8-second videos a day. Over the course of about 10 days, I burned through each of those in order to get the final eight clips in the video above. My prompts for each scene are listed below.

I wrote what I thought was a good prompt and tried it. Then I adjusted the prompt based on what didn’t meet my expectations.

I was really intentional about each scene. I wanted it to feel like it could be my school without it actually being my school. Our main color is red, so I wanted each of the main characters to wear a red collared shirt as a visual thread. Our school is a Catholic, Jesuit all boys high school, so I matched that in the characters. I included Spanish architecture to match our buildings. There are guest stars in most of the scenes who clearly are not high school students. This was because I wanted there to be obvious indicators that this video was not real. And I also thought it was funny.

I edited the clips together in iMovie. I often trimmed the start or end of scenes to keep the pace I wanted, and sometimes to cut out actions I didn’t love. I made some minor audio level adjustments throughout, and added the school bell at the end. Try as I might, I could not get AI to include that sound in the scene itself. I faded to black at the end and that was a wrap.

What I Learned

As expected, I learned a lot just by getting my hands dirty, so to speak. I found that I could direct one character pretty easily. But directing two characters to interact with each other got a lot harder. As an example, getting the students to fist bump at the end of the first clip was one of the hardest parts of all of it, even with what I thought was clear prompting. A quick tip is to download videos as soon as they are done, even if you aren’t sure you’ll want them. Gemini holds them for two days and then they disappear. I could have made outtakes of the failed fist bumps/handshakes/high fives if I had downloaded them all.

You have to think about everything in a scene. If I were to make the cafeteria scene again, I would specify high ceilings. I didn’t think to include that in my prompt, so of course it gave me standard ceilings. If you want something specific on a shirt or wall like a number of bookshelves, you have to tell it what to do. Without that direction, it can be blank or random and might not fit at all.

If you want it to seem like the camera is moving in a certain way, you need to tell it that. If you want background characters to act naturally rather than stand still and stare at the camera or move in unison like a dance troupe, you have to direct that. If you don’t want sporadic, often jumbled captions, you have to tell it to exclude those.

In short, you have to think about everything in every scene and then how it all works together. Like a real director. Or perhaps a student really engaged in a class project.

The Future of AI Video

As I noted in my previous post, there are major implications that come with being able to easily create and publish realistic videos. You can use it to make novelty introduction videos for your high school class. And you can just as easily make videos that are intended to mislead, misinform, scare, or stir chaos. AI literacy is a massive need for schools.

On the brighter side, student projects are going to get so much better—especially if students act as producers and use the tool to creatively, thoughtfully, and maybe even humorously demonstrate understanding.

Before that happens though, we need greater access to Veo 3 or tools like it. That will come, I’m sure.

In the meantime, the AI game has changed again.

Who we are:
Dr. Dani Kachorsky, Ph.D.
Mica Mulloy, M.Ed
Jake Kelly, M.A

AI Prompts

ChatGPT image prompt

Please create this image in a square format and in a geometric abstraction style: A male teacher with fair skin, short brown hair and blue glasses stands center stage in a darkened, surreal amphitheater glowing with ethereal blue and orange tones. He is not in the foreground—he is deeper into the frame. He raises his hands like a symphony conductor with a baton gripped in his hand, and instead of instruments, abstract video elements float around him—glowing holographic storyboards, Veo 3 prompt boxes, digital teenage characters, cinematic camera directions swirling through the air. All of these elements are made with sharp, straight, glowing lines. Blue and orange light wraps around him like a cosmic current, illuminating a sense of awe, control, and curiosity. The background fades into mist and floating symbols of education—a school bell, a syllabus scroll, and AI circuitry weaving into stained glass patterns behind him. There is a professional tone rather than an elementary school tone. Do not include any text.

Google Veo 3 video prompts

Clip 1
A 17-year-old high school student is walking down his school's hallway. The school has a modern Spanish architectural theme. He's wearing a red collared shirt, tan shorts, and regular sneakers. His medium-length, messy blonde hair adds to his casual, confident look. He's also wearing a backpack. He has a very outgoing and confident personality. Other students, a diverse group of males, are in the background, chatting and laughing casually. They all appear friendly and happy to see each other. The camera's perspective is in front of the student, moving down the hallway with him. He looks directly into the camera, smiling enthusiastically. He says to the camera, "What's up boys! I hear this Artificial Intelligence class you are taking is amazing!" As he continues walking, a friend comes into the scene from the right, clearly excited to see him. The two students dap each other up in a friendly, casual way as they continue walking down the hallway.

Clip 2
A 17-year-old high school student is sitting in a brightly lit classroom at a standard student desk toward the front of the room. The room has bookshelves with textbooks on the walls. He is wearing a red collared shirt and blue shorts. He has short dark brown hair and has darker-toned skin. Desks are in rows. Class is about to start and his classmates, all males, are coming in, sitting down, and casually chatting with each other. Everyone is in a good mood, happy to see each other, and ready to start class. The student is smiling. He looks into the camera and says, "Bro, your class is totally one of a kind, and your teacher really knows his stuff." At the end of the clip a large Bigfoot walks in, pulls a chair back from a desk next to the student and casually sits down. The student looks at the Bigfoot and approvingly smiles and nods. Do not include any closed captions.

Clip 3
Two students are in a brightly lit, modern high school chemistry lab conducting an experiment. Main character one is an anthropomorphic fuzzy tan-colored teddy bear about 4 1/2 feet tall. Character two is a 16-year-old male who is about 6 feet tall with messy hair, wearing a red collared shirt, tan shorts and safety goggles. The bear is wearing lab safety goggles and pouring a green liquid solution from a beaker into a test tube. There are other pairs of average teenage male students in the background, all wearing safety goggles and collaboratively doing the same experiment in different stages at lab stations. Several of the groups have a small plume of green smoke pop out of the test tube and they look surprised when it happens. Character 2, the 16-year-old student, looks at the camera and says in a friendly, confident teenage voice, “You are going to learn so much in your class.” Then the bear pauses, looks at Character Two and says, “Do you ever get the feeling that you aren’t real?” Character two shrugs, then they continue the experiment. Do not include any closed captions.

Clip 4
A 16-year-old male high school student sits at a cafeteria table with a tray of lunch food in front of him. He has a medium build, short and messy brown hair. He is wearing a red collared shirt. The cafeteria is large and brightly lit with big windows in the background and has tan walls. There is a large crucifix on one of the walls since it is a Catholic school. The cafeteria is filled with average high school male students. They are eating and excitedly talking to each other. Some are laughing. The 16-year-old main character sets his piece of pizza down on the tray, looks at the camera with a smile, and says, “Yo, one of the best things about your class is learning the right way to use artificial intelligence, not just using it to take shortcuts.” Then he picks his pizza back up.
At the end of clip a Storm Trooper walks across the scene behind the 16-year-old student with a tray of lunch in his hands, looking for a place to sit. Do not include any closed captions.

Clip 5
A middle-aged Jesuit Catholic priest with short gray hair, wearing a black shirt and traditional white priest collar, sits prayerfully in a church pew at the front of church. It is an old, dark, narrow and long Spanish-style church with stained glass windows along the walls. The camera is at the front of the church looking at him, and the pews behind him. The church is empty and quiet except for the priest and the school mascot—a large, red horse with a jersey that says “BROPHY.” The horse is sitting several rows back and to the right of the priest and he is in quiet, peaceful prayer the whole time, holding a large rosary in his hands. The priest pauses his prayer, looks up at the camera and says in a calm, wise voice, “What I like most about our approach to Artificial Intelligence is the intentional tie to our Catholic faith.” Then he goes back to prayer. Do not include any closed captions.

Clip 6
A high school football player takes his helmet off after a play and jogs off a football field at night toward the camera. The field is outside and lit by tall, bright stadium lights. There are a couple of hundred fans in the bleachers on both sides of the field cheering on the game. The player is average height and 16 years old. He has medium-length hair and is sweaty from playing football. He is wearing white football pants, a red jersey over his standard pads with the “BROPHY” and the number 11 on the jersey, and a red football helmet in his hands. The rest of the team is on the field behind him, getting ready for the next play. There is a referee on the field who is a robot and looks just like “Johnny 5” from the movie “Short Circuit.” The robot referee is wearing a striped referee shirt, has a whistle and is rolling across the field. The player jogging toward the camera comes to a stop and says in a positive, friendly tone, “For us, creativity and human connection are at the core of artificial intelligence.” Do not include any closed captions.

Clip 7
A 16-year-old African American male high school student is sitting at a table in a high school library. The library is large and has high ceilings. It has Spanish architecture-influenced arches along the tan walls. There are rows of bookshelves behind the student and other study tables, where other average male students are sitting studying and chatting quietly. The 16-year-old is wearing a red collared shirt and headphones to listen to music. His personality is outgoing and confident. He has a notebook and an “AP Calculus” textbook on the table in front of him along with pencils and his backpack. He looks at the camera and says, “Boys, you aren’t just learning about AI, you are revolutionizing how we learn!” We see an anthropomorphic fuzzy tan-colored teddy bear about 4 1/2 feet tall walking down one of the rows of bookshelves in the background looking for a book.

Clip 8
A clean-cut, confident 17-year-old high school student with short brown hair, a red collared shirt, blue shorts, and tennis shoes stands in the middle of an outdoor school courtyard smiling. The camera is in front of him. It is a sunny day with blue skies. The courtyard has a large, round concrete section in the middle where the student is standing. There are sections of grass and some trees around the courtyard, and various two-story classroom buildings in the background surrounding the courtyard, which are all tan and feature Spanish architecture arches. The courtyard has about 100 diverse male high school students in the background casually walking in different directions between buildings, unaware of the camera. They are all happy and talking to each other. The 17-year-old main character looks into the camera and excitedly says, “Alright fellas, it’s time to get to work! Let’s go!” As soon as he finishes talking, the camera smoothly lifts into the air as if connected to a drone and continues to rise as all the students go to classes in the different buildings. Walking among the students is one Stormtrooper and one 7-foot-tall Bigfoot. Do not include any closed captions.