AI Video Generation: Veo 3

[UPDATE] So, we screwed up; we were using Veo 2! We never actually toggled it to Veo 3 and because we weren't actually attempting results with audio yet, we didn't realize it. No judging! Check out our ACTUAL Veo 3 results here.

The implications of AI on the art world have been profound, to say the least. This technology is transforming and disrupting countless businesses and industries, seemingly overnight.

B.C. Design has incorporated AI-based image generation into a number of our previous projects, specifically, to enhance our mock-ups. Until recently, we have not been able to create AI generated videos that we felt could pass a "minimum usability threshold". (MUT)

That is to say, the technology for videos had not advance far enough for it's use to make sense for our clients.

That is...until now.

This past week, Google released its Veo 3 AI video generation model, which to our delight, was able to create usable video. This model, released May 30th of this year, "is Google’s latest AI video generation model, announced at Google I/O 2025. It transforms text or image prompts into high-definition videos, now with native audio integration. This means Veo 3 can generate synchronized dialogue, ambient sounds, and background music, producing clips that feel remarkably lifelike." (Datacamp.com)

Our MUT guidelines are simple:

• short-form video (5-8 seconds)

• foreground and background characters

• realistic, anatomically correct, human faces and bodies

• realistic clothing, motion, and settings

• generally, passable as background footage to naked eye

The technology is progressing toward a point where long-form, feature length films are possible, at least, theoretically. For B.C. Design, the most practical use (in the present) are short-form videos for clients that make use of multiple AI generated clips edited together. A social media reel or "story" is ideal.

In the first example, our Creative Team asked Veo 3:

"3 african american college students, laughing and smiling, walking downtown Washington DC in the afternoon, wearing burgundy, gold and white sweaters and jackets, cinematic, 5 seconds"

In it's full length, the video has a number of imperfections, glitches, and giveaways that it is AI. The students appear to awkwardly smile too long. That said, when we consider the use of multiple similar videos, edited in quick succession, said anomalies are far less noticeable, thus giving us "passable" footage.

In the second example, our Creative Team asked Veo 3:

"2 african american college students, walking past US Capitol building in downtown Washington DC, evening, wearing burgundy, gold and white sweaters and jackets, cinematic, 5 seconds"

This clip could easily be worked into an series of edited clips or be used outright at it's entire length. The unspoken goal was to create background footage of "UDC" students for upcoming video projects in order to save the client both time and money. If we treat this as "B-roll", this footage may have called for: filmmakers/film equipment, transportation costs/time, arranging use of extras/actual UDC students, relevant filming permits, make-up, lighting, etc. Those resources are better served on principal photography. In a hypothetical video shoot for UDC, our film crew could be on campus shooting footage while our Design Team is generating b-roll shots/scenes in various locations across the city.

In the final example, our Creative Team attempted to create 30-60 second promo video for Seneca Village State University, one of our "mock-up" brands:

The final result is not perfect. We cannot incorporate custom logos at this time, though we imagine that will change sooner rather than later. Color consistency is also an issue. Again, in a short-form promo, will clips changing in quick succession, with text or other overlays, there are valid uses of footage generated in this way. The clips took 2-5 minutes to generate. Editing in Capcut took about 10 minutes. The music is royalty free and sourced via a Google Search. In total, this is about 15 minutes of labor.

B.C. Design-Creative is excited to continue exploring and researching how AI-based video generation can serve our clients, our partners, and our business.

Stay tuned!