It can talk


The rules of content, creativity, and connection have changed.​

Read time: 5 mins

Hi Reader,

Veo 3 videos are probably the biggest thing on the internet over the past week.

The way we create will never be the same again.

AI video is here to stay.

And it will only get crazier from here.

Most people will miss this opportunity. They think it’s too technical. Too expensive.
But the truth is: audiences care less and less about how something is made.
They care if it’s entertaining, educational, or emotionally on point.

So let’s break it down.


From the experts

"Art challenges technology, and technology inspires the art." - John Lasseter, Pixar co-founder and animator

"Technology will not replace great teachers, but technology in the hands of great teachers can be transformational." - George Couros, educator and author

"Soon it’ll be unthinkable not to have intelligence integrated into every product and service. It’ll just be an expected, obvious thing." - Sam Altman, OpenAI CEO


It can talk

AI has two superpowers.
The ability to make use consume more than we wanted.
The ability to help us create better and faster than ever.

This last week, there was one thing that took over my schedule.

I'm not a video maker. But the ability to create high-quality videos with a prompt and for the cost of a coffee and a sandwich gives me the ability to amplify my work. I find it an opportunity worth exploring.

Now, it can talk. Listen for yourself.

video preview

Where to access Veo 3

Veo 3 is part of Google One AI Ultra (€274.99/month, or €139.99 for the first 3 months). This includes 12,500 credits (enough for ~125 generations).

But here’s the smarter way:

  • Platforms like Freepik and Krea now integrate the Veo 3 API.
  • You only pay per prompt. No subscription required.

How to prompt Veo 3

To avoid spending a fortune on generating with Veo 3, let’s go over some best practices.

The more descriptive you are, the better.

What’s the format of the video? Who are the characters? What are they saying? How are they saying it? What does the environment look like? What camera is being used to film it? Is it hand-held?

You get the point. The more descriptive you are with your prompt, the better results you will get.

The easiest way to come up with the best text-to-video prompts is to use ChatGPT or your favorite LLM.

Include all of the details you can think of, and let ChatGPT format the prompt.

Here's one I made earlier.

I asked ChatGPT:

"If you were the devil, how would you destroy the minds of future generations without them realising?"

It gave me a 9-point response.

I used those points to build 9 Veo prompts, and it provides the following:

Added sound, edit it in CapCut... and here's what came out:

Veo 3 project that is blowing up

Stormtrooper vlogs. Veo-generated vlog series of Greg the Stormtrooper is the best Star Wars content in a long time.

It went from 0 to 300k followers in a week after posting 20 videos. StormTrooperVlogs lets us view a well-known IP through the eyes of the most throwaway and faceless character (stormtroopers) in a different but familiar genre to anyone who has scrolled short-form videos in the 2020s (“day in the life of an office worker” vlogs).

The office worker vlog format works well with Veo 3 because Google’s video-generating AI tool can only create 8-second clips at a time. It’s also beneficial that the Stormtroopers are wearing masks because perfectly aligning lip movement with audio is still a challenge.

There are clear trademark and copyright questions, but I suspect StormTrooperVlogs meets the criteria for “fair use”. It is both transformative and parody (stormtroopers are vlogging about work while shitting on their boss Vader).

Key stat: the two seasons of Andor cost over $600m while StormTrooperVlogs probably cost less than $1,000 in tokens for each clip (this assumes the creator needs 20 usable Veo clips, which probably requires 100 total prompted videos…at ~$5 for an 8-second prompted video, that would be a total of $500 and then add on time spent editing).

The future of ads

But just because this was cheap doesn’t mean anyone can do it this quickly or effectively. You still need experience, skill and vision to make it look like a real commercial.

I’ve been working on content in some form for 11+ years, and just because something can be done quickly, doesn’t mean it’ll come out great. But it can if you have the right team.

I compare this to recent experience with marketing and creative agencies, and I'm now convinced of one thing.

The very near future is small teams making viral, brand-adjacent content 100x faster than agencies, getting 80 to 90 percent of the results for much less overhead and staff costs. (Here's an example of my own - this one made with Runway.)


Escape the algorithm

Video: made with AI but not about AI (4 mins)

Video: During Game 3 of the NBA Finals, online prediction market Kalshi streamed an advert made with Google’s Veo 3 video generator. The ad appeared on the YouTube TV livestream of the game.

video preview

Prompts: Specifically for video AI:

A cinematic handheld selfie-style medium shot, set on a snowy battlefield at dusk. A stormtrooper in full white armor holds the camera at arm’s length, his helmeted face perfectly framed as snowflakes swirl gently in the cold air. His armor is lightly dusted with frost and ash. Behind him, a vast frozen landscape stretches into a shallow-focus blur—explosions flicker in the distance, and streaks of missile trails arc across the twilight sky.
The stormtrooper slowly pans the camera sideways, revealing another stormtrooper crouched in the snow, carefully sculpting a snowman with exaggerated focus. Bits of snow cling to his gloved hands and leg plates as he works.
Back on camera, the first stormtrooper yells with a sense of urgency: “Okay so we’re in the middle of an active firefight, people are screaming, and Greg’s building a darn snowman.
A high-energy, spontaneous TikTok-style street interview.
Daytime setting, filmed vertically on a smartphone. A confident, charismatic interviewer approaches a random pedestrian in a busy public area (like Broadway, Venice Beach, or downtown Nashville) and asks an unexpected, provocative, or playful question, often joyous or shocking.
The camera is handheld, slightly shaky, with quick zooms and punchy cuts. The interviewee delivers an outrageous, meme-worthy response with bold facial expressions and charm. Bystanders laugh or react in the background. Humor is raw and unapologetic. Urban setting, natural lighting, with ambient street sounds and fast-paced editing. The interviewee's response details just how much she loves tech commentary and newsletters from Has.

Definitely not viral but still a bit of fun.

Tool: cobalt helps you save anything from your favorite websites: video, audio, photos or gifs. Just paste the link and you’re ready to go. No ads, trackers, paywalls, or other nonsense. Just a convenient web app.


If you got this far,

you're part of 71% of subscribers, so thank you, Reader.

For creators, it’s game-changing. For everyone else? Kinda terrifying.

We’re entering an era where deepfakes, fake news, and AI propaganda are cheap and convincing. The line between what’s real and what’s made up is about to vanish.

Do you think this will wake people up, or blur the line even more?

Developing your own criteria happens by opening up a tool and trying it yourself.

That's what we discussed at an event I took part in last Friday.

If a section has particularly interested you, click the link of the corresponding section. It helps me prepare future newsletters.

If this email was forwarded to you, please consider subscribing.

See you in two weeks.

Peace,

Has


Whenever you're ready, this is how I can help you:
- Get my toolkit (free). It contains the tools I'm using today
- Book a video call. Let's work on a project together.