Events · Recurring format
AI Media Nights
Image and video, made with AI - and the pipelines behind them.
Being planned
One talk, one room, a drink in hand. Somebody shows something they made with generative AI in twenty to thirty minutes, and the rest of the evening is the questions and the conversation.
What it’s about
- Video - Seedance first of all, and the others it gets measured against: Veo, Kling, Sora, Runway, and open-weight models like Wan that run on your own GPU. What a minute of film takes now, in time, money and patience - and the sound that increasingly comes with it.
- Image pipelines - node graphs in ComfyUI, Flux and Stable Diffusion, LoRAs trained on your own style, ControlNet, upscaling, and keeping the same character consistent from the first frame to the hundredth.
- Under the hood - calling the models through their APIs, batch jobs, running them locally versus in the cloud, and what it costs per second of video.
And the questions that come with it: who owns the result, and what it does to the people who used to make these things by hand.
Who it’s for
Developers who want to build with these models, and everybody who makes things: designers, photographers, filmmakers, marketing. The evenings are better for the mix - one side knows what looks right, the other knows how to wire it up. You do not need to write code to follow along.
How an evening runs
- The talk - twenty to thirty minutes. Show the result and the workflow behind it, not slides about it.
- Questions and discussion - the longer half. Settings, graphs, tricks, dead ends.
- Drinks and the rest of the evening - free for members, guests donate towards drinks and snacks.
Give a talk
Made something with an image or video model, or built a pipeline around one? Write to hello@hagenberg.international and tell us in two lines what it is. We sort out the room, the date and the invitations; you turn up with the thing.
Building rather than making?
Apps built with AI tools, and how the models actually work, are the AI Builder Nights.