Cast
Angelica
Angelica takes a leading role in the students’ technology project. She is interested in how the technology works and pays attention to the details of recording and building a world model. She understands AI well and is willing to challenge the guys on technology discourse. Her humour is usually dry and matter-of-fact. The actor should be able to make technical dialogue sound familiar and conversational.
- Playing age
- 20–26
- Language
- Good conversational English. You do not need to be a native speaker.
- Filming
- Summer 2027
- Days
- Around 13 days
- Where
- Northern/Eastern Europe
- Pay
- From €175 per shooting day
- Expenses
- Travel and accommodation are covered
All six principal roles include outdoor scenes in forest and by a lake, swimming, and evening scenes around a campfire.
Audition sides
The scenes below have been prepared for self-tapes. You can record them on a phone and add the link to the application form below.
Side 1: Three ways to study
Lines of other participants: Nadja (4), Sonja (3)
Setup. The second day of a camping weekend at a remote lake. Six university friends are building a world model of the lake for a group project, with digital doubles of themselves in it. Back at the camp, the software is training the doubles. Angelica, Nadja and Sonja walk along the shore to find somewhere to swim. Angelica knows more about AI than the others. For more context, read the extended synopsis.
EXT. SHORE PATH — DAY
A narrow path along the water. Angelica, Nadja and Sonja walk in single file with towels over their shoulders.
SONJA Did you hear Peter got an A in signal processing?
NADJA I got a C.
SONJA He read the whole book the night before. Cover to cover. He says he remembers all of it and understands none of it.
ANGELICA He’s a fully connected network.
NADJA Is that an insult?
ANGELICA It’s the oldest kind there is. Everything connected to everything. It takes in every fact at once and has no idea which ones belong together. You could shuffle the sentences and the pages, the same way every time, and it would learn just as well.
NADJA So how do I do it better?
ANGELICA Like a convolutional network that keeps track of close-by connections. It reads the book sentence by sentence, page by page, and then makes more sense of it all. It’s how they got good at understanding pictures. Detecting small things that look the same everywhere, like edges, then detecting big stuff by combining the edges, like eyes and such.
NADJA So it takes the edges of an eye and fits them together into an eye.
ANGELICA Yes, with switches in between. The “non-linearities”. They’re what make the edges turn into an eye, and not just a pile of edges. That’s where it gets it.
But the modern networks don’t need to read the book cover to cover, sentence by sentence. Instead they attend to earlier sentences and facts. So when it reads “her”, or “Theorem 4”, it can go back in the text and check who she is, or what the theorem says.
She makes quotation marks in the air.
ANGELICA “Attention is all you need.” That’s the motto.
SONJA Wow. Are you sure that’s not about my ex?
Nadja laughs. They come out onto a flat rock at the water’s edge. Sonja drops her towel and walks straight in.
Side 2: The plateau
Lines of other participants: Carl (1), Max (1)
Setup. The first night of a camping weekend at a remote lake. Six university friends are building a world model of the lake for a group project, and a phone by the fire is recording their voices for it. Peter has just explained backpropagation, using a sausage on a stick to show the route: a network makes a guess, measures how wrong it was, and sends the error backward so it does better next time. Angelica has trained networks herself and adds something Peter left out. For more context, read the extended synopsis.
EXT. CAMPFIRE — NIGHT
ANGELICA And sometimes it gets stuck. You keep training and the error doesn’t move. It just sits there, flat, for ages. A plateau. It looks like the model has learned everything it’s ever going to learn. And then, if you keep going anyway — bang. It suddenly gets it.
CARL How do you know if it’s a plateau, or if it’s just finished?
ANGELICA You don’t. You keep going to find out.
MAX That’s me! I plateaued like crazy in Calculus, first semester. Then — bang — I got everything in the last two days and nailed a D.
Angelica laughs a little. Then she looks at the fire.
ANGELICA Peter. Your sausage is on fire.
Peter pulls his stick out of the flames. The sausage is black.