Help us bridge the gap between AI and humans in telling stories about movies!

Hi everyone,

We are looking for MTurk workers with Master qualification (!) to complete a storytelling task.

In each HIT, you will be shown a sequence of images (frames from movies, which you may or may not recognise). You will be asked to write a coherent, visually grounded story describing what's happening in the images.

Your story should (1) mention relevant characters and events visible in the images, (2) flow naturally, but coherence, flow and narrative style are up to you. We are very much interested in how you interpret and connect the scenes as a story.

How to find our HITs: search for "CLASP AMT" on MTurk. We publish a few new HITs every day.

We pay 4$ per story and we also provide an additional bonus depending on the quality of your work. Each HIT includes detailed instructions to guide you through the task.

Your stories will directly support our research on visual storytelling and narrative understanding. In a world filled with generic AI-generated content, your contribution is very important.

Cheers,
Nikolai

 

Masters qualifications haven't been given out for years. You may do better looking for MTurk Workers with a 99% approval (which is required of Masters) and at least 20,000 approved HITs.