Kalship

banner
stack
drake
superbowl
facts

🧐 Inspiration

Short-form content is addictive because discovery is frictionless—one swipe and you're onto the next idea. While people scroll, they're constantly making predictions in their heads. Why not bring Kalshi to where that attention already lives...? What if prediction markets were embedded directly into the short-video feed experience?

That's Kalship: Kalshi for Youtube Shorts.

🤯 What it does

Kalship is a scrollable short-form video platform where each short is paired with a contextual prediction market card. It turns passive scrolling into an interactive loop of content → conviction → simulation.

💭 How it works

1. Scroll the reel
Users browse a YouTube Shorts-style vertical feed as usual.

2. See a related market instantly
For each short, Kalship surfaces a contextual bet card on the side (e.g., YES/NO market tied to the topic in the video).

3. Take a position
Users choose the side they believe in and submit their pick.

4. Generate the “if true” outcome video
After a bet is selected, Kalship creates an AI-generated short showing what the world could look like if that prediction is true.

5. Keep watching while generation runs
Everything is async, so users can continue scrolling immediately — no waiting, no blocking.

6. Receive generated clips back in-feed
When rendering is done, the generated “future outcome” short is inserted into the reel as fresh short-form content.

This creates a seamless loop: watch → bet → generate → watch.

🤦‍♂️ Challenges we ran into

Veo 3.1 isn't as good as we thought it was: The hype is fake. Veo AI had us crashing out. We had Sam Darnold throwing backwards spirals to Drake Maye at one point...

Queues: We spent massive amounts of time debugging the numerous queues.