Hypergamblification

September 2025 • Concept, Design, Development

September 2025 • Concept, Design, Development

Prediction markets claim to democratize information. Markets aggregate knowledge better than experts, the argument goes, and platforms like Kalshi position themselves on that premise. In practice the interfaces look like DraftKings. The markets trend toward spectacle. Will it snow in April. Will the government shut down. The line between informed forecasting and gambling gets thin when the product is designed to feel like a sportsbook. This project follows that logic to where it leads. A Chrome extension that watches what you read and surfaces real Kalshi prediction markets on top of the content, styled like a betting app. The gap between reading the news and gambling on it is just wiring. As these platforms proliferate, the question is what everyday information consumption starts to look like.

My contribution

Concept, Design, Development

The team

Sean Lee

Surya Narreddi

Experience

Sports betting apps have the most psychologically optimized UI patterns in consumer software. The pulsing green dot. The uppercase "LIVE" label. Green-for-yes, red-for-no. Fake bettor counts. The overlay borrows all of it. Small images get a compact Yes/No badge. Large images get the full treatment: glassmorphism ticker, betting pills with hover glow, volume figures, close date, "Bet Now." The gradient bleeds into the image from the bottom. Everything is designed to feel urgent and actionable. The entire aesthetic is replicated in CSS and applied to whatever you happen to be reading.

The extension in use on AP news

How it Works

The extension watches what you scroll past. Every image on the page is tracked by distance from your viewport center. When one enters view, the extension reads the surrounding page: headlines, captions, nearby text. If there's enough context it searches Kalshi directly. Under 200ms. When context is thin it falls back to GPT-4o-mini vision, which looks at the image and its surroundings and generates broad search terms. "Trump," "tariffs," "recession." Prediction markets trade on events, so the keywords stay broad.

The extension watches what you scroll past. Every image on the page is tracked by distance from your viewport center. When one enters view, the extension reads the surrounding page: headlines, captions, nearby text. If there's enough context it searches Kalshi directly. Under 200ms. When context is thin it falls back to GPT-4o-mini vision, which looks at the image and its surroundings and generates broad search terms. "Trump," "tariffs," "recession." Prediction markets trade on events, so the keywords stay broad.

Browser (Chrome Extension)Fly.ioOutputfast path · ~50msslow path · ~1-2syesnoContent Script(every page)IntersectionObserverviewport trackingContext rich?DOM Extract~50msCapture→ dataURLService WorkerbackgroundGPT-4o-minivision · ~1sKeywords"trump" "tariffs"Series Cache~1ms · in-memoryKalshi API~200ms · liveOverlay4 tiersYESNO
fast pathslow pathexternal

Project Architecture

You can email me. I’m active on Twitter, occasionally on LinkedIn, and surfing the internet on Are.na.

©2019-2026 SURYA NARREDDI.

You can email me. I’m active on Twitter, occasionally on LinkedIn, and surfing the internet on Are.na.

©2019-2026 SURYA NARREDDI.