Superteam Demo Day: CHOMP (Kiko Zang)

Can we still trust what we see on the internet? According to CHOMP co-founder Kiko Zang, the answer is a resounding no—and their Solana-based quiz game with embedded prediction markets might just be the solution to restoring honesty to our digital lives.

Summary

At Breakpoint 2025's Superteam Demo Day, Kiko Zang presented CHOMP, a deceptively simple quiz game that functions as a "Trojan horse" for more accurate internet information. The platform addresses what Zang identifies as the internet's core problem: not misinformation itself, but rather the massive gap between perception and reality that shapes our beliefs online.

CHOMP has achieved remarkable engagement metrics since launch, with 50,000 users contributing over 2 million data points. Perhaps more impressively, the platform boasts thousands of users maintaining week-long daily streaks, dozens with streaks exceeding 100 days, and two users who have returned every single day for over a year—despite the product only being live for just over 12 months.

The platform's approach is elegantly simple: users participate in a fast-paced quiz game, knowing that honest responses lead to better market outcomes. This creates a self-reinforcing mechanism for truthful data collection. The aggregated data creates what Zang calls a "new data primitive"—opt-in, validated, honest human input that can serve multiple purposes, from training AI agents to creating more contextually aware internet services.

The presentation highlighted CHOMP's early success in surfacing hidden information, including election-related insights in July 2024 that preceded other prediction markets. Looking forward, the team sees significant revenue potential, particularly in providing human reinforcement training data for AI agents that need to behave more naturally.

Key Points:

The Real Problem: Missing Information, Not Misinformation

Kiko Zang challenged the conventional narrative around internet trust issues. While most discussions focus on misinformation and disinformation, CHOMP's thesis is that the true culprit is missing information. There exists a significant gap between reality and our perception of it online, with our beliefs being reinforced by this distorted view.

This framing shifts the solution from policing content to aggregating better, more honest signals from users. By leveraging prediction market mechanics where honesty is financially incentivized, CHOMP creates conditions where truthful input is the rational choice for participants.

Gamification as a Trojan Horse for Truth

CHOMP's core innovation lies in its approach to data collection. Rather than asking users to consciously participate in verification exercises or prediction markets—which have traditionally been domain of chart-watchers and traders—the platform disguises these mechanisms within an engaging quiz game.

Users understand that honest answers improve their outcomes, creating alignment between user incentives and data quality. This "Trojan horse" approach has proven remarkably effective, with engagement metrics that rival consumer social apps rather than typical crypto applications.

Unprecedented User Retention

The platform's retention statistics tell a compelling story. Beyond the headline numbers of 50,000 users and 2 million contributions, the depth of engagement is striking. Multiple users have maintained daily engagement for over 365 consecutive days on a product that's only been live for slightly more than a year.

This level of retention suggests CHOMP has found product-market fit in a space where user engagement is notoriously difficult to maintain. The game mechanics have successfully created habitual usage patterns that could prove invaluable for sustained data collection.

AI Training Data as a Revenue Model

Looking beyond the immediate quiz game, CHOMP positions itself as infrastructure for the AI age. The platform generates what Zang describes as a "new data primitive"—validated, opt-in human input that can be used to train AI agents to behave more naturally and make better decisions.

As AI agents proliferate, the team believes there will be ongoing demand for human reinforcement training data. CHOMP's established user base and engagement patterns position it to become a supplier of high-quality human feedback data to AI developers.

Election Prediction Success

The presentation included evidence of CHOMP's predictive capabilities, referencing a graph showing the platform surfacing accurate election information in July 2024—ahead of other established prediction markets. This early validation suggests the aggregated wisdom of CHOMP's user base can produce meaningful, actionable insights.

This success demonstrates the platform's potential beyond entertainment, positioning it as a legitimate source of crowd-sourced intelligence on real-world events.

Facts + Figures

• Over 50,000 users have engaged with the CHOMP platform since launch
• Users have contributed more than 2 million data points through the quiz game
• Thousands of users have maintained 7-day or longer daily engagement streaks
• Dozens of users have achieved streaks exceeding 100 consecutive days
• At least two users have maintained daily engagement for over 365 days straight
• The platform has been live for just over one year
• CHOMP surfaced election-related information in July 2024, preceding other prediction markets
• The platform combines quiz game mechanics with underlying prediction market infrastructure
• The founding team consists of two co-founders: a "thinker" and a "doer" (Kiko Zang)

Top quotes

• "Do you trust what you see on the internet today? I'm glad that's your answer. Otherwise, I'll be like, what internet have you been using?"

• "It's over, isn't it? The internet is a bit fucked. We can't really trust what we see."

• "We believe that is not the true problem. The true problem is the missing information."

• "There is a huge gap between reality as we know it and what we see on the internet and that perception then reinforces our beliefs."

• "A lot of us here believe that markets is what creates more valuable and more accurate information. But rest of the world do not look at charts. They are scrolling."

• "We have a little fun game that is more of a Trojan horse."

• "The more honest they can be, the better their market outcomes can be."

• "We've created this new data primitive that creates opt-in, validated, honest human input for all of us."

• "My co-founder is the thinker, I'm the doer. And together we are building the new context engine for the internet."

• "Finally, all of us humans and agents can get what we deserve, which is an honest internet."

Questions Answered

What is CHOMP and how does it work?

CHOMP is a quiz game built on Solana that incorporates prediction market mechanics under the hood. Users play a fast-paced trivia-style game where their honest responses are incentivized by better market outcomes. The platform aggregates these user inputs to create more accurate information about various topics. Rather than requiring users to engage with complex trading interfaces or charts, CHOMP makes participation accessible and entertaining while still capturing the wisdom of crowds that makes prediction markets valuable.

Why does CHOMP focus on missing information rather than misinformation?

The CHOMP team believes the internet's trust problem stems not from false information being spread, but from critical context and truth being absent. When there's a gap between reality and what people perceive online, those distorted perceptions reinforce existing beliefs and create echo chambers. By focusing on surfacing missing information through aggregated honest human input, CHOMP addresses the root cause rather than playing whack-a-mole with individual pieces of false content.

How does CHOMP incentivize honest responses?

The platform creates alignment between user interests and truthful participation through its prediction market mechanics. Users understand that providing honest answers leads to better outcomes in the game. This creates a self-selecting mechanism where accuracy is rewarded, unlike social media platforms where engagement metrics often favor sensationalism over truth. The opt-in nature of participation also means users are consciously choosing to contribute validated information.

What business model does CHOMP pursue?

While the presentation didn't detail specific revenue streams, CHOMP identifies AI training data as a major opportunity. As AI agents become more prevalent, they require human reinforcement training data to behave naturally and make better decisions. CHOMP's validated, opt-in human input creates a "data primitive" that can be licensed to AI developers. The platform's proven engagement metrics make it a reliable source of ongoing human feedback data.

How has CHOMP demonstrated predictive accuracy?

The platform showcased its ability to surface accurate information early, specifically referencing election-related insights that emerged in July 2024—before other established prediction markets detected similar signals. This suggests that the aggregated wisdom of CHOMP's engaged user base can produce actionable intelligence on real-world events, validating the core thesis that gamified prediction markets can generate valuable insights.

What makes CHOMP's user engagement remarkable?

CHOMP has achieved retention metrics unusual for both crypto applications and consumer apps generally. Beyond 50,000 total users and 2 million contributions, the platform has users who have engaged daily for over a year straight. This suggests the game mechanics have created habitual usage patterns, and users find genuine value in returning regularly. Such retention is crucial for a platform that relies on ongoing human input to generate its core data product.