Author: Yetta, Investment Partner at Primitive Ventures; Wildon, Researcher at Primitive Ventures
Weekly trading volume of $1.5 billion; cumulative trading volume exceeded $10 billion within 60 days. While everyone is applauding Polymarket’s $13 billion valuation and its high-profile partnership with the New York Stock Exchange, this Eastern platform, still in its “seed stage,” has quietly achieved a bottom-up breakthrough. Here is our complete investment thesis on Opinion Labs.
The era of single-machine NPCs is coming to an end.
For decades, we have lived in a “single-player game.” Truth was produced by a few institutions, narratives were finalized in boardrooms, and most of us were merely NPCs following a predetermined script. Web2 gave us a voice, but it didn’t truly let us help shape the direction. How beliefs form and how consensus is built remains locked inside algorithmic black boxes and power structures.
But what if belief itself could flow and be priced?
Opinions aren't just about expressing attitudes—they're about placing bets.
Divergence is not just an emotional clash—it's a price split.
Consensus, not retweets, is a curve that can reverse at any moment.
Reality is no longer a script to be watched, but a multiplayer game.
The paradox of not being so "long-tail"
It’s an interesting discussion whether prediction markets are financial markets or content markets. As financial markets, they perform exceptionally well in a few specific scenarios. Without liquidity, there is no price; without price, there is no signal. Mechanisms such as order books, market makers, and depth function smoothly during high-profile global events like the U.S. election, but they begin to break down once outside these contexts.
Human attention is inherently fragmented. Most of the things we truly care about are highly dependent on specific contexts and regional backgrounds, such as cultural gossip, local politics, celebrity scandals, and various social topics. The internet has not converged these interests; instead, it has caused them to diverge infinitely. Content is growing explosively, but capital and liquidity remain perpetually scarce.
Thus, the paradox arises:
Markets require concentration to be efficient;
Human opinions are naturally decentralized.
If a platform can only maintain liquidity by relying on a few major events, it functions more like an “event exchange” than a true market where beliefs are traded. So the question becomes: how do you build a financial system on top of something that refuses to converge?
Why we invested in Opinion Labs
A year ago, we invested in @opinionlabsxyz, when Polymarket had just completed its first large-scale validation during an election cycle, making Prediction Markets one of the strongest narratives in Western VC circles. We believed that this content- and event-driven model would not be absent in the East, so we reached out to Opinion to make the investment. Six months later, discussions around Prediction Markets spread across Asia and to BNBChain. At that moment, Opinion was the only platform with a mature product capable of launching immediately and capturing that momentum.
More importantly, they achieved this scale with almost negligible capital expenditure. In terms of product velocity, execution density, and output per person, they are among the most efficient teams we’ve ever seen.
East ≠ West: Prediction Markets Are Diverging
Predictive markets are undergoing a clear East-West divide, and the reason is simple: assets can converge globally, but opinions cannot. The dollar, gold, and U.S. stocks can form unified prices, but what people are willing to bet on is inherently a cultural product. The true expansion of predictive markets depends not just on liquidity, but on shared attention.
In the United States, attention is extremely concentrated. Sports betting has cultivated retail habits over decades, and politics has long evolved into a nationwide reality show. Attention naturally converges on a few super events. Kalshi and Polymarket naturally focus on top-tier markets, serving high-frequency and professional traders with deeper order books.
The situation in East Asia is entirely different. China has limited space for political discourse, Japanese retail investors generally exhibit low levels of political engagement, and Korean users tend to focus more on speculation, entertainment, social issues, and pop culture. The differences across markets are substantial—retail investors care about different things, and their level of understanding to form informed opinions varies greatly.
This difference is very clearly reflected in the data. Under the event of the 2026 South Korean presidential election, Polymarket’s trading volume was approximately $400,000, while Opinion Labs reached $52 million. Prediction markets cannot be winner-takes-all, because belief formation is highly localized.
When shared attention is absent, liquidity does not automatically consolidate; instead, it naturally disperses. The structure we observe reflects this exactly: compared to Polymarket and Kalshi, Opinion’s trading distribution is significantly more dispersed, with volume not collapsing into a few top events, but rather continuously spread across the long tail of markets that offshore users genuinely care about.
AI as a key to scaling
Markets driven by humans can no longer keep up with this era. When opinions are infinite and contexts are highly localized, relying on manual curation and manual listing forces prediction markets to converge only on a few top events.
This is exactly why AI has become indispensable.
At Opinion Labs, AI transforms fragmented claims into structured markets: automatically generating clear settlement rules, defining failure boundaries, and aligning with staking mechanisms to constrain behavior and reinforce accountability. Market creation no longer relies on the judgment of a few editors or operations teams, but can be scaled exponentially. Numerous local markets with cultural and regional characteristics can be rapidly generated without being overwhelmed by spam.
When markets begin to extend into the long tail, the real challenge is settlement. Long-tail markets often rely on complex, fragmented, and unstructured sources of information. A single arbitration mechanism cannot handle this complexity. AI can scale the cross-referencing and verification of information, while hardware-level security and human governance serve as the final safeguard.
This means that prediction markets have the potential to expand horizontally for the first time—shifting away from compressing global attention onto a few super-events, and instead enabling beliefs themselves to be organized, traded, and verified in a highly decentralized world.
Prediction Markets 2.0: A Paradigm, Two Evolutions
On prediction markets, a clear divergence is emerging between the East and the West.
The West has chosen financialization and institutionalization. That’s why we hear the IB founder say that the most frequently traded contracts are weather/temperature contracts. More broadly, industries such as energy, agriculture, and shipping have long been exposed to climate risks, yet have consistently lacked sufficiently refined and tradable hedging instruments. In this context, the evolution of prediction markets has shifted toward whether they can scale liquidity, create robust derivative structures, and gain understanding and acceptance from institutional risk management systems. Entrepreneurs are focused on refining them into viable financial infrastructure.
The East has moved toward internetization and contentization. It functions more like an entertainment-driven mechanism for information consumption and expression. In this context, the core question is: at what moment are users willing to place a bet on what? It is a form of content monetization: betting represents engagement, odds reflect narrative intensity, and trading volume reflects emotional consensus. The product does not solve complex financial engineering, but rather a content operations logic—how to transform trends, public opinion, and social discussions into sustained trading motivation.
Prediction Markets have entered 2.0. They are no longer single, winner-takes-all markets, but rather two distinct evolutionary paths shaped by culture. Together, they point to a larger shift: markets are no longer merely venues for trading outcomes, but have become atomized mechanisms for handling uncertainty itself.