The Return of Independent Judgment

Today, prediction markets allow participants to place bets on the outcomes they believe are most likely to occur. How people decide on that most probable prediction may vary based on different factors. The participant’s location, the media they are exposed to, or the topic the person is knowledgeable about. People are forced to consume information filtered through algorithms, and editorial decisions. Access to trustworthy information is difficult to find for the average person. What if there were another signal, one generated not by publishers, influencers, or institutions, but by thousands of people putting real skin in the game? That is the question prediction markets could tackle. Could people have agents that learn and grow with the prediction market participant?

The moreLikely project aims to address the informational accessibility that prediction markets could provide. There is a need for tools that enhance the user experience. Currently, a user enters the market looking for an opportunity to predict. What they choose will vary based on the individual. Even if they are confident in a prediction, the user may do some base-level research to reassure their assertions. That research is difficult as a user consumes content from multiple sources, with each content varying in its usefulness. Some users may use AI tools like ChatGPT by simply asking the questions the user believes is needed to reassure their original beliefs.

Fragmented Information

For most of human history, understanding public opinion was impossible. People formed beliefs based on conversations with friends, local communities, and whatever information happened to reach them. Then came newspapers, radio, television, and eventually social media. Each innovation expanded access to information, but it also concentrated influence into the hands of those who controlled distribution.

Today, there is an abundance of reachable information. Yet many people feel less certain about what the public actually believes than ever before. Headlines compete for attention by reconstructing the truth in the hopes of growing user engagement. Regardless of where someone falls politically, economically, or culturally, there is a growing sense that many narratives arrive pre-packaged before individuals have had a chance to form their own conclusions.

Multiple sources can produce noise within the same timeframe on the same topics or stories. Numerous narratives for the user to consume and evaluate. That noise is attention grabbing but does not improve understanding of the information. This is an important factor when evaluating prediction markets. If a person creates a post, “ The New York Knicks will win the NBA Finals”, there is little to no cost of that person being wrong. With a prediction market, “ The New York Knicks have a 25% chance of winning the NBA Finals.”, this is backed by capital and can be tested against reality. Noise can be true, entertaining, emotional, or important. It’s just not necessarily useful for understanding what’s actually happening.

With the help of AI intelligence, prediction markets unlock a new informational power for the average user. Helping them gain incentive-aligned information to anticipate future outcomes confidently. Markets create signal because participants have incentives. With the addition of AI-powered tools, users can make predictions with greater confidence, identify new opportunities, and discover emerging markets.

Markets are the new Trustworthy Information Source

Unlike polls, opinions, or editorial analysis, prediction markets require participants to put capital behind their beliefs. Every market price becomes a real-time reflection of what thousands of participants collectively believe is likely to happen. The distinction is not that the market signal is the truth. The market could still be wrong, just as the local news or the experts could be wrong. Tools that improve signal, meaning information that increases our ability to understand or predict reality, are a critical part of the process.

Beyond the Market Price

Prediction market raw data can be difficult to interpret. The average user does not have time to monitor hundreds of markets, track sentiment shifts, identify emerging trends, or understand why probabilities within a market are shifting. This could deter the next wave of prediction market users from participating.

For example, instead of simply displaying prices, intelligent tools can provide deeper insights:

• What markets should be a priority for this user?
• Which narratives are gaining momentum?
• Where are markets diverging from media coverage?
• Which topics are seeing unusual attention?

Information Sovereignty in Practice

Information sovereignty does not mean rejecting the latest analysis from well-known experts. It does not mean ignoring journalism. It means having access to independent signals that help people evaluate information for themselves.

It means providing an environment where the user can grow and understand before taking action. When individuals can compare the right headlines against market expectations or observe shifts in collective sentiment, they gain a different insight that encourages placing prediction bets. In an environment where narratives compete for attention, access to transparent market signals provides a valuable counterbalance.

The mission is simple:

People make better decisions when they have access to better signals. Prediction markets contain a remarkable amount of collective intelligence, but that intelligence is often buried beneath thousands of contracts, price movements, and fragmented data points.

So join us as we embark on this new journey. Explore the first experimental approach, an open-source recommendation tool. The tool aims to improve how users find the most relevant prediction markets. Instead of scrolling through hundreds of irrelevant markets, this tool uses large language models, LLMs, to improve the recommendation list. The more the user interacts with the list, the more information the LLM can use to improve the next list.

This tool was created for the Somnia Agentathon. An agent-focused hackathon focused on building applications on the new agenetic blockchain layer. This experimental implementation recommendation tool may be used with different LLM providers, but utilizing an AI agent on the Somnia blockchain adds a few benefits.

The user can regain their informational sovereignty with transparent agents. On Somnia, agents are auditable as their performance is accessible on-chain. Introducing other smart contract standards to the flow of the agent interaction, like ERC8004 agent framework, showcasing the growing benefits of using the agentic blockchain solution. Agents need a learning environment, so the blockchain makes the relationship between the user and the agent occur in a controlled, transparent environment.

The future belongs to individuals who can separate signal from noise. Prediction markets won’t replace traditional information sources. They will provide something increasingly valuable, a transparent, incentive-driven reflection of collective belief. In a world saturated with opinion, that may be one of the most powerful tools available for maintaining independent judgment for the broader population.