From Brackets to Markets

How AI is helping people turn niche knowledge into economic edge

-- Justin Patel, Venture Investor at Decasonic

March Madness is the right hook, but it is not the whole story

March Madness works as a hook because it compresses a behavior that already exists across the internet. For a few weeks every year, millions of people convince themselves they see something the market does not. For me, that usually starts with believing Michigan can win the tournament. From there, every injury update, coaching mismatch, and bracket path starts to look like a potential edge. That instinct has always been there. What is changing is the product surface around it. This year, fresh off a $22 billion valuation, Kalshi introduced a $1 billion perfect bracket challenge that took a familiar ritual (March Madness brackets) and pushed it further into a probabilistic, market-adjacent experience. At the same time, prediction markets are drawing more mainstream attention and more scrutiny, including new sports-integrity rules and legislative pressure around sports-related contracts.

That timing is what makes March Madness useful here. It gives people an intuitive on-ramp into a bigger shift. The real story beyond bracket culture is that prediction markets are becoming a more credible interface for acting on judgment, and AI is starting to help users turn niche knowledge into something that is structured, persistent, and economically useful. That sits directly inside Decasonic’s recent work on personal superintelligence and durable adoption in prediction markets.

March Madness just makes the shift easy to see. The bracket is still the wrapper, but the behavior underneath it is moving from casual opinion toward priced conviction. Pricing conviction enables AI to drive at context layer improvements, an area of focus for our Q2 2026 internal AI but also deal sourcing.

Brackets were always social. Markets are economic.

The difference between a bracket and a prediction market is not just product design. It is a shift in what the user is doing.

A bracket is mostly a static expression of belief. You make your picks, share them, talk trash with your friends, and then watch the results unfold. The value is social. The participation is episodic. The feedback loop is short, but it does not really compound.

You can be right, but the system does not do much with that information after the fact.

A market behaves differently. It updates as information changes. It absorbs new evidence. It lets conviction get stronger, weaker, or more nuanced over time. It also gives the user a direct mechanism to express that change economically rather than only socially. That may sound obvious, but it matters because it changes prediction from content into action.

This is why I think prediction markets are more important than the usual “it is just sports betting with better branding” critique suggests. The category is not interesting because it digitizes a “bet” with a binary view of the future developed by the crowd, rather than a bet against a sportsbook. It is interesting because it creates an interface where fragmented human judgment can be priced, compared, and acted on in real time. That is a much bigger product category.

Our recent blog on durable adoption made a similar point in a different way: in prediction markets, tentpole moments generate attention, but what matters is whether platforms convert that attention into repeat behavior, cleaner resolution, and deeper user habit over time. In other words, the question is not whether a big event can drive volume. The question is whether a user comes back because the market has become useful.

That distinction becomes even more important once AI enters the picture.

AI changes the value of niche knowledge

For most of the internet era, niche knowledge was hard to scale. Someone might know college basketball unusually well. Someone else might track crypto-native sentiment better than mainstream research desks. Someone else might understand the cultural undercurrents around a creator, a product launch, or a policy shift before the rest of the market catches up. Those edges were real, but they were difficult to operationalize consistently.

Usually they lived in someone’s head, or at best in a messy stack of tabs, feeds, spreadsheets, and group chats.

AI changes that in a meaningful way.

The personal superintelligence framing I wrote about earlier this month was really about this: the next layer of AI value does not come only from stronger models. It comes from systems that get closer to a user’s actual workflow, context, memory, and intent. When AI can sit closer to what a user tracks repeatedly, what they have been right about before, what information they care about, and how they prefer to act, it starts feeling like leverage.

That matters a lot in prediction markets. AI can help a user monitor the exact information streams that matter to their edge. It can summarize changes against an existing prior rather than from scratch. It can surface when the market has moved meaningfully without the underlying facts changing much. It can track where a user historically has strong read-through and where they do not. It can help turn instinct into process.

Just as importantly, prediction markets themselves can improve how both humans and AI manage future events. A good prediction market is not only a place to express a view. It is also a live coordination system around uncertainty. It aggregates scattered information, forces clearer probability estimates, updates as new evidence arrives, and creates a feedback loop around whether a forecast was actually right. That makes it useful not just for traders, but for any system, human or machine, trying to reason more clearly about what is likely to happen next.

That is where the connection to personal superintelligence becomes much more direct.

Earlier internet systems were largely social. They helped people broadcast opinions, react to information, and participate in public discourse. Personal superintelligence is a shift away from that model and toward systems that help an individual think better, remember more, and act with greater precision on what they uniquely know. Prediction markets fit that shift especially well because they sit at the point where private judgment meets live economic expression.

This is why prediction markets may end up being one of the clearest applications of personal superintelligence. The user brings differentiated context. AI helps organize, monitor, and sharpen that context. The market provides a live mechanism for testing and expressing it. Over time, that creates something more powerful than a feed, a chat thread, or a one-off take. It creates a system where knowledge can be refined, measured, and acted on repeatedly.

In that world, the value of AI is not that it predicts the future in some magical way. The value is that it helps users make better use of what they already uniquely know, while prediction markets provide the structure that turns that knowledge into a more disciplined view of the future.

The real opportunity is bigger than sports

Sports are a convenient entry point because these are events that are mainstream, passionate, legible and frequent. Everyone understands what it means to have a read on a game, a team, or a tournament. But the deeper opportunity is broader.

Prediction markets sit naturally at the intersection of politics, macro, culture, crypto, product launches, regulation, sports, and public narrative. Anywhere information is changing quickly and conviction is unevenly distributed, there is room for a market. The appeal is not that people like to “bet.” The appeal is that they want a live system for expressing uncertainty.

That is part of why mainstream institutions are moving closer to the category even as regulators and lawmakers push back. Last week, MLB signed a multi-year deal with Polymarket, making it the league’s official prediction market exchange while also pairing the deal with an integrity framework and information-sharing with the CFTC. In parallel, lawmakers are trying to curb sports-related prediction contracts, and platforms such as Kalshi and Polymarket have tightened anti-insider-trading rules. The category is not fringe anymore; it is now important enough that leagues, regulators, and incumbent gambling interests all want to shape its boundaries.

That is usually what happens when a product category begins to matter.

The important point for builders and investors is that prediction markets are moving beyond a narrow “sports or politics” frame. They are becoming a broader probability interface. In time, I think the most valuable platforms will be the ones that make it easier for users to carry their knowledge across categories instead of treating every event as a one-off interaction.

That is where the AI layer becomes structural.

The next winning product will not just list markets

Most early debates around prediction markets were framed too narrowly. Which platform has more volume? Which one has better regulation? Which one is more mainstream? Which one is more crypto-native?

Those questions matter, but they are not the only ones that matter.

The more important question is what the winning product experience looks like once the category matures.

I do not think the long-term winner is simply “the exchange with the most contracts.” I think the winner is more likely to be the platform that helps users understand where they actually have edge, reduces the friction around creating and acting on that edge, and makes the whole experience feel closer to a modern internet product than a niche financial tool.

That is why the “personal terminal” framing is useful.

For institutions, the terminal historically mattered because it compressed information, context, and action into one surface. For consumers and independent traders, that surface has been more fragmented. You have feeds, chats, dashboards, screenshots, newsletters, spreadsheets, X lists, Discord servers, and alerts, but very little that ties them together into a coherent action layer.

Prediction markets plus AI can change that. The natural end-state is not just a website where odds move around. It is a product that knows what domains you care about, remembers your priors, tracks the markets most relevant to your expertise, helps you create or discover the right contracts, and increasingly lets software participate alongside you.

That is a very different ambition than “come here to place a bet.”

It is also the reason I think the category will split less along today’s labels and more along product philosophy. Some platforms will skew toward regulated mainstream access. Some will skew toward attention and media virality. The most interesting platforms will push further and try to become a real system for turning opinions into markets and markets into durable user behavior.

What product leadership in prediction markets will look like

The next phase of prediction markets will not be won simply by listing more contracts or capturing short-term volume around tentpole moments. It will be won by the platforms that make markets deeper, broader, easier to access, and more native to how people already consume and act on information online.

That is the more important product question now. As the category matures, leadership will come less from novelty and more from whether a platform can support repeat participation, cleaner market resolution, richer interfaces, and better distribution.

Recent roadmap posts from Opinion offer a useful example of what that product direction can look like. Rather than treating prediction markets as a single destination for event trading, the company frames them more as infrastructure for what it calls the “Multiplayer Internet”: a system that can turn economic, cultural, and local insight into tradable markets at scale.

That framing is notable because it points toward a broader view of mainstream adoption. A leading platform in this category likely needs to do several things at once.

First, it needs deeper liquidity and stronger market quality. That includes mechanisms such as maker rebates, better incentives for participation, and cleaner dispute-resolution systems that preserve trust as the number of live markets scales. Growth in prediction markets is not just about launching more markets. It is about making sure those markets remain reliable and usable as supply expands.

Second, it needs a much better user experience around context. Sports and esports flows, tournament pages, schedule context, AI-assisted analysis, and pre-market auction mechanics all point to a broader idea: users do not just want a list of markets. They want an interface that helps them understand where they may actually have edge and act on it more naturally.

Third, the next generation of products likely cannot remain trapped inside a single destination site. Through initiatives like Builder Keys, Opinion is pointing toward a more embedded model in which markets can live across sports apps, media products, community sites, and other digital surfaces where context already exists. That feels directionally important. If prediction markets are going to become a real interface for applied judgment, they should be able to travel to wherever information, attention, and intent are already concentrated.

Finally, the category will likely expand beyond simple binary event contracts. Opinion’s framing around “omni-markets”, including automated markets across sports, digital assets, and esports, along with more granular market structures, points toward a future where users have more precise ways to express differentiated views. That matters because the long-term upside here is not just enabling people to express more nuanced judgment across a much wider range of outcomes.

This is also where the personal superintelligence theme becomes more tangible. If AI is increasingly helping users organize context, track fast-changing information, and refine their own judgment, then the natural complement is a market layer capable of absorbing that judgment. In that world, the most important platforms are not just marketplaces for speculation. They are systems that help users turn niche knowledge into tradable expression across sports, macro, crypto, culture, and other domains where edge is fragmented and constantly updating.

That is the broader product shift to watch. The category leaders will likely be the ones that combine liquidity, trust, context, distribution, and AI-native participation in a way that makes prediction markets feel less like a niche corner of finance and more like a mainstream interface for applied judgment.

Where this goes from here

The reason I like using March Madness as a starting point is that it makes the demand side obvious. People already want to test whether they have an edge. They already want fast feedback loops around judgment. They already like competing on conviction. They already spend an enormous amount of time turning fragmented information into takes.

What has been missing is the product layer that turns that behavior into something persistent and economically useful.

That is the bigger opportunity in prediction markets. Not simply more event contracts or more speculation, but better products for applied judgment.

From here, product leadership will matter more than novelty. The platforms that lead this category will be the ones that make markets easier to understand, easier to trust, and easier to participate in repeatedly. They will help users identify where they actually have edge, give them better ways to express that edge, and integrate more naturally into the workflows, media surfaces, and digital environments where conviction is already formed.

This is also where the personal superintelligence theme starts to matter in practice. As AI gets better at organizing context, tracking fast-changing information, and helping users act on what they know, niche intelligence becomes more valuable. The person who follows one corner of the world unusually closely can do more with that than post about it. They can build a repeatable system around it. Over time, the advantage shifts from simply having information to having better tools for compounding it.

Prediction markets are one of the first places where that shift can become visible at scale. The category leaders will not just be exchanges with more contracts. They will be products that combine liquidity, trust, context, and AI-native participation in a way that makes prediction markets feel less like a niche financial tool and more like a mainstream interface for acting on judgment.

That is the direction we think this market is moving in.

Disclosure: Decasonic is an investor in Opinion Labs. This article is for informational purposes only and is not investment advice.

The content of these blog posts is strictly for informational and educational purposes and is not intended as investment advice, or as a recommendation or solicitation to buy or sell any asset. Nothing herein should be considered legal or tax advice. You should consult your own professional advisor before making any financial decision. Decasonic makes no warranties regarding the accuracy, completeness, or reliability of the content in these blog posts. The opinions expressed are those of the authors and do not necessarily reflect the views of Decasonic. Decasonic disclaims liability for any errors or omissions in these blog posts and for any actions taken based on the information provided.