functionSPACE

functionSPACE is an open-source protocol and developer framework for prediction markets. Every market is self-liquid out of the box — no need to source market makers or bootstrap liquidity. You define the market, integrate it into your app, and it just works. Any numerical outcome can become a market, for example: - NBA game totals - ETH liquidation volumes - quarterly GDP growth - daily high temperatures - Apple earnings Offer your users binary questions, price ranges, complex structures all share the same underlying surface. You build the experience; the protocol handles the rest.

@0xtarzan1999 Open systems will win, you don't need $500m of liquidity in a market to get an accurate price ...cheaper open markets > gated API's - add the AI internet in the mix and the economics aren't there at all

@functionspaceHQ When the main product is information, why don't they build a product around it and make extra money?

Kalshi has ~10,000 live markets. They want 50,000–100,000. Interestingly they say, "95% of their MM liquidity comes from ~2,000 individuals, not institutional market makers." People with niche expertise. An Ariana Grande superfan, a Kansas based inflation forecaster i.e. Belief-driven participants. What if the liquidity model was rebuilt around how these people actually think? Pair that with un-gated market creation and distribution channels and 100k markets starts to look like the floor not the ceiling. 5 things we find particularly insightful from the Cheeky Pint episode: 1. "80% of Kalshi users don't trade." They check the odds. The primary product is information, not the trading activity. 2. The best inflation forecaster on Kalshi is a guy in Kansas who never traded before. Dispersed private knowledge finding a price - Hayek in practice. 3. Kalshi explicitly wants to move beyond binary. Futures, swaps, options are on the roadmap. Binary is the starting point, not the end state. 4. Tarek cites the "infinite markets" thesis. As society gets more complex, you need markets on the individual factors driving outcomes, not just the aggregate. 5. Long-tail liquidity is how they think about moat. Head markets attract makers organically. The long tail needs something structurally different.

We shouldn't sleep on this, very bullish for embedded prediction markets as distribution pathway

Here's who to follow if you want actually good takes on prediction markets: @VitalikButerin @DaedalusRsch @probaaron @aradtski @functionspaceHQ @TideMarkets @ZEITFinance @Jake_Nyquist @Nekt_0 @Ranger_Global @sybilpm @nic__carter @jeff__opinion @masonnystrom @RohOnChain @aaronjmars @klos @voliti_co @gemchange_ltd @hyperstiti0ns @benfielding @semajeth @0xjacobzhao @0xnagu @SolSt1ne @mikhryc0x @mikita_crypto @marvellousdefi @theoddchain @curiouscamilo @linfluence @jolimmmm @outpxce @socialgraphvc @DougieDeLuca @BetBreakNews @dbarabander @sidbharth @smaaaaliy @thejay @trashman18 @drNeyo @13_niakris @Baheet @a1research__ @akshayraj_v0 @reachavci @njokuScript

Join our early builders program: builders.functionspace.dev

functionSPACE is an open-source protocol and developer framework for prediction markets. Every market is self-liquid out of the box — no need to source market makers or bootstrap liquidity. You define the market, integrate it into your app, and it just works. Any numerical outcome can become a market, for example: - NBA game totals - ETH liquidation volumes - quarterly GDP growth - daily high temperatures - Apple earnings Offer your users binary questions, price ranges, complex structures all share the same underlying surface. You build the experience; the protocol handles the rest.

@michael_lwy @0xsmac 10/10 - recommend

@0xsmac so many problems onprediction.xyz

the reason people misunderstand prediction markets is because they appear simple on the surface but in reality have so many uniquely difficult problems and if you understand all but one of these problems you can convince yourself they’re going to take over the world

Want to build a binary UX while leveraging the value of a unified liquidity contract? Use math to determine the right ranges don't guess up front OR let users choose their own range your choice, just drop in the Binary Market Kit

@0xtarzan1999 interesting - what do you see as the common factors between pm traders and perp traders?

Prediction Markets are live on Hyperliquid testnet. I told you they are coming for PMs. Most of the Prediction Markets platforms are trying to build from scratch and onboard traders, mostly the same group of traders common to Polymarket, Kalshi, Opinion Labs, Predict Fun, Limitless, Myriad, etc. But when you are building something new, your leverage= attention + distribution. Hyperliquid has tremendous attention and distribution. They have perp traders, who can easily try Prediction Markets.

@StockMKTNewz why is that?

Its kind of interesting to me that prediction markets were picked up by retail investors before institutional investors

@thenarrator what are some oracle solutions you are seeing which can actually achieve this?

the hyperliquid playbook, mapped to prediction markets: 1. build the best infrastructure first hyperliquid built a custom L1 with sub-second finality and a full onchain order book it wasn't just another perp dex, it was structurally better than everything else. the pm equivalent: build a matching engine and resolution system that's genuinely faster, cheaper, and more reliable than polymarket or kalshi. 2. be first to list, always hyperliquid listed every trending token before anyone else. when a token was pumping on ct, hyperliquid already had the perp live. traders went there because that's where the market existed when they needed it. the pm equivalent: when news breaks, your market should be live within minutes, not days. 3. no vc, return everything to users users own the platform. the pm equivalent: don't build for investors, build for traders return fees to the community through rewards, accuracy-based incentives, or direct revenue share the platforms that optimize for investor returns lose to the ones that optimize for trader retention. 4. protocol-owned liquidity solves cold start (only if you have the capital) hyperliquid built HLP, a protocol-owned vault that actively market makes across every perp. it handled 40% of volume and offered 17% APY to depositors with no reliance on external market makers. the pm equivalent: build a protocol-owned liquidity vault that seeds every new market automatically. when a market goes live on a breaking story, the vault provides initial depth. users arrive to a market that already has liquidity instead of an empty order book. 5. composability attracts builders hyperliquid launched HyperEVM and gave developers direct access to the trading layer through precompiles. hundreds of teams started building on top. the ecosystem created its own gravity. the pm equivalent: open apis, public data feeds, embeddable markets, developer tools. let anyone build a terminal, a news app, an alert system, a bot on top of your markets. polymarket understood this and they became the backend for metamask, trust wallet, jupiter. 6. community-first distribution hyperliquid grew from 300k to 1.4m users in 2025 with zero paid marketing. dozens of regional communities, hundreds of teams building, grassroots adoption fueled by aligned incentives. the pm equivalent: let your best predictors become your growth engine. every correct prediction is a distribution event and every top predictor with a following is an unpaid ambassador. 7. fee structure as a competitive weapon hyperliquid's fee-to-buyback loop created a virtuous cycle with more volume, more fees, more buybacks, higher token price, more users. the pm equivalent: dynamic fees that reward makers and scale with uncertainty. return fees to the community or use them to deepen liquidity.

@kelxyz_ Or simply - build on functionSPACE

A prediction markets team could break the Polymarket/Kalshi duopoly by: - Curating the experience around the top 1% of high impact macro markets: elections, politics, war, economics - 10 - 100x the liquidity on those markets - Bring leverage into the market

This report is a great signal of whats next in our space that you wont see from mainstream research. Our contribution centres on the core idea for resolving markets. We'd love to see more discussion and innovation in one of the hardest problems in prediction markets. Shout out to the @predictionindex team for including us. @justigor

@thenarrator when everyone has a browser agent or extension, does the depth of information become even more important?

someone is going to build a browser extension that overlays prediction market probabilities on every news article you read headline says "recession likely" the extension shows the market has it at 23%. to me this looks like an entirely new trust layer for the internet.

@blockchainbrett agree, the line between builder/creator is getting thinner every day. We're building the dev tooling for them to be able to do this

maybe creator owned markets isnt neccesarily the creator making the market about themselves but curating other markets they believe in for their audiences to take positions alongside them maybe not meant for the artist, but for the creators that share their worldviews

@robinmarketsxyz @defiance_cr @0xFudrick Amazing read, favourite part "if you want to hear about trading, you have to start following people who have like a hundred followers on Twitter" - looking forward to more of these!

Episode 2 of Prediction Market Wizards together with @defiance_cr is now transcribed and live on our Substack. A lot of fantastic insights to start your week with: open.substack.com/pub/robinmarke…

In functionSPACE, predictions are represented as vectors over a probability distribution. They are information equity, not yes/no tokens. Because they exist on a shared surface, they can interact, cancel, offset, aggregate, and be traded as continuous assets.