Hart Lambur

"9 people can control enough voting power on UMA to actually decide the outcome — which is not quite as decentralized as I would imagine many people would want. And this is just another risk as it relates to prediction markets." — Steve ⚠️ https://x.com/i/broadcasts/1dGYllboLkYKX?s=20

很好的思考。借着你思路。在考虑一个分层的复合型思路。 主流市场，占 80%，依然采用标准 UMA Optimistic + DVM。保持速度和低成本； 中高价值或高争议风险的，则动态上调 bond（按 volume / open interest / 潜在 payout 放大），甚至引入 KYC/白名单机制给 proposer 和 disputer（降低匿名攻击），此外为了保险，还可以并行验证（另一个 oracle 或 AI 初步判断）。 最后是超级大额的，远超 UMA 经济安全阈值，比如单市场 volume 超 5000万或 open interest 极高的，可升级到有明确问责主体的

- migrating existing open markets (esp high volume ones with multiple Bs at stake) seem very painful - positions already tokenized against the UMA adapter - unsure if can manually migrate liquidity/positions to new markets - changing the entire market-creation pipeline + getting users to move would

Polymarket 空投玩家现在已经卷出新高度： 不仅天天交易，还在从各个维度完善自己的账号—— 抵押 bond 去挑战提案， 质押 UMA 抢争议票权， 把“社区参与权重”玩成了真治理。 这已经不是单纯预测了，这是真金白银的 Skin in the Game。 而 predict 那边好像就简单多了： 测试各种策略拿分就行，没那么多麻烦。就是策略迭代太快，一公布就卷废。 今早 o1 也出新情况了—— 同分情况下突然“不同权”，临时加了一堆额外条件。 （虽然对 o1 特殊情况来说挺科学的，但也暴露了纯积分制的软肋） 盲刷、积分、积分加条件，3种模式，你更喜欢哪一种？

到底是 SEC 的文件更权威还是 Polymarket 更权威呢？Polymarket 的审核流程就是 discord 开展辩论，交给一个市值仅有 5500 万美金的加密货币项目 UMA 的大户进行投票 也就是说，Polymarket 的安全性和结果公正性，一个目前估值 120 亿的项目，承载了上亿交易量和对应结果的预测市场，结果的公正性仅由市值五千多万的项目的少数集中持有者来维护 目前，已经有多起 Polymarket 篡改事实，结果与客观事实不符的判定出现，而这起事件被广受关注、很显然 PM 官方不占理的原因在于 1. SEC 文件的客观性和事实权威性，远比其他能够主观理解的事件更高

Kalshi, Polymarket and the UMA whales will make for a great story one day.

GPU proving. Multi-asset fee payments. A wallet. Developer tooling. 34 audit findings closed. Our partners @Nethermind released their Q1 report last week, covering six active workstreams across protocol, research, and tooling. Read the full report here: https://t.co/C9whkvi3km

polymarket will sign their grave if they allow the UMA cabal to vote NO on this unreal incompetence

Looks like @Polymarket is cooking in silence 👀🧐 a way better UMA resolver is coming and they are testing on testnet currently three things stood out 👇 1. one resolver handles everything, single outcome, multi-outcome, all natively. cleaner architecture, less surface area 2. createPublicUmaRe

Direct Capital Flows