Entity
1.6K posts







OpenAI开源Symphony:把项目需求直接转化为AI自动化交付流 这是软件产品的需求自动化实现了,在类似Sprint的看板里添加需求后,Agent自动认领需求,开启完全隔离的独立开发流程并更新需求的进度,开发完成后AI有“自证”的步骤,包括测试、功能演示、代码Review、开发文档,最后由人类验收无误后合并项目。 对比之前软件开发的人类团队,这个流程用Agent替代了开发和测试,人类主要负责提出需求、设计系统和架构、业务验收。这次AI不是服务于单个人了,而是直接介入整个项目的流程。目前看还是很大胆,但是长期看必定会替换项目里的绝大多数人类。 Github:github.com/openai/symphony




Crazy that @GiggleAcademy received more than $1m donations in the first 12 hours of the feature being online. Way beyond my expectations. Lots of good people out there. Lots of good crypto people. 🙏 90% of the donations so far came from a meme coin. I still don't get how a $5m market cap coin can generate $900k in fees in 12 hours for donations. I don't know who created it, or how the smart contract was coded. But thank you! 🙏 This will forever change my perception of meme coins. Meme coins have utility! 🤐 Giggle will put these donations to good use. They will be used to incentivize contributors and accelerate growth. They will not be use to pay employees. I will cover employee and server costs for the foreseeable future personally. @GiggleAcademy did not create a token, and will not create a token in the foreseeable future. I don't want to complicate a free education platform with tokenomics. I believe Giggle is perfectly capable of sustaining itself in the future (5-10 years later) by donations from Alumni networks who got (better) jobs because of the education Giggle provided, and the general public. Let's together make education free for all of our future generations, and make it high quality! Thank you, and onwards!

明显感觉到一个趋势: @Lighter_xyz 、@grvt_io 等一众layer2 Perp项目在拼命秀肌肉,颇有和追赶 @HyperliquidX 的架势。 但这明显有点“起了个大早却赶晚集”的尴尬,Perp DEX的繁荣本应该属于layer2才对,怎么又被 $ASTER 等其他Alt-L1给截胡了? 1)关键症结不复杂,Arbitrum、Optimism等layer2四大天王都志在搞通用Rollup,但这就决定了其TPS、出块时间的上线,因为交易Finality、状态同步、复杂计算开销等都会成为拖累,咋可能给Hyperliquid这种亚秒级别媲美CEX体验的产品较劲。 所以,一味追求General- Purpose是原罪,Lighter就很机智,不再纠结所谓通用化,从ZK电路和订单簿撮合等层面都做了专用交易优化,所以交易量才能达到媲美Hyperliquid的成绩。其他layer2-Based Perp DEX要想收复旧失地,放弃通用化设计是第一步; 2)大部分layer2 Perp还在死磕AMM那套老逻辑或者尝试混合Orderbook设计, @GMX_IO 作为曾经的layer2 Perp老大哥已经给出了验证,死磕AMM的天花板就在那。因为,无常损失、MEV攻击、滑点等问题直接影响了机构级资金的大单流入和高频交易,这在Hyperliquid的全链CLOB+HLP模式下几乎不存在; 所以,Layer2 Perp想要彻底翻身索性彻底拥抱原生CLOB架构,原先所谓的渐进式改良,dYdX的出逃不久已经证伪了? 3)Layer2的老一套DeFi挖矿激励已经失去了吸引力,像Hyperliquid这套新Perp玩家,除了底层交易性能哪些,关键是玩法多元,诸如,MEME文化代入、积分空投、代币回购等等,给了用户很大的体验改进和玩法预期; 何况, Layer2的流动性是严重分散的,大家用各种激励手段抢来去,直接增加了用户的跨链桥接成本且肥了一众做第三方跨链桥的项目,并不能直接带动其自身协议的用户粘性; 以上。 Layer2阵容的Perp DEX也着实不少,除了上边提到的项目,还有 @SynFuturesDefi @OrderlyNetwork 、 @paradex 等等一众项目一直在Build路上,GRVT最近还刚拿了19M融资,可见Layer2 Perp从来没放弃过抢夺话语权,真的,被Hyperliquid打趴下也就算了,至少和 @Aster_DEX 、 @SunPerp_DEX 等相比总得有一战之力吧? 别忘了,以太坊生态沉淀的巨量稳定币和DeFi TVL是其他L1所不具备的绝对性优势。




















