

长安(Web3版)🌊RIVER
1.9K posts

@changan_web3
没啥能力,全靠机遇,所有的内容仅是个人观点不构成任何投资建议仅此一个账号,其余均为假冒谨防受骗!



been at Abu Dhabi Finance Week and catching up on some of the views around InfoFi as always, appreciate all the observations and suggestions - the public discourse and feedback is how we learn and improve below are some of my latest thoughts, with perspective from also meeting with top web2 brands recently - globally, distribution is increasingly decentralized - shifting from pay-for-platform to pay-for-creators, with a particular focus on user-creator overlap. Whop invented clipping-to-earn and SideShift is powering UGC content creation - both pure web2 US platforms, with a strong foothold on US campuses, and collectively powering billion dollar+ volumes annually. so, there is a mega trend - the rise of AI has reduced the marginal cost of content to basically 0. all social media platforms are now ramping up anti-bot mechanism - can see AI ran content across all platforms, most notably LinkedIn, X and YouTube. Even on TikTok though which overall does a decent job, the comments section is flooded with slop, often simply summarizing the original video clip. so, there are obviously challenges - marketing in crypto remains highly inefficient. before, it ran mainly on agency-KOL marketing. today it’s a hybrid - sometimes a dual track. the birth of InfoFi has unequivocally provided so many average user-creators their foothold into the space and an opportunity to get exposure while learning/using/sharing about the protocol. when I was in Korea earlier this year (Kaito’s no 1 market actually), I was surprised there were so many part-time professional workers (lawyers, bankers etc) using Kaito as their side hustle, as their way of learning about crypto and connecting with like-minded friends. hence why, since a few months ago, we introduced onchain multipliers, rewarding not just content creations, but also onchain usage of the protocols, as a shift towards a user-evangelist model. today we are working closely with many of the innovative apps in the space to bring their apps to users globally. what I think isn’t working well, compared with our web2 counterparts - I do think content creation needs a minimum threshold on either social influence or onchain usage. in web2 this was done through vetting and interviews. in web3, perhaps because of the ethos of being full permissionless, some of the teams opt for no threshold at all. going forward, we will work towards a minimum level of reputation or onchain usage to qualify. in fact, uniquely in web3, we actually have a ton of data, including at the wallet level. one metric that we will soon introduce is called “follower net worth” - frankly it’s something that’s impossible to fake. looking around, it is clear that Kaito today has the most strict anti-bot mechanism, and we encourage other platforms to be equally strict: - replies don’t count - reputation threshold for gating - strict nerfs on engagement farming and group farming - weekly cap on contributing tweets to disincentivize volume and many more we already open-sourced our reputation protocol - anyone is able to freely integrate. lastly, I do want to say that in this global shift towards UGC content and the mass proliferation of AI content, there is a truly unique opportunity for crypto. as Balaji said - “AI makes everything fake, and crypto makes it real again” truth is with wallet provenance, it provides the best way to tell the holistic profile of a particular account. In fact, today we know way better than X in filtering out bots for the exact same reason. I believe there’s an opportunity for crypto to play the role of telling what’s real vs fake on global social media at the end of the day, if decentralized social isn’t the answer, maybe it’s crypto doing what crypto does the best - provenance - in the inevitable future of the world of social media AI onwards








关注几个机会:确定型机会,国家和时代的红利 宇树科技(Unitree Robotics) 成立于 2016 年杭州,主打高性能四足/人形机器人,自研电机、减速器、控制器等核心部件。 2023 年全球四足机器人出货占比达 69.75%,累计出货超 5 万台。年营收超 10 亿,已连续五年盈利。 C 轮融资超 10 亿,估值约 120 亿元。 2025 年 7 月启动 IPO 辅导,预计 2026 年上半年上市,目标「人形机器人第一股」。 长鑫存储(CXMT) 成立于 2016 年合肥,主营动态随机存储器(DRAM)芯片,是国内首家推出自主 LPDDR5 的企业。 2024 年营收超 10 亿美元,融资 108 亿元,估值约 1399 亿元。 股东包括合肥国资、阿里、兆易创新等。 预计 2025 年出货量增长 50%,市占率升至 8%。 已启动 IPO 辅导,最快 2026 年登陆科创板,冲击“存储芯片第一股”。 摩尔线程(Moore Threads) 成立于 2020 年北京,专注 GPU 芯片设计,是国内首个支持 DirectX12 的 GPU 厂商。 产品覆盖 AI 训练、科学计算、图形渲染等领域。 2024 年融资超 50 亿元,估值约 255 亿元,背后投资方包括红杉、腾讯、字节等。 2024 年 11 月启动 IPO 辅导,预计 2026 年上市,目标“国产 GPU 第一股”。 众擎科技(EngineAI) 成立于 2023 年深圳,聚焦通用人形机器人与智能解决方案。 团队来自小鹏机器人、UC Berkeley、清华等。 2024 年推出首款双足机器人,完成近 1 亿元天使轮融资。 当前仍处早期阶段,暂无 IPO 计划,重点在产品迭代与市场落地。 宇树科技、长鑫存储、摩尔线程 已正式启动 A 股 IPO 辅导流程,预计 2026 年上市,代表机器人、存储芯片与 GPU 三大“国产替代”方向。 众擎科技 仍处早期融资阶段,短期内暂无上市计划,但具备潜力。 这些公司均受益于国家 “双碳 + AI + 半导体” 政策红利。

@Bitget_zh 这是可以说的吗 🫢






伊麗莎白「補捐」事件據說又有反轉了! 要是真的是 24 小時撤回、或模擬器假打款… 那這慈善就不是給受害者的,是來驗證大眾智商的 😂 既然她人目前也在迪拜, 不如幣安直接來一場「現場驗證」: 錢轉哪、怎麼轉、有沒有 100% 轉出去? 一鏡到底,至少也得對那一百多條人命有個交代吧? 現在看下來——玩套路、玩造假, 在座的各位在傻白姐面前都是弟弟 😂 @abyssofgambling @czreth @heyibinance @cz_binance @binance





直到现在 我依然不太相信是P图 我一直期待的反杀怎么还没出现 🙃🙃🙃