Group 1 - The rapid advancement of AI technology, exemplified by the release of GPT-5, contrasts with the slow adaptation in real-world applications, highlighting significant barriers to practical implementation [2][3][5] - The quality of official economic data in the U.S. is deteriorating, which raises concerns about the reliability of metrics that influence critical economic decisions such as interest rates and social security [5][9] - The Consumer Price Index (CPI) data collection methods have become increasingly reliant on estimations, with 35% of the data being approximated rather than directly collected, indicating potential inaccuracies in economic reporting [6][8][9] Group 2 - The recent non-farm payroll (NFP) report showed a disappointing job growth of only 73,000 positions in July, with significant downward revisions for previous months, marking one of the worst performances since the pandemic [9][10] - The Bureau of Labor Statistics (BLS) continues to use outdated methods for data collection, which introduces various uncontrollable factors that can skew results, such as response rates and the accuracy of reporting by businesses [11] - There is a growing trend in the U.K. towards automated data collection for CPI, utilizing supermarket scanner data to enhance accuracy and coverage, suggesting a potential path forward for improving economic data collection methods [15]

2. 一切都在发展变化中，一些"结论"都是标题党：面对"百大变局"，大家的处境并不一样。这个颗粒度 细分的世界越来越碎片化。 比如我们举三个典型代表： 从它们身上可以得到几个观察： 首先没那么悲观。原因有三： 1. 大家心态都还算稳：不能说完全不焦虑，但都在积极想办法应对。前几个月抢出口，现在抓紧时 间布局本地化。 01 最近一个多月，我一路小跑，从新疆到江浙到大湾区，从国企到民企，从橡胶到袜子到发电机，调研了 很多制造业企业。重点关注大家的关税应对和出海策略，特别是那些早有危机感、甚至在贸易战1.0阶 段就已经走出去的"先知们"。信息量大，心情复杂。 怎么说呢？如果快问快答，能立刻脱口而出的有两个感觉： 1. 企业家对宏观的敏感度完胜经济学家。身处一线，必须时刻看路，生死攸关是感受在皮肤上的。 1. 美国市场不可能放弃：几乎所有企业都表示不可能割舍。利润高、订单大、消费能力强、又是最 大的"全国统一大市场"——迄今为止依然是全世界最好的买家。 2. 不能低估产业链转移的可能性：虽然中国目前产业链最完整、协同效率最高，但并不能排除假以 时日（可能5～10年，甚至借助中国经验会更快）也能成形。也不排除事情会向完 ...