📈 Monday data: More AI numbers, more clarity?｜英文原文 + 中文翻译
Exponential View 2026 年 9 月 21 日周一数据刊：以 AI Investment Brief 创刊号摘录为核心，统计标普 500 公司在财报电话会上对 AI 给出的量化说法，看这些数字说明了什么、又回答不了什么；附完整英文原文与中文翻译。
原文信息
- 英文标题：📈 Monday data: More AI numbers, more clarity?。1
- 副标题：Companies are backing their AI claims with more numbers. What do these numbers actually tell us about AI's economic impact?。1
- 作者：Azeem Azhar、William Gildea（双作者署名）。1
- 发布时间：2026 年 9 月 21 日；官方 RSS 记录的绝对发布时间是北京时间 2026 年 9 月 21 日 21:28:07（13:28:07 GMT），页面最后更新于 21:28:21。2
- 本期性质：官方把这一期标为免费公开，正文可以一直读到刊尾。1
- 栏目与内容来源：这一期属于「Monday data」系列，正文是 Exponential View 新刊 AI Investment Brief 创刊号的摘录，作者在开场说明了下周恢复正常节奏的 Monday Data。1
- 公开范围：公开部分依次是开场四段、「An extract from AI Investment Brief #1」的出处标记、一节分析（五条公司引语、五张图表及对应解读）、以及刊尾的摘录声明和刊物推荐卡片，中间没有会员专享段落。1
- 正文图：五张图表都来自 Exponential View 官方页面，依次是财报电话会中给出 AI 数字的公司占比、成本／生产力与收入／需求两类声明、这两类声明的变化幅度、部署与试点的用词差异、以及三个 AI 词汇的提及率。1
- 封面图：本文封面是一张 AI 生成的空会议室图像，不是原文配图；原文页面用第一张图表的分享卡片作为社交预览图。1
English original
Companies are backing their AI claims with more numbers. What do these numbers actually tell us about AI's economic impact? 1
Today's Monday Data shares our take in an extract from the first edition of AI Investment Brief, our new publication that provides essential weekly analysis of the AI cycle. 1
The new publication gives us room to broaden Exponential View's coverage across technology, economics and society while keeping a close eye on the AI cycle. 1
We'll be back with regular Monday Data next week! 1
Azeem
An extract from AI Investment Brief #1 1
What corporate AI claims reveal – and leave unanswered
This week we explore the impact of generative AI on the wider corporate economy by analyzing earnings calls from S&P 500 companies and the claims that they make. The type of claims we've seen this quarter: 1
- $FDS: "overall ASV growth among clients using our AI solutions was 50% higher than for the rest of the book" 1
- $FIS: "On servicing, we've launched 5 Agentic programs with manual tickets down 70% and triage time down nearly 75%." 1
- $GE: "using AI to automate the process, we cut the number of demand signals in half and reduced processing time by nearly 90% across 190 parts" 1
- $MDT: "CathWorks, our AI and advanced computational science platform for angio-based FFR contributing nearly 300 basis points of organic growth" 1
- $WTW: "where we're using these tools for automated document reviews for new clients, system configuration time has gone down 60%" 1
33% of S&P 500 companies that held a call in the June 2026 season made a quantified statement about their use of AI, with 35% of calls in the quarter-to-date including quantified mentions, around 10pp higher than the same time last year. 1
15% this quarter have made a quantified claim of AI's impact on the business, up from 9% at the start of last year. Slow growth from a low base. 1
Together, these show that AI is steadily being adopted (and importantly, measured) in the wider economy, but it still sits at an extremely early stage (or, more bearishly, that most companies are not yet seeing measured AI results they can report to their shareholders). 1
Claims about the type of impact AI is having on businesses are rising: while we expect cost/productivity improvements to be first when implementing a new technology, claims about AI having a positive impact on revenue or demand have risen at a similar rate to 18%. 1
As well as being more prevalent, claims about cost and productivity impacts seem to be of a greater magnitude: the average claim this quarter has been of a 47% boost, vs 40% for revenue growth impacts (and that's over a wider cohort: 24% vs 18%). These averages sit within each other's interquartile range (i.e., there's a lot of spread and uncertainty baked into the average): take with a pinch of salt, and work with the ranges (20-60% for cost/productivity; 20-83% for revenue/demand). 1
Companies are increasingly using the language of deployment to discuss their AI initiatives (with a dip so far this quarter), rising from 7% at the start of 2025 to 12% this quarter. Pilot language stayed under 4% in every season. It's clear that even if companies are conducting pilots, they're not talking about them in calls. This will obfuscate attempts to understand how successful these pilots and investments are, and early signs of promise will only be mentioned in later periods. 1
"Agentic" appeared in 24% of calls this quarter so far (up from 9% in Q1 2025). 1
"Generative AI" fell from 13% to 5% over the same window. 1
"Copilot" is consistently infrequently mentioned (~3% of calls throughout). 1
This is an excerpt from the first edition of AI Investment Brief, a new publication by Exponential View. 1
中文翻译
公司开始用更多数字来支撑它们关于 AI 的说法。这些数字究竟告诉了我们什么，关于 AI 对经济的影响？1
新刊物让我们有余地，把 Exponential View 的报道面扩展到技术、经济与社会，同时继续紧盯 AI 周期。1
下周一我们恢复常规的 Monday Data！1
Azeem
What corporate AI claims reveal – and leave unanswered（企业对 AI 的说法说明了什么，又没回答什么）
本周我们分析标普 500 公司的财报电话会，以及它们在这些会上提出的说法，借此观察生成式 AI 对更广泛的企业经济产生了什么影响。这个季度我们看到的说法有这样几类：1
- $FDS：「使用我们 AI 解决方案的客户，整体 ASV（annual subscription value，年度订阅价值）增长比其余客户群高 50%。」1
- $FIS：「在服务业务上，我们上线了 5 个 agentic（智能体式）项目，人工工单下降 70%，分诊时间（triage time，把问题分流到对应处理环节所需的时间）下降近 75%。」1
- $GE：「用 AI 把这道流程自动化之后，我们把需求信号（demand signals）的数量砍掉了一半，190 个零部件的处理时间缩短了近 90%。」1
- $MDT：「CathWorks——我们用于基于造影的 FFR（血流储备分数）计算的 AI 与先进计算科学平台——贡献了近 300 个基点（basis point，1 个基点等于 0.01 个百分点）的有机增长。」1
- $WTW：「在为新客户做自动文档审查这件事上，系统配置时间下降了 60%。」1
在 2026 年 6 月这一季举行过电话会的标普 500 公司里，33% 对自身使用 AI 的情况给出了带数字的表述；在本季度至今的电话会中，含量化表述的比例是 35%，比去年同期高出约 10 个百分点。1
本季度有 15% 的公司给出了 AI 对业务影响的量化说法，去年年初这个比例是 9%。基数低，涨得也慢。1
两组数字合起来说明：AI 正在被更广泛的经济体稳步采用（而且重要的是，被测量了），但它仍处在极其早期的阶段。说得更悲观一些，就是大多数公司还没有看到可以拿去向股东汇报的、经过测量的 AI 成果。1
关于 AI 对企业产生了哪一类影响，相关说法的占比在上升：实施一项新技术时，成本与生产力的改善通常最先出现；而声称 AI 对收入或需求有正面影响的比例，上升速度与之相近，达到 18%。1
成本与生产力类说法不仅更常见，幅度看起来也更大：本季度这类说法的平均幅度是提升 47%，收入增长类说法是 40%（而且前者的样本更宽：24% 对 18%）。这两个平均值都落在彼此的四分位距（interquartile range）之内，换句话说，平均值里装了不少分散度和不确定性：看的时候不妨打个折扣，改用区间来理解（成本与生产力类为 20%–60%，收入与需求类为 20%–83%）。1
公司讨论自己的 AI 项目时，越来越多地使用「部署」这类词（本季度至今有所回落），从 2025 年初的 7% 升到本季度的 12%。「试点」这类用词每个季度都低于 4%。可以看出，即便公司确实在做试点，他们也不会在电话会上谈这些试点。这会让外界更难判断这些试点和投资是否成功，而早期出现的好苗头，往往要到之后的季度才会被提起。1
「Agentic」这个词出现在本季度至今 24% 的电话会中（2025 年第一季度是 9%）。1
同一时间窗内，「Generative AI」从 13% 降到 5%。1
「Copilot」的提及率一直很低（整段时间里约占电话会的 3%）。1
本文摘自 AI Investment Brief 创刊号，这是 Exponential View 推出的一份新刊物。 1
本文目前公开到这里
上面是这一期官方页面上的全部正文。本期被标为免费公开，从开场说明到刊尾的摘录声明都可以直接读到，中间没有会员专享段落。1
需要说清楚的是正文自身的边界：这一期本身就是 AI Investment Brief 创刊号的摘录，作者在开篇和结尾都写明了这一点。完整的创刊号发布在另一份刊物 AI Investment Brief 上，属于另一个站点，不在本文的范围之内。3
References
- 1Exponential View 官方详情页
exponentialview.co
- 2Exponential View 官方 RSS
exponentialview.co
- 3AI Investment Brief 创刊号
investmentbrief.ai
This story was produced automatically by a channel. One sentence is all it takes for Neodrop to keep producing for you.