科技巨头扩展人工智能需要数千亿美元，但美国经济中最小的企业只需几个AI工具，就能以具有成本竞争力的商业模式起步运营。
It takes hundreds of billions of dollars for the tech giants to scale artificial intelligence, but it takes just a few AI tools for the smallest companies in the U.S. economy to get up and running with a cost-competitive business model.
Alex Willen 预计，他无员工运营的12个电商品牌今年将产生250万至300万美元的收入。Kristin Ginn 于3月离开微软，创办了AI采用咨询公司 trnsfrmAItn。七个月后，她在没有员工或固定承包商的情况下，已获得或签约约25万美元的业务，持续至年底。
Alex Willen expects the 12 e-commerce brands he operates with no employees to generate between $2.5 million and $3 million this year. Kristin Ginn left Microsoft in March and started trnsfrmAItn, an AI-adoption consultancy. Seven months later, she has roughly $250,000 in business received or contracted through year-end with no employees or regular contractors.
斯坦福大学经济学家 Erik Brynjolfsson 表示：“最大的误解是认为这只是关乎个人生产力；实际上是关于利用AI执行此前需要整个团队才能完成的任务。”关于AI在个体创业者商业世界影响的早期研究值得关注。
"The biggest misconception is that it's just about individual productivity; it's actually about leveraging AI to perform tasks that previously required a whole team," said Stanford economist Erik Brynjolfsson. Early research on the impact of AI within the solopreneur business world bears watching.
乔治梅森大学默卡托斯中心劳工政策项目主任、劳工经济学家 Liya Palagashvili 指出，新企业成立数量的上升与不太可能成为雇主型企业的申请直接相关。
According to Liya Palagashvili, a labor economist and director of the Labor Policy Project at George Mason University's Mercatus Center, new business formation is rising in direct correlation to applications unlikely to become employer firms.
在高度接触AI的行业中，没有早期招聘迹象的商业申请从2024年第一季度到2026年第一季度增长了26.8%，从每季度约22.2万份增至28.1万份。
dustries highly exposed to AI, business applications without early signs of hiring rose 26.8% from the first quarter of 2024 through the first quarter of 2026, from roughly 222,000 to 281,000 per quarter.
同期，这些行业中显示更强最终成为雇主型企业迹象的申请下降了6.4%。在 Palagashvili 用作对照组的建筑业和批发贸易中，非雇主型申请下降了0.4%。她说：“AI所做的就是将这些能力直接交到个人手中。”
Over the same period, applications in those industries with stronger indications of eventually becoming employer businesses fell 6.4%. In construction and wholesale trade, which Palagashvili used as a comparison group, non-employer-type applications declined 0.4%. "What AI is doing is giving those capabilities directly into the hands of the individual," she said.
帕拉加什维利警告称，这些发现是初步的和描述性的，并不能确定人工智能导致了增长。基于人口普查的衡量指标识别出没有与短期工资单或招聘计划相关的指标的申请；它并未显示这些企业最终仍然是一人公司或从未雇佣员工。但威伦展示了经济状况可能是什么样子。
Palagashvili cautioned that the findings are preliminary and descriptive and do not establish that AI caused the increase. The Census-based measure identifies applications without indicators associated with near-term payroll or hiring plans; it does not show that those businesses ultimately remained one-person firms or never hired employees. But Willen shows what the economics can look like.
人工智能公司Anthropic的Claude帮助威伦处理其企业中的记账、客户服务、广告和库存工作，而OpenAI的图像工具完成了曾经交给摄影师和设计师的工作。
Anthropic's Claude helps with bookkeeping, customer service, advertising, and inventory across his businesses, while OpenAI image tools perform work that once went to photographers and designers.
威伦说，如果没有人工智能，“我可能还需要另一名全职人员，还得雇佣承包商来处理摄影、设计、记账和广告管理。”他说，一次传统的产品摄影拍摄可能需要2000美元到5000美元，而设计师为新产品图片收取300美元到1000美元。
Without AI, Willen said, "I'd probably need one other full-time person, plus I'd need to use contractors for photography, design, bookkeeping, and ad management." A traditional product photography shoot could cost $2,000 to $5,000, and a designer $300 to $1,000 for new listing images, he said.
威伦说，当人工智能直接消除其中一项开支时，“那笔钱就直接进了我的口袋。”人工智能还让他能够追求以前从未考虑过的收入渠道。在收购了销售双筒望远镜背带、背包及其他户外配件的狩猎装备品牌Lone Trail后，威伦说，其Shopify网站曾约占销售额的1%。他没有关闭这个渠道，而是使用ChatGPT和Claude重建了网站，并利用人工智能帮助运行Meta广告。他说，在最近的一个30天期间，Shopify约占Lone Trail收入的30%。
When AI directly eliminates one of those expenses, Willen said, "that money just goes into my pocket." AI has also allowed him to pursue revenue channels he would not have previously considered. After acquiring Lone Trail, a hunting-gear brand selling binocular harnesses, packs, and other outdoor accessories, Willen said its Shopify site accounted for roughly 1% of sales. Rather than shut down the channel, he used ChatGPT and Claude to rebuild the site and AI to help run Meta advertising. In a recent 30-day period, he said, Shopify accounted for roughly 30% of Lone Trail's revenue.
威伦说，“这最终将至少带来六位数的年收入，没有人工智能这些收入都不会存在。”
"That's going to end up being at least six figures of annual revenue that would not exist without AI," Willen said.
金的企业展示了新进入者在不先围绕其构建组织的情况下，能多快实现有意义的收入。她利用人工智能协助处理提案、定价、战略及其他工作，并正在探索使用人工智能代理来完成部分劳动密集型客户流程。
Ginn's business shows how quickly a new entrant can reach meaningful revenue without first building an organization around it. She uses AI to help with proposals, pricing, strategy, and other work and is exploring an AI agent to perform part of a labor-intensive client process.
会计和发票软件公司 FreshBooks 进行的一项调查在更广泛的范围内发现了类似行为。79% 的个体创业者表示，在雇佣他人完成自己无法胜任的任务之前，他们曾尝试过使用 AI（如果将微型企业主也计算在内，这一比例为 86%）。FreshBooks 首席执行官 Shaheen Javadizadeh 表示，AI 让他们能够“看起来比实际规模大得多”。
A survey conducted by accounting and invoicing software company FreshBooks identified similar behavior more broadly. Seventy-nine percent of solopreneurs said they had tried AI before hiring someone for a task they could not perform themselves (the same figure was 86% among when including microbusiness owners). AI allows them to "look much bigger than they are," FreshBooks CEO Shaheen Javadizadeh said.
但现有的小企业数据却讲述了另一个故事。为小企业提供薪资和人力资源软件服务的公司 Gusto 跟踪了企业在采用 AI 后的情况，发现一年后，这些企业的员工人数比了解 AI 但未采用 AI 的可比企业高出约 7%。在员工少于 10 人的企业中，采用 AI 的企业员工人数增长比可比的非采用企业高出约 10%。
But among existing small businesses, the available data tells another story. Gusto, a payroll and HR software company that serves small businesses, tracked businesses after they adopted AI and found their headcount was about 7% higher after a year than a comparable businesses that knew about AI but had not adopted it. Among firms with fewer than 10 employees, AI adopters grew headcount about 10% more than comparable non-adopters.
Willen 的经历也凸显了新商业模式中的一种权衡：减少人员并不意味着消除了对大型技术平台的依赖。其总体收入中约 80% 仍来自 Amazon，他利用 Meta 和 Shopify 拓展直销渠道。Ginn 没有员工，但她表示，LinkedIn 是给她的咨询业务带来客户的最大渠道之一。
Willen's experience also highlights a tradeoff in the new business model: operating with fewer people does not eliminate reliance on larger technology platforms. Roughly 80% of its overall revenue still comes through Amazon, and he uses Meta and Shopify to expand direct sales. Ginn has no staff, but LinkedIn has been one of the biggest drivers of customers to her consultancy, she said.
Brynjolfsson 表示，小型生产者的激增实际上可能会增加连接他们与买家的平台的价值。
Brynjolfsson said the proliferation of smaller producers can actually increase the value of the platforms that connect them with the buyers.
“存在着这种产品长尾，但与此同时，像 Amazon 这样的大型平台比以往任何时候都更强大，因为它们聚合了所有这些产品并将它们汇集在一起，”Brynjolfsson 说。
"There's this long tail of products, but at the same time, big platforms like Amazon are more powerful than ever because they aggregate all of these and they bring them together," Brynjolfsson said.
但这种关系并不像劳动力节省直接流向大型科技公司那样简单。Willen 表示，他节省下来的部分钱变成了利润，而 Ginn 每月在付费 AI 工具上的花费不到 60 美元。
But the relationship is not as simple as labor savings flowing directly to Big Tech. Willen said some of the money he saves becomes profit, while Ginn spends less than $60 a month on paid AI tools.
AI 可以使更多小型生产者的存在变得经济可行。Brynjolfsson 表示，随着核心业务生产成本降低，小企业将更加专注于客户获取和分销。
AI can make it economical for more small producers to exist. And as core business production becomes cheaper, small businesses place greater focus on customer acquisition and distribution, Brynjolfsson said.
布林约尔松认为，虽然独资创业者将蓬勃发展，但 AI 的使用同时也将导致平均企业规模变得比今天更大。
Brynjolfsson believes that while solopreneurs will thrive, use of AI will at the same time lead to the creation of even larger firms, on average, than exist today.
在 9 月发布的一份题为《AI 在社会中的知识利用》的研究报告中，布林约尔松和人类学研究所研究员佐伊·希齐格（Zoë Hitzig，研究 AI 的社会与经济影响）指出，随着 AI 将此前分散在工人手中的知识编码化，并使更多信息能够被汇总并在中央层面采取行动，企业平均规模将扩大，行业集中度上升，而地方管理自主权下降。研究人员表示，早期证据在很大程度上与这些模式一致。
In a September research report titled "AI's Use of Knowledge in Society," Brynjolfsson and Zoë Hitzig, a researcher at the Anthropic Institute who studies the social and economic impacts of AI, argue that as AI codifies knowledge previously scattered among workers and enables more information to be aggregated and acted upon centrally, firms will grow larger on average, while industry concentration increases and local managerial autonomy declines. The researchers say early evidence is largely consistent with those patterns.
这对中型企业是个坏消息。布林约尔松预测，AI 可能导致中型企业出现“空心化” —— 一端是能够做更多事情的极小团队，另一端是利用 AI 和规模优势的巨型组织。
That's bad news for medium-sized companies. Brynjolfsson predicts AI could contribute to a "hollowing out" of midsize firms — very small teams capable of doing more at one end and giant organizations exploiting AI and scale at the other.
他说：“中间的‘肥肉’ —— 那些专注于内部协调的管理层级 —— 很可能会大幅减少。”
"The fat in the middle — the layers of management focused on internal coordination — will likely be significantly reduced," he said.
亚当·布兰奇（Adam Branch）曾在两家截然不同的科技公司内部经历过这种转变的一个版本。
Adam Branch has experienced a version of that shift from inside two very different technology companies.
在创立利用 AI 实现遗留系统现代化的企业软件公司 rhino.ai 之前，布兰奇建立了 Incentive Technology Group，这是一家传统的技术服务公司，员工曾增长到约 500 人。他说，在那种模式下，赢得更多业务意味着雇佣更多人。
Before founding rhino.ai, an enterprise software company that uses AI to modernize legacy systems, Branch built Incentive Technology Group, a traditional technology services company that grew to roughly 500 employees. Under that model, winning more work meant hiring more people, he said.
布兰奇说：“有了 rhino.ai，我们正在构建 AI 软件，来完成那些团队过去所做的更多工作。”
"With rhino.ai, we're building AI software to do more of the work those teams used to do," said Branch.
rhino.ai 约有 50 名员工。布兰奇估计，同等的现代化工作现在每位员工的生产率可提高三到四倍。
Rhino.ai has roughly 50 employees. Branch estimates comparable modernization work can now be performed with three to four times the productivity per employee.
他的团队估计，在 ITG 时期，人工成本约占运营支出的 85%，技术约占 5%。而在 rhino.ai，布兰奇表示人工成本约占 40%，技术（包括云服务、软件和 AI 模型）约占 25%。
His team estimates labor represented roughly 85% of operating spending at ITG, and technology about 5%. At rhino.ai, Branch puts labor at roughly 40%, while technology, including cloud, software, and AI models, represents about 25%.
AWS 在 AI 原生初创企业中也看到了收入与人数关系类似的松动。其“增长引擎”研究——一项针对 20 个国家 3413 名初创企业创始人和高管的独立全球研究——发现，这些公司的年均收入增长率为 156%，而初创企业整体为 65%；超过一半的公司人均收入超过 40 万美元。该报告聚焦于成立不足五年、以最先进形式使用 AI 的公司。
AWS sees a similar loosening of the relationship between revenue and headcount among AI-native startups. Its Engines of Growth research, an independent global study of 3,413 startup founders and senior leaders across 20 countries, found those companies grew revenue an average 156% annually versus 65% for startups overall, while more than half generated over $400,000 in revenue per employee. The report focused on companies under five years old that use AI in its most advanced forms.
“增长过去直接取决于你能雇佣多少人，”AWS 专家与合作伙伴副总裁 Ruba Borno 说。“现在创始人可以靠技术扩张，而不只是靠人数。”Brynjolfsson 并不认为这仅仅是大科技公司以牺牲其他所有人为代价获胜。他说，企业家可以保留部分生产力收益，消费者获得更多选择，平台则因提供技术并连接更多生产者与客户而受益。
"Growth used to be tied directly to how many people you could hire," said Ruba Borno, vice president of AWS Specialists and Partners. "Now founders can scale with technology, not just headcount." Brynjolfsson does not see the result as simply Big Tech winning at everyone else's expense. Entrepreneurs can keep some of the productivity gains, consumers get more choice, and platforms benefit from supplying the technology and connecting more producers with customers, he said.
“平台和创造工具的人在赢，因为他们能卖出这些工具；个体创业者在赢，因为他们能使用这些工具；消费者在赢，因为他们得到了以前没有的新产品，”他说。“这主要是把蛋糕做大了，所以是三赢——三方共赢。”他说，AI 可能改变的是这个更大蛋糕的形状。
"The platforms and the people who are creating the tools are winning because they are able to sell them, the solopreneurs are winning because they are able to use them, and the consumers are winning because they are getting new products that they didn't get before," he said. "It's mostly a bigger pie, so it's a win-win-win — a triple win." What AI may change is the shape of that bigger pie, he said.
“每当出现技术变革、平台变革，显然会使创造这些平台和技术的公司受益，”Brynjolfsson 说。“但也会使能够从头开始的新进入者受益。”
"Whenever there's a technology change, a platform change, obviously it benefits the companies that are creating those platforms and technologies," Brynjolfsson said. "But it also benefits new entrants who can start fresh."