麦肯锡公司的一位高管表示，人工智能正在帮助员工完成更多工作，但未必能为雇佣他们的企业带来广泛收益。麦肯锡北美区高级合伙人兼主席埃里克·库奇（Eric Kutcher）在周三的公司媒体日活动中告诉记者，除了软件开发等特定领域外，许多企业尚未实现预期中的生产力提升。“个人生产力确实在提高，但他们只是改变了部分工作内容，”他说。
AI is helping workers get more done, though not necessarily delivering broad benefits for the organizations that employ them, according to a top McKinsey & Company executive. Eric Kutcher, a senior partner and chair of McKinsey North America, told reporters at the company's media day on Wednesday that outside specific areas such as software development, many of the hoped-for productivity increases have yet to materialize across companies. "You have people that are getting more productive as individuals, but they're just changing some of the work that they're doing," he said.
库奇指出，除非企业审视整个组织的工作范围，并识别出系统中的低效环节，否则很难实现整体产出的显著提升。“除少数领域外，我们尚未实现企业级生产力提升，”他表示。他提到，早期许多企业试图通过向员工提供Anthropic的Claude和OpenAI的ChatGPT等工具来大幅提升生产力，甚至在某些情况下要求他们尽可能多地使用这些工具。
Until companies look at the scope of work across an organization and identify inefficiencies in the system, it's hard to generate sweeping increases in output, Kutcher said. "We have not gotten enterprise-level productivity, with the exception of a few areas," he said. Early on, he said, many companies' attempts to supercharge workers' productivity meant giving them access to tools like Anthropic's Claude and OpenAI's ChatGPT and, in some cases, telling them to use them as much as possible.
库奇说，一些公司甚至会统计谁使用的AI token最多，“却没有意识到这些token是有成本的”。员工与企业成果之间的这种差距也反映在麦肯锡的数据中。今年8月，这家咨询公司的AI部门QuantumBlack报告称，在一项调查中，十分之八的受访者表示人工智能提高了他们的生产力。但与此同时，调查还发现，认为人工智能对其企业核心运营利润（息税前）有贡献的受访者比例与一年前基本持平，为37%。
Kutcher said that some companies even ranked who was using the most AI tokens "without realizing those tokens cost money." That gap between workers' and companies' results shows up in McKinsey data. In August, the consulting firm's AI division, called QuantumBlack, reported that in a survey, eight in 10 respondents said AI had increased their productivity. At the same time, the survey found that the share of respondents who said that AI had contributed to their organization's profits from core operations — before interest and taxes — was essentially flat from a year earlier, at 37%.
库奇表示，这项技术的益处已在软件开发等领域显现，开发者利用人工智能生成了更多代码。他说，这可以转化为新产品和新功能的更快发布。
Kutcher said the technology's benefits have materialized in areas such as software, where developers have used AI to generate more code. That can translate, he said, into quicker releases of new products and features.
越来越多的企业似乎正试图将软件开发人员通过人工智能实现的生产力提升转化为更广泛的业务成果。麦肯锡的调查显示，从2025年到2026年，在使用人工智能且已超越实验和试点阶段、进入“规模化”模式的企业比例上升了六个百分点。库彻表示，许多企业“目前正在努力思考：‘如何将这种个人层面的生产力提升转化为组织层面的效率？’”
More businesses appear to be trying to turn the productivity boost software developers have seen from AI into broader results. McKinsey's survey showed a six percentage-point increase from 2025 to 2026 in the share of organizations using AI that were moving beyond experimentation and pilot phases into "scaling" mode. Many organizations, Kutcher said, "are trying to figure out right now, 'How do I shift from this individual productivity into the institutional?'"
为了实现这一转变，库彻指出，一些公司正在从头到尾重新思考其业务的关键部分。对于一家生命科学公司而言，这可能意味着考虑人工智能如何重塑研发流程；而对于一家石油和天然气公司，则可能意味着采用新的钻井方法。“你可以选定一个领域，然后问自己：‘我该如何以彻底不同的方式来做这件事？’”他说。
To make that shift, Kutcher said, some companies are rethinking a key part of their business, from end to end. For a life-sciences company, that might mean considering how AI could remake research and development; for an oil-and-gas company, it might mean new approaches to drilling. "You pick the area and say, 'How would I do this profoundly differently?'" he said.