The AI that was going to change everything and revolutionize our work He doesn’t seem to be doing any of that at the moment. What there is is a great polarization between those who believe in that promise and between those They do not see it at all clearly or they fear it. And if there is a place where this love-hate for AI is palpable, it is in companies, where CEOs see things in one way and employees in a quite different way.
what has happened. The consulting company Section has conducted a survey of 5,000 workers and managers in US companies with a fundamental question: How many hours of work per week is AI saving you? Survey results, displayed in The Wall Street Journalsay a lot about the vision of CEOs and employees about the impact of AI tools.
Source: WSJ.
CEOs love her, employees not too much. According to data from that survey, two out of three employees indicated that AI does not save them time at work or that at most it saves them less than two hours a week. These responses contrast with those of managers and CEOs: one third affirm that it saves them between 4 and 8 hours, another third affirms that it saves them 8 or more hours, and the other third affirms that it saves them 4 hours or less. The big difference is precisely in this negative view: 40% of employees say that they do not save any time, and only 2% of CEOs agree with that opinion.
AI screws up more than anything else, some say. A user interface designer named Steve McGarvey indicated in that text how managers “automatically assume that AI is going to be the savior (of the business).” His experience is different, however, and he tells how “I have lost count of the times I have looked for a solution to a problem, asked an LLM, and they gave me a solution to an accessibility problem that was completely wrong.”
And it’s not that big of a deal. This professional also indicates that he uses Perplexity as an assistant to research on various projects and that it has saved him time. However, part of their job is to ensure that visually impaired users can access websites, and chatbots have not been of help in that task.
The employees are somewhat afraid. There’s another important aspect to the findings: Employees were much more likely to report feeling anxious or overwhelmed by AI than excited by it. That 40% who responded that it did not save them time added that because of them they would never use AI again.
Employees are the ones who are most overwhelmed by AI, managers are the ones who are most excited about it. Source: WSJ.
For now AI is used like Google. But there is another problem and that is that many of these professionals are using AI as an alternative to the traditional search engine from Google. They do not use it for practical applications of their work—perhaps because they do not know how—and, for example, it was used much less for topics such as code generation or data analysis.
It saves me time, but like it doesn’t. Software companies like Workday participated in the survey and pointed out an interesting fact: this technology imposes an “AI tax” in terms of productivity. Although 85% of its 1,600 employees surveyed indicated that they save between one and seven hours a week thanks to AI, that doesn’t help them much: Much of that saved time ends up being used to correct errors made by AI or modify content generated by AI.
AI isn’t much use (yet). An additional and also recent survey conducted by PricewaterhouseCoopers focused on 4,500 CEOs. The result: only 30% of them were confident that there would be an increase in revenue thanks to AI, although they admit that having a good AI foundation can help boost that return on investment. The adoption of AI, however, at the moment is not compensating them too much, and only 12% of companies claim to have obtained benefits in revenue or costs, while 56% claim to have “not obtained anything” with that investment. These data are in line with those of the MIT study of August 2025 according to which 95% of pilot projects with generative AI were not paying off to companies.
But. The data is negative, but there may be factors that point to a change in trend. The surveys do not indicate how much time users are spending learning how to use AI versus the time it saves them. The benefit may be negative now, but in the long term it will be positive. Furthermore, there are sectors in which AI has clearly become a clear tool to assist workers, as in the field of programming. Although there is, of course, a necessary phase of code review that AI generates, the massive use of these tools indicates that productivity may have gained in whole.
Image | Redd F
GIPHY App Key not set. Please check settings