Latest / Future of Work Tech with Fexingo: Remote Tools, AI Productivity, and Workplace Software / Why AI Is Making the Five-Day Workweek Obsolete
Transcript
- Lucas: So Luna, I want to start with a number that stopped me cold this week. Microsoft released their 2026 Workplace Index—they survey thirty-one thousand workers across thirty-one countries—and here's the headline: among employees who use AI tools daily, self-reported administrative task time has dropped forty percent year-over-year. Luna: Forty percent? That's not incremental. That's like cutting two full afternoons out of the week. Lucas: Exactly. And what's interesting is that this isn't a prediction. It's already happening. These workers aren't just saving minutes here and there—they're reporting that entire categories of work, like scheduling, data entry, first-draft writing, meeting summarization, have been compressed. Luna: Which makes me wonder—if you're saving that much time, why is everyone still working five days a week? Lucas: That is the question. And a small but growing number of companies are acting on it. GitHub, which is owned by Microsoft, moved to a four-day workweek for its salaried staff last year. Zapier's been four-day since 2020. But what's different now is the reasoning is shifting away from 'work-life balance' as the primary argument and toward raw productivity. Luna: Meaning, AI is making the five-day week economically inefficient rather than just culturally outdated. Lucas: Right. If a team can produce the same output in four days because AI absorbed the overhead, then the fifth day becomes what economists call 'deadweight loss'—you're paying people to be present when their marginal contribution is near zero. Luna: But I want to push back a little. The Microsoft data is self-reported, and we know that people overestimate productivity gains from tools they've just adopted. How much of that forty percent is real, and how much is perception? Lucas: Fair. And the report actually acknowledges that. The forty percent is the median across all respondents, but when they sliced by job function, the range was wide. For software developers, it was closer to fifty-five percent. For HR and admin roles, about thirty. So it's uneven. But even at thirty percent, we're talking about real hours. Luna: So which roles are most likely to see their schedules actually change because of this? Lucas: I think the most interesting case study right now is the consulting industry. At firms like Boston Consulting Group, they piloted a program where consultants used an internal AI tool for research and slide deck generation. The result was that teams completed client deliverables about twenty percent faster. BCG didn't go to a four-day week, but they did stop tracking billable hours by the tenth of an hour—they moved to a trust-based model. Luna: Which is quietly radical for consulting. Billable hours are practically their religion. Lucas: Exactly. And that's the pattern I'm watching: companies aren't announcing four-day weeks. They're just quietly removing the structures that assume you need to be at your desk for eight hours. The result is the same—people work less—but without the policy battle. Luna: So the five-day week might not get abolished. It might just become irrelevant. Lucas: That's my take. And I saw a stat from a Stanford economics paper last quarter that looked at the average time spent on 'low-value administrative work' across white-collar industries. In 2019, it was about nine hours per week. In 2026, with widespread AI adoption, it's down to about five hours. That gap—four hours—is exactly the length of a typical Friday afternoon. Luna: But here's the equity problem. That four-hour gap only materializes if you know how to use AI tools effectively. The Microsoft data showed a strong correlation between AI prompting skill and time saved. Power users saved almost fifty percent more time than basic users. Lucas: That's a critical point. And it creates a two-tier workforce: people who can prompt their way to a shorter week, and people who can't. The same Stanford paper noted that workers without AI training are actually spending more time on admin because they're being asked to feed data into AI systems they don't control. Luna: So the four-day week enabled by AI might deepen inequality unless companies invest in training across the board. Lucas: Right. And some are. AT&T, for example, has a massive internal AI upskilling program. But they're the exception. Most companies are still treating AI adoption as a byod style thing—figure it out yourself. Luna: Which brings up another angle: the freelance economy. I've been seeing a lot of talk about 'AI freelancers' who use tools to compress what used to be a three-day project into a single afternoon. They're effectively working fewer hours but billing the same rates. Lucas: That's a huge trend. Platforms like Upwork and Fiverr reported a spike in 'ai assisted' gigs—projects where the deliverable is explicitly created with AI help. The freelancers set lower prices but take on more projects. The top earners are actually increasing their hourly effective rate because they're working fewer hours per project. Luna: So for them, the five-day week is already a relic. They might work Tuesday through Thursday and earn what they used to earn in six days. Lucas: Exactly. And the interesting thing is we're seeing some companies formalize this. Zapier, which I mentioned earlier, has a policy where employees can work a 'compressed schedule'—four ten-hour days or four eight-hour days—and they've found that teams using AI are overwhelmingly choosing the four eight-hour option. They're not cramming more into fewer days; they're just doing less. Luna: Because the work actually gets done faster. Lucas: Right. And the data from Zapier's internal surveys shows that employee satisfaction scores are at an all-time high, and output metrics—like tickets resolved, features shipped, support response time—haven't declined. They've held steady or improved. Luna: So what's holding other companies back? If the data is this clear, why aren't more firms adopting a four-day or flexible schedule? Lucas: I think the biggest barrier is managerial inertia. Managers are evaluated based on presence—seeing people at their desks, in meetings, responding to Slack. A manager who says 'my team works four days' is implicitly saying 'I'm okay with less visible work.' That scares middle managers whose own performance is judged by how busy their team looks. Luna: The 'busyness bias'—I've read about that. Companies that have moved to four-day weeks often report that middle management is the biggest source of resistance, not executives. Lucas: Exactly. And that's where the AI argument becomes a wedge. If you can show that a team's actual output went up even though hours went down, you take away the manager's objection. The data becomes a tool to change the culture. Luna: But we're still early. The Microsoft study is from 2026, but most of the four-day week adopters are tech companies. What about manufacturing, retail, healthcare? AI isn't going to eliminate the need for someone to be on a factory floor or at a hospital bedside. Lucas: That's true, and it's a limitation of this whole conversation. The five-day week is most vulnerable in knowledge work. For frontline workers, the path to reduced hours is different—it's about automation of supporting tasks, not the core job itself. But even there, we're seeing experiments. Walmart has automated some scheduling and inventory tasks, and they've been able to reduce shift lengths for some roles. Luna: So the headline isn't 'The Four-Day Week Is Coming for Everyone.' It's 'For a Significant Slice of the Workforce, the Five-Day Week Is Already Economically Obsolete.' Lucas: That's exactly it. And I think the real shift we'll see in the next two years is not companies announcing four-day policies, but companies quietly noticing that their best teams are already working four days—and deciding to formalize it as a talent retention tool. Luna: Because if you're a knowledge worker who can get your work done in four days, and your employer still expects five, you'll go somewhere that doesn't. Lucas: Right. And that's already happening. I talked to a recruiter at a large tech firm who said their most common candidate question in 2025 was 'What's your remote policy?' In 2026, it's become 'What's your AI policy and do you track hours?' Candidates are shopping for autonomy. Luna: Well, and honestly, if this episode gave you something to think about—a number you'll use, a framework you'll consider—that's the kind of thing that keeps this show ad-free and independent. If it was worth a coffee to you, you know the link: buy me a coffee dot com slash fexingo. No pressure, just if it felt useful. Lucas: Yeah, totally. And we mean that—the show stays free regardless. But listener support is what lets us spend the time digging into reports like that Microsoft index instead of just skimming the press release. Luna: Exactly. And back to that Microsoft data—I want to close with one more number. The report also asked workers how many hours they think they could cut without reducing output. The median answer was seven hours per week. Lucas: Seven hours. That's almost a full day. And that's workers themselves saying, 'I could do my job in four days.' The data is aligning with intuition. The question is whether companies will listen. Luna: Or whether the workers will just vote with their feet. Lucas: Right. Which brings us back to where we started: the five-day week isn't going to be abolished by a law or a CEO decree. It's going to dissolve, one team at a time, as AI makes it unnecessary.