AI productivity is usually sold as faster work, cheaper work, or more work. Mike Su asks the more awkward question: if AI can turn a week of white collar output into a much shorter sprint, can workers take Friday off? The post is short and playful, but the argument lands because it goes straight at the missing line in most AI adoption decks: who gets the saved time?
Table of Contents
The short version
- Mike Su’s “Can we have the day off?” asks whether claimed 10x AI productivity should translate into a four-day workweek for white collar workers.
- The strongest version of the argument is about distribution, not model capability. If AI agents compress work, employees will ask for time, pay, or both.
- Hacker News readers turned the joke into a labor debate: some saw a serious bargaining question, while others argued market competition will push companies to demand more output instead.
- For builders, the product lesson is blunt. AI tools that only promise management more throughput may make employees feel less secure, even when the software is useful.
AI productivity and the four-day workweek
Su’s post starts from a familiar claim: AI is supposed to raise white collar productivity by a large multiple. If that premise is true, he asks, why should the gain only appear as more output for the employer?
The concrete proposal is deliberately simple. People work Monday through Thursday. On Thursday they prepare prompts and tasks. On Friday, AI agents keep working while the humans take the day off. It is partly a joke, but it exposes a real gap in the current workplace conversation.
Most companies talk about AI productivity as capacity. Ship faster. Write more. Support more customers. Close more tickets. Employees hear a different message: use the same hours to do more, with no guarantee of higher pay, more leave, or better job security.
That is why the Friday framing works. It turns an abstract productivity claim into a payroll and calendar question.
What happened
The original essay, published on May 27, 2026, is a short personal blog post titled “Can we have the day off?” Su writes that if AI can produce the same work in a fraction of the time, then a four-day schedule should be a reasonable ask. He even imagines Friday as an “AI workers’ day,” where agents run while humans are out of the office.
The post does not present a benchmark or a policy plan. It is closer to a pressure test for the way AI is being marketed inside companies. If executives believe AI can multiply output, employees can reasonably ask whether some of that gain becomes time off.
That makes the piece useful beyond the joke. For more IT and AI workplace briefs, the IT & AI archive tracks similar shifts in automation, developer tools, and product operations.
Why this is worth watching
AI productivity gains are easy to claim and hard to divide. A team can adopt coding assistants, research agents, summarizers, and workflow bots without ever agreeing on what happens to the saved hours.
That silence creates a management problem. If a company tells employees that AI will make them vastly more productive, but keeps the same schedule and raises the output target, the tool starts to look like surveillance with a nicer interface. If the company offers a share of the gain, through shorter workweeks, better compensation, or fewer low-value tasks, adoption has a better chance of feeling like a deal instead of a threat.
The four-day workweek is only one possible answer. The larger question is whether AI productivity becomes a worker benefit, an owner benefit, or a mix of both. That question will shape how teams talk about agents, copilots, and automation over the next few years.
What Hacker News readers are arguing about
The Hacker News thread was large: more than 1,300 points and hundreds of comments when checked. The first serious thread picked up the post’s main point almost exactly. If employees help introduce AI into their workflows, one commenter argued, they should ask what they get in return: days off, higher pay, or some other concrete share of the gain.
A second camp was more cynical. They argued that productivity gains usually flow to owners, especially when workers are worried about layoffs. Several comments connected the issue to older automation cycles: computers, software, and the internet made many tasks faster, but the standard workweek did not shrink much for most employees.
The useful objection in the discussion is competition. Some readers argued that a company offering Fridays off could be outpaced by rivals that use AI to work faster all week. Others pushed back, pointing out that many companies already waste huge amounts of time on busy work, weak coordination, and rework. More hours do not automatically mean more useful output.
There was also a policy thread. Some readers moved from employer-level bargaining to unions, worker protections, taxes, UBI, and social safety nets. That jump matters because it suggests the four-day workweek may be hard to win company by company if the market rewards whoever turns AI into raw output first.
Treat the thread as sentiment, not proof. But the sentiment is clear enough: workers are starting to ask whether AI productivity will give them leverage or simply raise the bar.
The practical read
If you run a team, do not pitch AI productivity only as acceleration. Say what happens to the saved time. Will it reduce after-hours work? Remove recurring busy work? Change sprint scope? Create a trial four-day schedule after the team proves the workflow? Vague promises will not survive contact with calendars.
If you build AI tools, this is a product positioning issue. A tool that says “your manager can get 10x more from you” and a tool that says “your team can finish the same work with fewer wasted hours” may have similar features, but they land very differently.
For employees, the move is to make the bargain explicit. Track which tasks AI actually shortens, how much review work remains, and where quality still depends on humans. Then ask for a share of the gain in terms that can be measured: time off, compensation, narrower scope, or fewer low-value meetings.
AI productivity will not automatically create a shorter workweek. Someone has to ask for it, price it, and design the workflow around it.