AI 996 Culture: Why Longer Hours Aren’t an Advantage

More output can conceal weak choices

AI increases what a small team can attempt. A developer can spin up a prototype faster. A marketer can test more campaign variants. A product team can explore multiple concepts before committing. Those capabilities are useful - but they don't automatically create more time or make every opportunity worth chasing.

A familiar pattern follows: something that used to take six months now takes six weeks, so the organization fills the newly available space with more tasks. The productivity "win" turns into a higher expectation, not less pressure. The team ships more, but the backlog doesn't shrink - it grows to match the new capacity.

This isn't automatically a reason to avoid ambitious goals. It's a reminder that productivity only helps if leaders decide what to do with the gain. The options are straightforward:

• Deliver the same work with less effort.
• Improve quality or reduce operational risk.
• Take on additional work with a credible business case.
• Give people more room to recover and keep performing well.

Many companies default to the third option. It's fine when the opportunity is genuinely valuable and the team really has room to take it on. It becomes a problem when every efficiency improvement instantly turns into "now we should do even more."

Hours are an input, not an outcome

Working long hours can signal commitment, but it doesn't prove someone is doing high-value work. One person can spend 12 hours grinding through low-priority tasks, endless meetings, or rework caused by unclear decisions. Another person might make a hard call in 30 minutes that saves the team a week.

This matters even more in AI work. Tools can make individual contributions easier to see. A small team may build a prototype, analyze feedback, or automate a workflow without the bigger process that used to surround those tasks. When outcomes are easier to observe, measuring people by online presence becomes harder to justify.

None of this means quality is easy to measure. Reliability, collaboration, risk reduction, and long-term impact can take time to show up. But difficulty isn't the same thing as permission to use hours as a substitute. The fix is to define outcomes more carefully - not to lean on schedules as a proxy.

A company that only hits its goals when everyone works 996 isn't necessarily dealing with a commitment problem. It may be dealing with a scope, staffing, sequencing, or management problem. Before asking people to work longer, look at the system that keeps generating more work than the team can sustainably finish.

Burnout is not the only cost

Talk about unsustainable work often collapses into burnout alone. Burnout matters, but the organizational impact is broader. A team under permanent pressure has less time to review decisions, document systems, learn from incidents, and pay attention to how customers actually use the product.

They also lose people who can't - or don't want to - make work the center of their lives. Parents are an obvious example, but it's wider than that: people caring for relatives, managing health constraints, or protecting commitments outside the office. When a culture prizes unlimited availability, it filters out talent before those people even apply.

This is a product and business issue, not just a personal one. Teams are shaped by who gets to join and stay. If you narrow that group, you narrow the assumptions built into a product: which problems get noticed, and which risks get treated as real. On top of that, experienced employees often bring lessons from previous launches, scaling attempts, and mistakes you don't want to repeat. When that experience leaves, you pay for it.

And no, the answer isn't "parents are always better" or "people without caregiving responsibilities are less disciplined." The practical point is simpler: constraints force decisions. When someone has a hard stop, they learn to separate essential work from work that's just available. A team that never has to make that call can drift into meetings, detours, and urgent requests that chew up time without moving the business forward.

A sprint and a permanent pace are different systems

There are real reasons to work unusually hard for a limited period. A security incident, a critical customer issue, a launch, or a genuine market moment can justify a short sprint. The key word is limited. A sprint needs a clear objective, a defined duration, and recovery afterward.

If everything is treated as exceptional, there's no recovery phase. Deadlines may still get met, but the organization won't know whether the pace is sustainable until people start leaving or performance slips. By then, the cost is already in the system.

Here's a test leaders can use: if the team takes a weekend off, does the business miss its goals? If the answer is yes, the organization may be fragile - relying on a few people's unpaid extra capacity instead of having realistic plans, enough coverage, and a delivery process that holds up.

High-pressure environments also reward the wrong behavior. If promotion and recognition go to the people who are visibly online at all hours, employees learn that availability matters more than impact. A written policy for flexible work won't override what people actually see being rewarded.

Employees: make trade-offs explicit

When a job matters and the team is talented, it's easy to keep saying yes. Work produces urgency on demand: a deadline, a customer request, a launch, or a colleague who needs help. Personal commitments can feel easier to move, so they get moved again.

One decision tool that cuts through the heat is the five-year regret test. When a work commitment conflicts with something important outside work, ask which choice you'll be more likely to regret five years from now. This isn't saying "work always loses." Some opportunities are meaningful, and it can be valuable for kids - or other family members - to see someone commit to what matters. The test creates distance from the urgency of the moment and makes the trade-off visible.

It also helps to state boundaries clearly. Tell colleagues when you're unavailable and what you're unavailable for. "I'm taking my child to soccer" is clearer than a vague calendar block that invites people to assume they can still reach you. The same goes for medical appointments, caregiving, or a regular time to exercise and recover.

A boundary isn't a guarantee that everyone will respect it. Some employers do reward constant availability, even while saying otherwise. Employees should assess that trade-off honestly, especially when choosing a role or deciding what they're willing to accept. Still, being explicit makes expectations legible. People can't consistently respect a limit they never see.

Leaders: make the stated culture observable

Leaders shape working norms more than any policy. If a manager claims that time off matters but sends messages at midnight and works through every holiday, the team will notice. If the people who get promoted are always available, the organization is rewarding availability - whether or not it says that on a values page.

Good leadership practices are concrete:

• Define what counts as an emergency and who needs to respond.
• Set priorities so the team knows what won't be done.
• Take vacation and disconnect, rather than modeling constant availability.
• Don't treat a delayed response as lack of commitment when no urgent response was agreed on.
• Recognize people for the quality and impact of their work - not just visible effort.
• Check whether deadlines and staffing make the expected workload realistic.

Managers should also be careful about "absorbing" extra work themselves. Sometimes doing that protects the team. But if a leader repeatedly handles nights and weekends without saying it's temporary, employees may interpret it as the baseline they're expected to match. When the sprint is over, make the exception explicit - and fix the underlying capacity or prioritization issue.

The real test isn't whether senior executives can set boundaries. It's whether a junior employee can leave at the agreed time without harming their prospects. Culture is defined by what behavior gets rewarded and tolerated, not by what benefits it advertises.

Measure contribution, not the green status bubble

AI changes what a person or team can accomplish, but it doesn't remove the need for judgment. Companies should be explicit about what "good work" means: outcomes, quality, customer impact, reliability, and collaboration. Online presence can matter in roles that require coverage, but it shouldn't stand in for contribution across the board.

The measurement problem is real. Prototypes are easy to count; the quality of the product decision behind them can be harder to assess. A feature can ship quickly and create support costs later. Sometimes the best call is not to build something at all - and it saves money without leaving a shiny artifact. Leaders need measurement that accounts for those differences instead of rewarding raw volume.

The same discipline applies to AI tools. If AI reduces the time required for a task, decide what the benefit should do: improve throughput, improve quality, lower cost, or increase employee capacity. Don't assume every saved hour must instantly be filled. That's a business choice, not a technological inevitability.