Elena Verna Draws a Line on Work Hours in the AI Race

Elena Verna is protecting their evenings while working in AI. Their October 22 essay about work hours puts a staffing decision in front of marketing leaders: decide what happens to the time AI saves before adding more assignments.

Verna, a growth operator at Lovable, challenges 996 culture, the expectation of working 9 a.m. to 9 p.m., 6 days a week. Their own choice is specific:

“But I generally don’t work nights or weekends.”

They say Lovable has respected those boundaries. This is a public account of an ongoing routine, not an announcement that the company has changed its working-hours policy.

The time has a purpose. Verna describes making dinner, taking their children to sports, and keeping time for exercise. They put those commitments on the calendar and explain them to colleagues. The essay makes family life visible in an industry conversation often organized around what can be built next.

More campaigns need a decision

Consider a hypothetical marketer who can now prepare a campaign brief in an afternoon instead of a day. Adding another brief might be a good use of that capacity. So might spending the afternoon checking customer evidence behind the first one.

The manager needs to choose. Leaving the choice implicit makes it hard to assess whether the new tool improved the work or simply increased the number of assignments in progress.

Our earlier coverage of marketing’s growing AI workload examined the extra responsibilities around automation. Verna’s account brings that problem into a personal decision about what an employer can expect from someone who cares about the job and also has commitments outside it.

For a small marketing team, the same choice appears when a launch needs another asset and someone has already finished the day’s agreed work. Having the software to make it doesn’t establish who has time to approve it, or whose evening the new request will consume. That’s a staffing decision the manager still owns.

An older study explains the pressure

There is relevant research behind the concern, although it predates this essay. In a February 26 university interview, UC Berkeley Haas doctoral researcher Xingqi Maggie Ye described an 8-month study at a 200-person U.S. technology company, co-authored with associate professor Aruna Ranganathan.

Ye observed work and conducted more than 40 interviews. Employees took on unfamiliar tasks and carried work into moments that previously served as breaks. Keeping several AI-assisted assignments moving also made the day more demanding.

The researcher described how expectations can shift until:

“what was once extra effort becomes standard performance.”

This was qualitative research at one company, still described as in progress in February. It doesn’t establish how common that experience is among marketers, or prove that a particular weekly schedule produces better results.

It does identify something a marketing manager can examine without claiming a universal rule. Compare the team’s agreed workload with the work it now accepts. Include the time spent checking generated material. A higher count of finished drafts gives an incomplete picture when the review queue keeps growing.

A senior person’s boundary has limits

Verna acknowledges that their experience and credibility give them room to refuse work. A junior colleague may feel less able to make the same choice. The issue also carries Chargebee sponsorship; it’s an operator’s argument, not independent evidence about an industry’s employment practices.

Their account makes the promotion decision especially relevant. A marketing lead can ask for evidence of good judgment in a campaign review. They can assess whether a colleague found an error before a launch. After-hours responsiveness answers a different question about availability.

That distinction belongs alongside the responsibilities discussed in our reporting on shared AI skills in marketing job ads. Before adding a new AI responsibility to a role, decide what work it replaces and how the contribution will be assessed.

There will be launches that need an unusual push. Verna allows for those sprints. The useful management task is to name when the sprint ends and what work waits until afterward. Otherwise, an exception can become an expectation without anyone agreeing to it.

For the next AI-assisted campaign, set the expected result and the available working time together. Judge the result against that agreement before expanding the next assignment. That gives a marketer in their first role a fairer starting point than having to negotiate every evening back after the workload has already grown.