Happy Thursday, everyone.
Something keeps coming up in conversations about AI and work this year. People describe finishing tasks faster than before, feeling pleased about that, and then noticing something underneath the efficiency that's harder to name. The work was done. But they weren't quite in it.
Three threads worth paying attention to.
This week in workplace whiplash 🌀
🏦 BofA bans back-to-back remote days
Starting mid-September, employees can't work remotely two days in a row, including Fri-Mon. Not more office days. Just no consecutive ones. They say it's about real estate.
💊 Starbucks cut GLP-1 coverage. BofA just expanded it.
Two employers, one benefit, opposite decisions in the same week. A third of employees say they'd switch jobs over this.
💰 54% of workers would take a pay cut for job security
Monster's 2026 report. Separately, 41% say AI is already replacing parts of their role. The two surveys weren't linked. They don't need to be.
🤝 In partnership with Indeed
Humanity at the heart of AI. As AI transforms everything from hiring to daily workflows, humanity is the key to making it all work. Join Indeed FutureWorks from Sept. 2–3 to learn how leaders can prioritize humanity as tech shifts. Get free tickets to the virtual event.
🔋 When capability becomes expectation
Almost 70% of HR professionals' ChatGPT conversations are about work outside traditional HR responsibilities: analysing financial data, building communications for other teams, editing content that once belonged to a different department. Nothing in the job description changed. The job did.
This is what job creep looks like in the AI era. Parkinson's Law tells us work expands to fill the time and resources available. AI is creating new capacity, and that capacity is being filled by work nobody consciously signed off on. Each individual request makes perfect sense on its own. It's only when you zoom out over six or twelve months that you see what the role has actually become.
Roles need reviewing as often as the technology changes, not just at annual appraisal.
💬 When AI takes the seat next to you
New research found that 37% of employees have already used AI for companionship at work. Not just for tasks. For the kind of quick, low-stakes conversation that used to happen between people. "What would you do here?" "Does this make sense?"
Those conversations do more than answer questions. They build the informal trust that holds teams together. They bring information forward that never makes it into formal channels. When they move to AI, they become faster and less surprising. The problem gets solved. The relationship doesn't develop.
The conditions for human connection don't maintain themselves. They need designing for.
🧠 When the work starts to feel flat
There's a specific kind of fatigue that comes not from overwork but from underwork of a particular kind. When AI handles the parts of a task that require effort, iteration, and real judgement, the brain doesn't get the same reward signal it would from working through those things itself. The task is done. The satisfaction is missing.
Research on cognitive engagement is clear: people don't just want outcomes, they want the experience of earning them. AI is making outcomes easier to reach. In some cases, it's removing the part of the process that made the work feel worth doing. Engagement and output are different things. You can have one without the other.
🚀 What this means for leaders
These three things are connected. Together they describe a workforce that is producing more, connecting less, and finding it harder to explain why work feels different from how it did two years ago.
Job creep is a workload and recognition problem. Roles need reviewing as often as the technology changes. Looking back over six months often reveals responsibilities that accumulated without anyone deciding they should.
Connection loss is a culture design problem. It requires deliberate investment in human interaction, in structures that create the conditions for it rather than platforms that simulate it.
Cognitive flatness is a conversation problem. People need space to talk about how AI is changing their relationship to their work, not just whether they're using it efficiently, but what it's taking away as well as what it's adding.
AI is a real capability multiplier. What it multiplies is determined by the choices leaders make around it.
💬 Which of these feels most true in your organisation right now?
How's the depth of today's edition?
See you next week,
The It's Not the Work Team
If you’re new here, I’m over the moon you’ve joined us! To help me craft content that’s actually useful (and not just noise in your inbox), I’d love it if you took 1 minute to answer this quick survey below. Your insights help shape everything I write.
✨ Insane Media is more than one voice
💡 Dive into our other newsletters - where psychology meets the founders, creator economy, e-commerce marketing, and AI founders.
