Last week I saw two headlines that made my head spin: “Mass Hysteria. Thousands of Jobs Lost. Just How Bad Is It Going to Get?” and, a few days earlier, “Can A.I. Be Pro‑Worker?” It’s dizzying.
Here’s one reframe that helps me breathe: the pace of AI capability isn’t the pace of human adoption. Headlines report AI’s pace; our calendars reflect our actual pace. The anxiety isn’t only about the tech—it’s about how our brains process it.
Four common biases are quietly turning up the volume:
- Availability heuristic: Extreme headlines are easy to recall, leading us to overestimate extreme outcomes. A layoff story and a takeover prediction stick, while the slow, boring reality of workflow improvements doesn’t.
- Negativity bias: We weight threats more than gains. That tilts us toward doomscrolling and skipping simple experiments that could actually simplify our day.
- Exponential growth bias: We watch jaw‑dropping capability demos and assume adoption will climb just as fast. It won’t. Organizations move in S‑curves, with real friction—budgets, training, process change, compliance.
- Illusory truth effect: When a claim repeats (“Half of jobs gone soon”), it feels truer each time, regardless of evidence quality. Repetition turns speculation into “fact,” and drives our anxieties.
A quick story
Recently, I was having a conversation about a simple role‑mapping exercise in which AI could help us analyze our own duties and flag tasks where human judgment remains central. Then I was asked about “the singularity.” That leap—from mapping tasks to imagining overlords—wasn’t about the exercise; it was about our collective cognitive load. We’re swimming in headlines and forecasts about AI capability. But capabilities speed ≠ adoption speed. Humans (especially in organizations) adopt much more slowly, with a predictable psychology of change. If you’ve read Bridges’ Managing Transitions: Making the Most of Change, you know the arc: endings, neutral zone, new beginnings. That arc exists, even when the tech curve is steep.
How do we steady ourselves and keep learning?
Two counterweights help:
Base‑rate thinking: Instead of treating AI as a unique technology, compare it to prior waves—automation, mobile, cloud, even past internal rollouts. How long did it take your org to adopt the last big tool? What stages did you skip at your peril?
- Action: Reflect on your own experience. What is an “adoption reality check” when you think about past waves of personal tech. It might be the the switch to ATMs instead of bank tellers, CDs to streaming platforms, or Friendster to Facebook. This doesn’t minimize change; it contextualizes it.
Growth mindset—with constraints: Treat AI as a skill‑building practice inside safe‑to‑try limits. Set small goals, measure, and actively seek and reward yourself for finding what doesn’t work.
- Action: pick a software to try. Define success metrics (time saved, quality, error rates, lessons learned) and schedule a stop‑or‑scale decision. Constraints protect you from panic buying and from quietly letting risky experiments sprawl.
If you’re feeling overwhelmed, try one of these for the next month or two:
- Pick one bias to watch. When you catch it, write the headline down, then ask: where have I seen this before? What was the impact on my role and organization then? What might be similar today?
- Pick one counterweight to practice. Add an “adoption reality check” question in your next AI plan or proposal, or run a tightly scoped pilot with a clear end date.
Sometimes we have to go slow to go fast. Calm is a competitive advantage. Make space to learn, test, and adopt intentionally—and let the headlines be headlines while you build what’s real.
Free resource: Go to https://repivot.notion.site/news for the latest AI @ Work headlines