AI at Work, Are we losing confidence, or building new muscles?

Lately I’ve noticed a new kind of email landing in my inbox. The name in the “From” field is familiar, but the message doesn’t sound like them at all. It’s flawlessly structured, immaculately polite, and somehow slightly…hollow. The tone is almost right, but not quite. It feels like I’m corresponding with the organisation’s AI policy rather than the human I know behind it.

At the same time, I notice my own habits changing. I draft something, then run it through an AI tool “just to refine it”. I double‑check things I already know. I can generate pages of copy for a website in minutes and then feel my brain lagging three steps behind, slightly dazed by the speed and volume.

So what is actually happening to our confidence – and our sense of voice – as AI becomes woven into everyday work?

The paradox: more help, less trust?

AI tools are brilliant at certain things. They can get us past the blank page, tidy up clunky sentences, offer alternative phrasings, and help us sound more professional across cultures and contexts. Used thoughtfully, they can level the playing field for people who haven’t had access to the same education, language or neurotypical advantages as others.

But there’s a shadow side that many of us are starting to feel:

• Messages sound polished, but strangely generic.

• We second‑guess our own words, even when we’re experienced and capable writers.

• We can produce far more content than anyone can meaningfully read, process or act on.

• Colleagues start to wonder, “Did you write this…or did something write it for you?”

  
  

AI can make us feel both more and less confident at the same time. More confident in the appearance of our writing, but less confident in the authorship of it. More productive on paper, but less connected in practice.

Three tensions to notice

Rather than deciding if AI is “good” or “bad”, it might be more useful to sit with a few tensions that many of us are living with right now:

1. Speed vs processing

   
   

   AI moves at a pace our nervous systems can’t naturally match. We can churn out documents, posts and emails at scale, but our capacity to think, feel, sense and integrate hasn’t changed. The risk is that we end up skimming along the surface of our own work.

   
   

2. Polish vs presence

   
   

   AI is very good at polish. It’s less good at the slightly messy, alive, contradictory tone that characterises real human presence. When everything becomes perfectly rounded, we can lose the tiny quirks and edges that tell people “this is actually me”.

   
   

3. Assistance vs authorship

   
   

   There’s a big difference between “AI as assistant” and “AI as ghost‑writer”. When the tool is supporting our thinking, we stay in the driving seat. When it’s replacing our thinking, we risk a slow erosion of skill, voice and confidence.

   
   

Reflection questions to gently audit of your AI habits

• When do I most often reach for AI e.g. ideas, first drafts, edits, summaries, translation?

  
  

• What personal “rules” am I already following, even if I’ve never named them?

• Where does AI make my life lighter e.g. less procrastination, clearer structure, fewer spirals of overthinking?

  
  

• What have I learned about my own style by seeing how AI mimics or “improves” it?

  
  

• What am I afraid will happen if I send something that is 100% mine, slightly rough edges and all?

• Which kinds of communication in my role really need to be in my own words (feedback, appreciation, difficult news)?

• Where might I deliberately choose to go slower, even if AI could go faster?

A closing invitation

AI isn’t going away. The question is not “Will we use it?” but “How will we be with ourselves while we do?”

Perhaps this is a moment to renegotiate the relationship between speed and depth, polish and presence, assistance and authorship. To remember that our slightly wobbly, honest, human writing still has a place and may be exactly what cuts through the noise.