AI Helped Me Sound “Better” and Feel Worse

Does everything feel like too much these days? Get When Life Sucks: 21 Days of Laughs and Light for free when you join the Tiny Buddha list.

It was close to midnight the first time it really hit me.

I was sitting alone at my kitchen table, still in work clothes, phone in hand. I’d come straight home after a long day of back-to-back meetings, staff conversations, and one decision I’d been avoiding for weeks—a call that would affect someone’s role, their income, and their sense of security. By the time I got home, I was too wired to sleep and too tired to change.

The house was quiet.

On the screen was a chat window.

Not with a friend. Not with a therapist. With an AI.

I’d just typed out a long, messy paragraph about a staff issue, the weight of leadership, and the guilt of feeling utterly drained when my job is literally about caring for others.

“I feel like I’m failing everyone,” I wrote.

Within seconds, the reply appeared: calm, validating, beautifully worded.

“It’s understandable that you feel this way given the emotional load you’re carrying…”

Something in me relaxed. Something in me hollowed out.

Because during the day, I run a large mental health service. I’m the person others come to when they’re overwhelmed, scared, or stuck. I’m supposed to be the one who knows what to do, who can hold complexity without flinching.

But that night, I realized I’d quietly handed my own inner life over to a machine.

Not dramatically. Just one exhausted conversation at a time.

When “Help” Starts to Replace “Self”

From my vantage point, I see a strange double life playing out.

In meetings, casual conversations, and WhatsApp chats, I hear people say things like:

“I wrote my message in AI first so I didn’t sound too emotional.”
“I checked with a chatbot if I was overreacting before I replied.”
“Sometimes it’s easier to talk to it than to anyone else.”

Leaders, colleagues, friends, we’re all quietly doing the same thing.

We turn to AI to:

• Find the “right” tone so we don’t upset anyone
  • Make our feelings sound reasonable, not “too much”
  • Get quick answers when we’re too tired to sit with questions

It’s not evil. It’s not weak. It’s human to want reassurance, comfort, and confirmation that we’re doing the right thing.

But as I watched this pattern in people around me and then caught it in myself at midnight in my kitchen, I had to face something uncomfortable:

In trying to hold everyone else together, I’d stopped knowing what to do with my own feelings.

AI hadn’t created that problem. It had just made it easier not to notice it.

The Pattern I Couldn’t Unsee

Once I saw it, I started noticing the same theme again and again.

A manager used AI to soften a piece of honest feedback so it sounded “less disappointed.”

A friend used it to rehearse telling their co-founder they were burning out and couldn’t keep working at the same pace.

Someone else, a senior clinician I work with, used it to draft a message to me because they were terrified of saying the wrong thing about their workload and feared it might come across as ungrateful or unprofessional.

Underneath all of these moments was the same quiet fear:

“If I say it how I really feel, I might lose something—respect, connection, my job, my relationship.”

So we hand our words to a system that never flinches, never blushes, never gets triggered. It gives us back something smoother, kinder, more balanced.

And slowly, almost invisibly, we start to trust that more than we trust ourselves.

The more I saw this in others, the more I had to admit:

I had been doing the same thing with my own life, not for days or weeks, but for years. Each time I chose polish over honesty, regulation over truth, I moved a little further away from myself. Over time, it left me clearer in my head but increasingly disconnected from my body, my instincts, and my sense of what I actually wanted.

The Night My Friend Asked What I Was Avoiding

One evening, after a particularly heavy week, I was on a call with a close friend.

We often talk about the chaos of building things that matter, team issues, cash flow, complicated decisions, and the emotional hangover of responsibility.

I did my usual summary:

“It’s been a big week, but it comes with the territory. We’re growing, and it’s a privilege, and I’m grateful…”

He was quiet for a moment and then said:

“That all sounds very polished. How are you actually?”

I paused.

My first instinct was to give a tidy, measured answer, the kind that sounds good on a podcast or in an email newsletter.

Instead, I noticed my mind reaching for familiar phrases I’d seen on screens:

“It’s understandable that I feel…”
“On the one hand… on the other hand…”
“A more balanced view would be…”

They sounded wise. They did not feel true.

For a few seconds, I couldn’t actually find my own words.

I had become so used to expressing myself in careful, well-regulated language—for staff, for partners, for stakeholders—that I’d almost forgotten how to speak as a person, not as a role.

I couldn’t blame AI for that. But it had certainly helped me avoid noticing it.

That conversation left me with a simple, unsettling question:

When did I stop trusting my own voice?

What I Was Really Afraid Of

When I finally stopped long enough to listen underneath the polished language, I found a very simple fear:

“If I let myself be fully honest, everything might fall apart.”

If I admit that I sometimes feel overwhelmed, will my team trust me less? If I tell a friend I’m too tired to support them tonight, will they think I don’t care?

AI had become a perfect hiding place for that fear.

I could pour out my unfiltered thoughts without risking anyone’s disappointment. I could receive advice and validation without feeling like a burden. I could feel momentarily “held” without having to navigate anyone else’s reactions.

But after each conversation, I noticed something:

My head felt clearer. My body did not.

Because my nervous system didn’t actually need more perfectly formatted sentences.

It needed to know that my real, messy self was allowed to exist in front of people, not just in private chat logs.

Learning to Come Back to Myself (Without Pretending We Live without Tech)

I didn’t suddenly delete every AI app and move to a cabin in the woods.

I still live in a world where technology is everywhere, and I still use it in my work.

But I made a quieter promise to myself:

“I will use technology to support my humanity, not replace it.”

That meant changing a few habits.

First, I started checking in with myself before checking in with a system.

Before I ask any tool, “What should I say?” I ask, “What am I actually feeling right now?”

Sometimes I write it down plainly: “I’m scared this won’t work.” “I’m angry, and I don’t want to be.”

Only after I’ve named it do I decide if I want help shaping it. If I do, it’s there to refine my expression, not decide what’s acceptable for me to feel.

Second, I let humans back into the loop.

If something really hurts, I reach out to a person before I reach out to a machine. Sometimes it’s as simple as: “Today feels heavy. Do you have ten minutes later?”

It doesn’t always fix the problem, but every time I choose a human over a chat window, I send a message to my nervous system: I am not alone in this.

Lastly, I started protecting a few spaces where the unedited version of me is allowed to exist.

For me, that looks like:

• No AI help for important emotional conversations with people I’m close to
• No technology in the first thirty minutes after waking and the last thirty minutes before sleep
• No using AI to rehearse difficult personal conversations

These aren’t rigid rules. Some days I break them.

But having any spaces where my words are allowed to come out wrong has reminded me that I can survive imperfection and that the people who care about me can too.

If You’re Quietly Doing the Same Thing

Maybe your circumstances are different from mine.

Maybe you’re running a small business, a household, a team, a life that other people depend on.

Maybe you’ve noticed you’re more comfortable typing your rawest feelings into a box than saying them out loud.

If so, here’s what I wish someone had told me sooner:

You’re not strange for finding AI comforting. It makes sense to turn to something that feels safe and predictable when people haven’t always been that for you.

You’re not “less mindful” for using technology. The issue isn’t the tool, it’s whether you’re still in the conversation with yourself.

The parts of you that feel too heavy, too dramatic, or too complicated are often the exact parts that most need to be met by a real, breathing, imperfect human being, including you.

You don’t have to stop using every supportive tool. You don’t have to suddenly pour your heart out to everyone in your life.

You could start much smaller:

• One honest breath before you pick up your phone
• One sentence of truth in a conversation where you’d usually say, “I’m fine”
• One person you let see you before you’ve tidied yourself up

Closing

AI can help you organize your thoughts.

Only you can decide that your messy, unfiltered inner world is worth listening to.

And if you forget, because I still do, often, remember this:

Underneath the emails, the roles, the prompts, and the noise, there is still a quiet part of you that knows when something feels off, and when something feels true.

That part deserves more than a cursor blinking back at it.

It deserves you.

See more posts