AI can build you an app with zero coding experience. It can help you launch a website, brainstorm business ideas, polish awkward emails, and maybe even stop you from sending that passive-aggressive message to your coworker.
It can help you think clearer, move faster, and work smarter.
It can also confidently lie to your face.
That’s the part people forget.
We’re living in a time where AI can be brilliant in one moment and dangerously convincing in the next. Which means the real skill isn’t using AI. It’s knowing how to use it.
And that’s what this post is about.
Not hype. Not fear.
Just a grounded look at how I use AI as a physical therapist, runner, and writer — where it genuinely helps, where it absolutely doesn’t, and why mindset matters more than the tool itself.
When AI Looks Smart… But Isn’t 🧐
A study by Tordjman et al. (2026) found that 17 radiologists from 6 different countries struggled to reliably distinguish real X-rays from AI-generated fake ones.
Let that sink in.
In the first round, many didn’t even realize fake scans were mixed in. Only 7 out of 17 radiologists (41%) recognized that AI-generated scans were part of the dataset — less than half. Funny enough, experience didn’t seem to matter either: the group ranged from newcomers in the field to clinicians with 40 years of experience.
In the second round — when they knew fake scans were present — accuracy improved, but mistakes still happened regularly. Even with that warning, only 75% were identified correctly, meaning roughly every fourth scan still fooled the radiologists. Once again, level of experience didn’t seem to make a meaningful difference.
Even more ironic?
The AI model that helped generate the fake images wasn’t perfect at spotting them either.
That should concern all of us.
Because fake medical images aren’t just a funny tech story. They could be used for:
- insurance fraud
- manipulated assessments
- contaminated research databases
- false claims with real-world consequences
AI is improving fast. Faster than most people’s critical thinking.
And that’s the real risk.
Things You Need to Understand Before Using AI ⚠️
AI is not a replacement for judgment.
It’s not wisdom. It’s not expertise. It’s not accountability.
It’s leverage.
Used well, it saves time, expands options, and helps you think more clearly.
Used badly, it accelerates bad decisions.
That’s why I use AI in very specific ways:
- as an objective training coach
- as a thinking partner in practice
- as a writing assistant
- as a tool for reflection and challenge
But never as the final decision-maker.
Critical thinking is the key here. If you’re unable to question what you’re being given, you can end up believing everything — and that can backfire hard. I let AI give me suggestions, perspectives, and options, but the final decision was always mine.
That part stays human.
Always.
Using AI as My Running Coach 🏃🏻♀️
As a physio, I understand training principles, injury management, progression, recovery, biomechanics, nutrition, and load management. I’ve also spent a lot of time researching training methods, running drills, and ways to improve and adapt my running form over the years.
Sounds useful, right?
It is.
Until you’re the injured one. Because suddenly, all that knowledge doesn’t really help you the way you hoped it would.
Because when you are the athlete, emotions enter the room.
Fear. Impatience. Frustration. Overconfidence.
The classic “I’ll just push through it.”
I was managing a likely tibialis posterior tendinopathy while preparing for a half marathon. I had the knowledge. I had a training plan, and I had a clear goal in mind: running 21k in under 2 hours without my tendinopathy killing my progress — or the goal itself.
What I lacked was objectivity.
That’s where AI helped.
1. It Helped Me Think Clearly 🔎
I wasn’t objective. I couldn’t be. I couldn’t just tell my brain to stop doubting, stop being frustrated, or stop fearing that I’d miss my goal because of another flare-up. Although I understand the theory of rehab, training, and load management, I still couldn’t switch off my emotions.
I tried to treat myself like I would treat one of my patients. It helped a little — but it wasn’t the same.
When pain flared up, I could spiral.
Should I train harder?
Should I rest?
Am I losing progress?
Am I being lazy?
Am I being stupid?
AI helped me slow down, review the facts, and challenge emotional thinking.
Not replace my brain. Just calm the noise.
2. I Fed It Real Data 📊
Before AI could help me, it needed context.
I didn’t just ask, “Hey coach, what should I do this week?” and hope for magic.
I gave it real data from previous runs — my longer efforts from autumn, interval sessions, threshold runs, pace, heart rate, how I felt during the run, how my ankle reacted, even what the weather was like. The more complete the picture, the more useful the feedback became.
I had already started week one of my half-marathon prep, so I fed that in too: what sessions I’d done, how recovery felt, what my gym training looked like, how the tendon responded, where I felt strong, where I felt off.
Basically, I gave it everything I’d normally look at myself — or everything I’d ask one of my patients about, or tell a real coach if I had one.
Because that’s the truth with AI: the better the inputs, the better the outputs.
Give it vague nonsense, and it’ll often return polished nonsense.
Give it useful data, and now you’ve got something to work with.
3. I Decided Last 🎯
This is the most important part of all.
I never followed AI blindly.
Yes, I wanted objectivity. Yes, I wanted something outside of my own emotional headspace. But that doesn’t mean I handed over the steering wheel.
AI could suggest adjustments, review my data, and help recalibrate the next training week — but the final call was always mine.
Sometimes I felt more tired than expected and skipped a session completely. Sometimes I dialed a run down because recovery clearly wasn’t there. At one point I got sick, and when I started training again, I knew I wasn’t ready yet. So instead of forcing myself through a session just because “the plan said so,” I pulled back and gave myself more time.
That matters.
Because objectivity doesn’t mean ignoring your body. It means learning to separate panic from useful signals.
There were days where I wanted to do more because I was impatient. There were days where I wanted to do less because I was doubting everything. AI helped me step back, look at the facts, and find the middle ground.
It gave me structure.
I brought the experience, intuition, and honesty.
That combination worked far better than either one alone.
Using AI as My Co-Therapist 🧠
After seeing how useful AI could be in my own training, I became curious how it might help me in practice.
Not to replace clinical reasoning. Not to diagnose people from a chatbot prompt. Not to become some weird robot clinic.
I wanted to remove friction from things I already knew how to do.
So I started small.
I used it with messy notes, rough ideas, treatment hypotheses, progressions, and documentation drafts (all anonymous, of course!). Things that already existed in my head — but needed structure, refinement, or a second angle.
Same hierarchy as before:
I’m the physio. AI is the assistant.
It doesn’t assess the patient. It doesn’t build trust. It doesn’t notice fear in someone’s face when they load a knee for the first time again.
That part is human work.
But for support tasks? It can be surprisingly useful.
1. It Challenges My Thinking 🔍
Sometimes when you work with patients regularly, you become very efficient — which is good, but it can also make you predictable.
You see patterns quickly. You follow instincts. You know what usually works.
That’s valuable.
But sometimes “usually works” can blind you to other options.
So I use AI to challenge my own thinking.
I’ll throw in my observations and ask:
- What else could explain this?
- Which tests would help measure progress better?
- What am I maybe overlooking?
- Are there other exercise routes worth considering?
- What prognosis would be realistic here?
Not because AI knows better.
But because having another perspective can sharpen your own.
And sometimes it reminds you of something obvious that somehow slipped past your tired Tuesday brain.
2. It Helps Create Structure 📈
A lot of good therapists carry systems in their heads.
We know when someone is ready for the next step. We know what milestone matters next. We know when confidence is the missing ingredient, not strength.
But translating that instinct into a clean, measurable roadmap?
That can get messy.
AI can help organize the path.
Not by telling me what matters most — it often isn’t great at that — but by helping turn ideas into something clearer, simpler, and easier to communicate.
That helps patients too.
Because patients don’t just need exercises.
They need clarity.
They need momentum.
They need to understand why they’re doing what they’re doing.
3. I Decide Last — Again 🧭
Just like with my running, I decide last.
AI can give me ten exercise ideas.
Only I know which two actually fit the person in front of me.
AI can suggest progressions.
Only I know whether that patient is physically ready, mentally ready, or secretly terrified.
AI can generate structure.
Only I know whether it makes sense in real life.
That’s why I see it as a thinking partner, not a thinking replacement.
AI is a force multiplier. Point it in the right direction, and progress gets faster and easier. Point it in the wrong direction, and it accelerates mistakes just as fast.
Use it to get unstuck.
Use it to challenge yourself.
Use it to generate options.
But don’t outsource judgment.
That’s still your job. You still know your Movement Potatoes best. 😉
Using AI as My Writing Buddy ✍️
This is actually where it all started.
Before training plans, before patient notes, before all the practical stuff — I used AI for writing.
Because I love ideas.
I love noticing patterns, having insights, connecting dots, writing first drafts when something clicks in my head.
What I don’t love?
Spending twenty minutes wondering if a paragraph sounds smart, weird, too emotional, too harsh, too long or too soft.
That middle part used to drain me.
1. It Lowers the Barrier to Starting 🚪
Sometimes I know exactly what I want to write.
Sometimes I just have a feeling. A sentence. A moment from practice. Something a patient said that sticks with me.
Enough to feel interesting.
Not enough to become a full blog post yet.
That’s where AI helps.
I can throw in rough notes, half-thoughts, bullet points, messy emotion — and suddenly I’ve got something to react to.
Now I’m not creating from nothing anymore.
I’m shaping something.
And that’s always easier.
2. It Helps With the Boring Middle 🛠️
I still come up with the ideas. I still write the first draft. I still decide where the post goes.
That’s the first 10% of writing. That’s all me.
Now AI helps with the middle 80% — the part that’s super boring for me:
- flow
- grammar
- structure
- cutting repetition
- transitions
- making chaos readable
Honestly? That matters.
With a full-time job, my own practice, training, responsibilities, and a dog with aggressive opinions about walk times, I don’t have unlimited hours.
AI helps me keep creating consistently.
3. Final Filter: Does This Still Sound Like Me? 🎙️
This is the last 10% — and again, that part is done by me.
It’s the non-negotiable part.
Before anything gets published, I read it and ask:
Does this still sound like me?
Is the message still honest?
Did it become too polished?
Did it remove the edge?
Did it turn me into a LinkedIn motivational candle?
If yes, I rewrite it.
Because I don’t want AI to become my voice.
I want it to support my voice.
Big difference.
My Honest View on AI 💭
AI can be incredibly useful.
It can save time, sharpen ideas, reduce friction, and help you think more clearly.
It can also flatter you, mislead you, and confidently produce nonsense.
That’s why mindset matters more than prompts.
If you use AI to avoid thinking, it weakens you.
If you use AI to think better, it can strengthen you.
That’s the line.
And most people don’t notice when they cross it.
📝 Key takeaways
- AI works best as a tool, not an authority.
- Better inputs usually lead to better outputs.
- In training, AI helped me become more objective.
- In practice, AI helps challenge ideas and create structure.
- In writing, AI helps polish without replacing creativity.
- Human judgment is still the real skill.
- If you stop questioning, you start following.
- Use AI to think better — not less.
One thought on “🤖 How to Use AI Without Losing Your Mind.”