A new competitor arrived

Something crazy happened yesterday.

I was back in the store after being away for four weeks and noticed a shift. A few customers started questioning our advice, saying they’d been told something else.

By who?

ChatGPT.

That’s right.

Runners have started asking AI for help picking running shoes.

At first, I laughed.

But then I remembered a recent consult I had: This runner told me that he was being recommended a shoe that he ended up hating. It turned out the shoe was on promotion and likely pushed for the store’s benefit, not his.

If runners are losing trust in retail because certain brands are being pushed for financial benefits, I totally understand why they’re looking elsewhere.

So… my curiosity was piqued.

I wondered: How well can ChatGPT actually do for me?

The moment of truth

I uploaded a rear video of myself running on a treadmill and gave it this prompt:

It paused for a moment, then decided to analyze my form based on 5 key metrics:

• Foot strike pattern

• Knee and hip alignment

• Arm swing symmetry

• Cadence and posture

• Pelvic and trunk stability

It extracted five key frames from the video and began its analysis.

🧠 Gait Analysis: ChatGPT’s Observations

1. Foot Strike Pattern
   Midfoot striker — efficient and typical for experienced runners. Minimal heel contact = good proprioception and strength.

2. Stride & Overstriding
   Foot lands close to the center of mass — not overstriding. Great for injury prevention and efficiency.

3. Knee Drive & Hip Extension
   Solid knee lift and hip extension. Suggests strong propulsion and good use of the posterior chain.

4. Posture & Trunk Control
   Upright trunk, slight forward lean from ankles (ideal). Pelvic stability looks good — no excessive tilt or hip drop.

5. Arm Swing
   Symmetrical and compact — supports forward momentum. No noticeable cross-body motion.

I was impressed. Sure, it missed a few things, like my slightly wide elbows and minor hip rotation during toe-off, but honestly, for an AI using just five frames, the feedback wasn’t bad at all.

ChatGPT summarized my current favorites like this:

👟 Shoe Preferences: AI’s Take

1. Endorphin Speed 3 – lightweight, versatile

2. Vaporfly 2 – race day, responsive

3. NB 1080v14 – max-cushion daily trainer

4. Evo SL – superlight and aggressive

It noted a clear preference for performance-forward shoes, especially lightweight and plated models.

Again: solid analysis.

Here’s what it suggested for short races:

✅ Race Day Shoe Recommendations

1. Adidas Adios Pro 3 – snappy toe-off, stable forefoot

2. Saucony Sinister – ultralight, great for 5Ks

3. Nike Streakfly – fast for 5–10K, less aggressive than Vaporfly

4. Asics Metaspeed Edge+ – ideal for higher cadence runners

5. Puma Fast-R Nitro Elite 2 – aggressive, bouncy, and stable

Here’s where I start to disagree with ChatGPT.

These shoes are old. Some are nearly impossible to find anymore. So the shoe advice felt outdated, which makes sense, since it likely pulls from a data cutoff that doesn’t include the latest released running shoes.

Final Thoughts

The gait analysis? Surprisingly decent.
The shoe advice? Outdated.
Overall? Pretty impressive for a free to use AI tool.

Of course, this is just an N=1. But it makes me wonder how AI will shape running retail in the future. I’ll definitely look deeper into markerless motion capture and computer-based gait analysis.

Would I trust ChatGPT to pick my next shoe? Not yet.
But would I use it to cross-check what I’m told in-store?
Honestly… maybe.

Thanks for reading.

-Tim 👟

P.S: Not ready to trust AI with your gait just yet? I offer personalized consults if you’d prefer a real human to take a look and guide you through the right shoe choice.