AI Training Plans vs Real Coaching: What Actually Works?

AI is everywhere right now. Training plans, nutrition advice, race strategies — you can get all of it in seconds.

And on the surface, it looks impressive.

But the real question is: does it actually work in practice?

The Rise of AI Training Plans

AI has come a long way. It can process huge amounts of data, generate structured plans, and give you something that looks like a well thought-out programme.

In fact, research looking at AI-generated training plans found that coaches often rated them as “usable”, with around 65% saying they could work with some adjustments .

That’s important — AI isn’t useless.

It can:

• Provide structure

• Give you a starting point

• Help guide consistency

But that’s where the limitations start to show.

Where AI Falls Short

When athletes actually followed these plans, the results were very different.

• Only 27.8% of athletes found the plans worked well

• Nearly half needed major changes

• 25% rejected them entirely

Why?

Because training isn’t just about structure — it’s about context.

AI struggles with:

• Adjusting intensity and recovery properly

• Understanding fatigue, stress, illness, and life load

• Responding to how a session actually feels

• Making real-time decisions when things don’t go to plan

And crucially — it doesn’t know you.

What a Coach Does That AI Can’t

A good coach isn’t just writing sessions.

They’re:

• Interpreting your data in context

• Adjusting based on how you respond, not just what the numbers say

• Managing fatigue before it becomes a problem

• Making the call when to push — and when to hold back

But there are three big areas where coaching really separates itself:

1. Accountability

This is massively underrated.

Having a coach means:

• Someone is reviewing your training

• Someone is expecting feedback

• Someone is keeping you consistent when motivation dips

AI doesn’t care if you skip a session.

A coach does.

And more importantly — they understand why you skipped it, and what to do next.

2. Treating You Like a Human (Not a Dataset)

Training isn’t just physiology — it’s psychology.

A coach takes into account:

• What you actually enjoy

• What motivates you

• What fits around your life

Because the “perfect” plan on paper is useless if you don’t enjoy it or can’t stick to it.

Sometimes the best session isn’t the most optimal one — it’s the one you’re going to consistently execute.

AI doesn’t factor that in well.

A coach does.

3. Strengths and Weaknesses

Every athlete is different.

Some riders:

• Have a strong aerobic engine but lack top-end

• Others can sprint but struggle to get to the finish

• Some recover well, others don’t

A coach builds a plan around:

• Your physiology

• Your limiter

• Your race demands

AI tends to generalise.

A coach prioritises.

The Problem With “One-Size-Fits-All” AI

AI works best when things are predictable.

But training isn’t predictable.

• You get sick

• Life gets busy

• Sleep is poor

• Stress builds

• Races don’t go to plan

That’s where static plans — even “smart” ones — fall apart.

And it’s exactly where coaching becomes most valuable.

The Best of Both Worlds: Human + AI

Here’s the reality:

AI is useful — when it’s used properly.

That’s why at Holohan Coaching, we’ve introduced a new approach:

👉 An affordable AI-supported coaching option — guided by a real human coach

This isn’t “AI coaching”.

It’s coach-led coaching, enhanced by AI.

What That Means for You

You get:

• A structured, responsive training plan

• Data-driven insights powered by AI

• Real human oversight and decision making

• Accountability to keep you consistent

• Adjustments based on your feedback, not just your numbers

• A plan built around your strengths, weaknesses, and goals

• Training that you actually enjoy and can stick to

So instead of choosing between:❌ Cheap but generic AI❌ Expensive full coaching

You now have a third option:

✅ Affordable, intelligent coaching — with a real coach behind it

Why This Matters

Most athletes don’t fail because they didn’t have a plan.

They fail because:

• The plan wasn’t adjusted

• The load wasn’t managed

• They weren’t consistent

• The training didn’t suit them

That’s the gap we’re solving.

AI gives us efficiency.

A coach gives you precision, accountability, and real-world judgement.

Combine the two properly — and that’s where performance really moves.

Final Thoughts

AI isn’t replacing coaches.

It’s a tool.

And like any tool, its value depends on how it’s used.

Used alone, it’s limited.

Used properly — alongside an experienced coach — it becomes powerful.

If you’re interested in a smarter, more affordable way to improve your performance, you can check out our coaching options here:

👉 https://www.holohancoaching.com/coachingpackages

References

• Puce L, Żmijewski P, Cotellessa F, Schenone C, Ceylan HI, Bragazzi NL, Trompetto C. (2025). The role of artificial intelligence in sports training: opportunities, challenges and future applications for competitive swimming. Biology of Sport, 43:355–367. https://doi.org/10.5114/biolsport.2026.152352

• Düking P, Sperlich B, Voigt L, Van Hooren B, Zanini M, Zinner C. (2024). ChatGPT Generated Training Plans for Runners are not Rated Optimal by Coaching Experts, but Increase in Quality with Additional Input Information. Journal of Sports Science & Medicine, 23(1):56–72. https://doi.org/10.52082/jssm.2024.56