A few weeks ago, I bought a 22-year-old Honda Deauville motorcycle.

My intention was simple. I wanted a reliable bike for exploring Ireland and filming episodes for my YouTube channel, Thinking on Two Wheels.

What I didn’t expect was that the motorcycle would become a teacher.

Over the past few weeks, as a complete beginner with no mechanical experience, I have changed the final drive oil, changed the engine oil and filter, replaced spark plugs, changed the coolant, investigated a cooling fan problem, tested electrical circuits, repaired a multimeter, and spent more time than I care to admit staring at bolts, wires, and components wondering what to do next.

The experience taught me something interesting about learning in 2026.

The Cost of Being a Beginner Has Fallen

Twenty years ago, I probably wouldn’t have attempted most of these jobs.

I would likely have taken the bike to a mechanic.

Not because I lacked interest, but because the barriers to learning practical skills were much higher.

Today, those barriers have fallen dramatically.

The first thing I discovered was the value of YouTube.

Need to remove a fairing?

Someone has filmed it.

Need to change coolant?

Someone has filmed it.

Need to replace spark plugs on the exact same motorcycle?

Someone has filmed it.

You can pause, rewind, watch again, and learn at your own pace.

It is like having thousands of teachers available whenever you need them.

The second thing I discovered was how easy it has become to source parts and tools.

For the Deauville, I ordered spark plugs, filters, a cooling fan switch, electrical connectors, wire, heat shrink tubing, and specialist tools online.

The parts were matched to the exact make and model of the motorcycle and delivered directly to my door.

I found myself ordering things that I didn’t even know existed a few weeks earlier.

At one point I needed piggyback spade connectors for some wiring work.

A month ago, I wouldn’t even have known what a piggyback spade connector was.

I can imagine myself walking into a hardware store trying to describe it:

“It’s sort of a connector thing for a motorcycle…”

Online, it was different.

I could search, compare photos, ask questions, check compatibility, click a button, and have exactly what I needed delivered a few days later.

In the past, lack of knowledge could stop you before you even started.

Today, many of those barriers are disappearing.

AI as a Learning Companion

Perhaps the biggest surprise was how often I found myself consulting AI.

Not because AI was doing the work.

It wasn’t.

I still had to pick up the spanners, remove parts, make mistakes, and solve problems.

But whenever I got stuck, I had someone—or something—to ask.

Questions like:

• Is this the correct bolt to remove?
• Why won’t this oil filter come off?
• What does this mechanism do?
• Is this coolant level correct?
• What should I test next?
• How do I use a multimeter?

Before buying the Deauville, I owned a multimeter but had no real idea how to use it.

A few weeks later, I was measuring battery voltage, testing switches, checking continuity, and troubleshooting electrical circuits.

AI didn’t diagnose the motorcycle for me.

It taught me how to use the tool that helped me diagnose the motorcycle.

That feels like an important distinction.

Knowledge Is Becoming Conversational

Before starting this project, I bought a technical manual for the Honda Deauville.

I expected it to become my primary source of information.

Surprisingly, I barely used it.

Not because the information wasn’t there.

It was.

The problem was that when you’re standing beside a half-dismantled motorcycle with oil on your hands, you don’t want to search through hundreds of pages looking for an answer.

You want the answer to your question.

So I found two workshop manuals online and used them to create a custom GPT specifically for the Honda Deauville.

Suddenly I wasn’t searching manuals.

I was having a conversation with them.

I could ask questions in plain English and receive answers grounded in information specific to the bike.

The manuals did not disappear.

In fact, they became more useful than ever.

AI wasn’t replacing the manuals.

It was helping me access the knowledge inside them.

Technology Doesn’t Have All the Answers

One of the most important lessons was that no single source of information was enough.

Sometimes YouTube was more useful.

Sometimes AI gave a clearer explanation.

And sometimes the answer emerged from combining several sources.

For example, AI suggested that I would need to remove the right fairing to access one of the spark plugs.

A YouTube video showed someone changing the same spark plug without removing the fairing.

The video was right.

The AI was wrong.

On another occasion, my oil filter refused to come off.

AI didn’t provide the solution, but a YouTube video gave me the idea of using a strap for additional leverage.

Looking around the garage, I spotted a ratchet strap I already owned.

I adapted the idea, hooked the ratchet strap around the filter, and used the extra leverage to break it free.

What struck me afterwards was that the solution came from several sources at once.

YouTube provided the spark of an idea.

The ratchet strap was already sitting in my garage.

The final solution emerged through experimentation and persistence.

Learning often works like that.

We gather ideas from different places, combine them with the resources available to us, and create a solution that nobody explicitly taught us.

Technology can help us think.

It cannot think for us.

Learning Takes Time

One mistake beginners often make is expecting things to happen quickly.

My experience with the Deauville taught me the opposite.

The oil filter gets stuck.

A bolt refuses to move.

A spark plug socket won’t fit.

The fan still doesn’t work after you’ve changed the switch.

Progress is rarely linear.

What helped most was slowing down.

Don’t rush.

Don’t force things.

Don’t damage the bike.

Pause.

Think.

Watch another video.

Ask another question.

Order the correct tool.

Try again tomorrow.

The goal is not to finish quickly.

The goal is to learn.

Patience became part of the toolkit.

Learning to Troubleshoot

Perhaps the most valuable thing I learned was not maintenance.

It was troubleshooting.

When the battery initially went flat, my first instinct might once have been to replace it.

Instead, I tested it.

I measured the voltage.

I checked the charging system.

I compared the readings with expected values.

The evidence suggested that both the battery and charging system were working correctly.

The same thing happened with the cooling fan problem.

Rather than randomly replacing parts, I learned to ask questions.

What do I know?

What can I test?

What does the result tell me?

What should I test next?

The multimeter became one of the most valuable tools in the garage.

Not because it gave me answers.

But because it helped me ask better questions.

At one point, the multimeter itself stopped working.

One of the probes had broken internally.

Rather than replacing the meter, I opened it up, found the fault, soldered the connection, and got it working again.

It struck me that I wasn’t just learning motorcycle maintenance.

I was learning how to investigate problems, diagnose faults, and repair things.

Confidence Through Capability

The biggest thing I gained from this experience wasn’t mechanical knowledge.

It was confidence.

A few weeks ago, changing brake pads, diagnosing electrical faults, or working on the cooling system felt intimidating.

Today, I still wouldn’t call myself a mechanic.

But I no longer see these jobs as impossible.

I see them as learnable.

And that is a powerful shift.

I now have a process.

Watch.

Ask.

Read.

Test.

Learn.

Adapt.

Repeat.

As I finished putting the motorcycle back together, I discovered there was one screw missing.

Fortunately, I knew exactly where it belonged.

After consulting my AI mechanic assistant—the custom GPT I had created using the Deauville manuals—I identified the correct replacement and ordered one online.

A small problem solved.

A few weeks earlier, that missing screw might have felt like a major setback.

This time it felt like one final reminder of the lesson the motorcycle had been teaching me all along.

Learning isn’t a straight line.

Things go wrong.

Parts get stuck.

Tools break.

Mistakes happen.

Progress is messy.

But if you stay curious, slow down, and keep going, you eventually get there.

Looking Ahead

Looking ahead, I suspect we are only at the beginning of this transformation.

AI will become more capable.

Delivery systems will become faster.

Technical information will become more accessible.

Digital assistants will become more specialised and knowledgeable.

The barriers to learning will continue to fall.

The people who benefit most may not be experts.

They may be ordinary people who are willing to remain curious, ask questions, and learn by doing.

The motorcycle runs a little better now.

But perhaps the biggest thing that changed wasn’t the bike.

It was me.

A 22-year-old Honda reminded me that in a rapidly changing world, one of the most valuable skills we can develop is confidence in our ability to learn.

And that confidence may be worth more than any individual skill we acquire along the way.