AI is an easy tool that is hard to master

After using AI every chance I could over the last two years, here's my conclusion as a professional writer: It is hard to master for writing well.

Someone recently approached me with a collaboration request. Would you want to develop a guide on AI and its use for tech communication?

Strangely, I came up blank.

How do you explain riding a bicycle?

Grasp AI by its horns

When ChatGPT came out, I was deeply worried. I recognised that the bar for "average" writing has been irrevocably raised.

And despite its many flaws, AI will only get better.

I've always advocated grasping challenges by their horns. So I did exactly that, using AI across the wide range of content I get paid to write daily.

• Editing.
• Blog posts.
• White papers.
• Brochures, websites.

So how has AI helped in my writing?

A tale of two stories

I had two stories to write, based on interviews I conducted with experts in their respective fields.

For the first piece, I used AI to answer highly specific questions in bullet point form. I then had it write them up into short blocks. I selected the most promising and stitched them together into a cohesive piece, then edited it further. I loved the result.

• Estimated time saved: 1 hour.

Enthused with the first story, I used the same technique for the second piece.

But it came out wrong: The AI was making inferences that were just a step too far for a niche, highly technical topic. In the end, I had to extensively rework everything to make it work.

• Time saved? None.

Frustration aside, I estimated I spent an extra 1.5 hours than if I were to skip the use of AI entirely.

The lesson here? AI use is both a skill and an art. And AI expertise lies in one's ability to apply the right mix for each scenario.

Learn techniques, not prompts

How do you explain riding a bicycle?

Simple. You don't.

• Get tips from someone who can do it.
• Have them hang around.
• You do it.

To be truly proficient with AI, one must develop the skill by actually doing it. Not by rote learning or applying rigid AI prompts developed by someone else.

Advice from experts can help, but it'll be far more helpful if they are versed in cycling, and not by a unicyclist or someone who needs training wheels.

As you train yourself on various AI techniques, you'll soon gain the competency to use the right strategies on the right occasions - and also know when not to use it.

And that's how you will master generative AI.

How are you doing on your AI learning journey?