AI is Great, But It Takes Time

One of the most awkward parts of building my AI garden visualizer was not actually the AI itself — it was the waiting time.

The image generation API I used could take close to a minute to return a result. From a developer’s perspective, the obvious solutions might be

• A loading spinner.
• Parallel processing
• Try to get the API to return progress during the wait
• Experiment with 3 different other models to see which one is the fastest
• Simplify the prompt, just show trivia changes
• Other clever solutions...??

But none of these is ideal - some of them require days of work and research. But - I haven't even got any users for the app!

Is There Another Way?

Most homeowners using this app are not garden experts. Many feel overwhelmed by messy backyards, overgrown plants, drainage issues, or simply not knowing where to begin. They do not necessarily want to study gardening for weeks. They want quick, practical guidance that helps them feel more confident immediately.

So instead of trying to “hide” the waiting time, I decided to use it.

Rotating Tips

During the AI processing phase, the app now displays simple illustrated garden tips every few seconds — almost like a lightweight PowerPoint presentation. Each screen is designed to be scannable within 7–10 seconds:

• Why mulch matters
• How tree drip lines work
• Why plants fail when planted too close together
• Simple before/after landscape ideas

Something like this,

Then use setInterval to display them

    // use setInterval() to display an garden tip every 7 seconds
    window.setInterval(() => {
      setTipVisible(false);

      // use setTimeout to fade in/out 
      timeoutId = window.setTimeout(() => {
        setTipIndex((prevIndex) => getRandomTipIndex(prevIndex));
        setTipVisible(true);
      }, 500);
    }, 7000);

The interesting thing is that users no longer feel like they are waiting. They are learning.

Making the App More Helpful

And I think this taught me an important lesson about software development: sometimes the best solution is not deeper engineering complexity, but changing perspective. Instead of asking, “How do I technically eliminate the delay?”, I started asking, “What would make this minute genuinely useful for the user?”

That shift completely changed the experience of the app.

Ironically, the long API call became an opportunity to strengthen the product’s identity. The app stopped feeling like a simple image generator and started feeling more like a helpful gardening companion.