Generative AI Isn't Perfect, But It's Fine If You Use It Well - Part 4

Last time,
we talked about how AI doesn't just start acting on its own.

This time,
"But I've heard it might say things that are wrong..."
I'd like to answer those kinds of concerns.

Actually, while generative AI is a very convenient tool, it is not perfect.

Sometimes, it might give you the wrong answer or
tell you a lie that sounds plausible.
But don't worry.

If you know "why that happens," you can use it well without being afraid.

In this article,

• Why generative AI says things that are wrong

• How to deal with "AI lies"

• Tips for using generative AI conveniently

will be explained in a gentle and organized way!

It only takes a few minutes to read, so please relax and read through it.

1. Why does generative AI give wrong answers?

Generative AI looks very smart, but in reality, it is "not actually looking for the truly correct answer."
What generative AI is doing is, based on a lot of information, "creating a plausible answer.".

For example, when you memorize a recipe and cook, have you ever accidentally used similar ingredients or improvised the amounts?

AI is a bit like that.

Based on information it has learned in the past, it thinks, "It's probably something like this," and puts it together in a plausible way.
However, sometimes incorrect content gets mixed in.

This doesn't mean the AI is trying to lie or trying to make a mistake on purpose.
It is simply "not perfect."

That is precisely why the step of "checking it yourself" is very important when using generative AI.

2. The reason why "plausible lies" happen

When interacting with generative AI, sometimes things come up that make you think, "Wait, is that really correct?"

For example...

• When I asked, "Tell me recommended picture books for a 3-year-old,"
  it introduced a picture book title that doesn't actually exist.
  

• When I asked, "Tell me a recipe I can use for a child's lunch box,"
  it suggested a menu with poor nutritional balance (all carbohydrates).
  

But the important thing here is that the AI is not trying to lie.

Generative AI is just creating a plausible answer based on patterns it learned in the past, thinking, "If this is the flow, maybe it continues like this?"

In other words, it's not making things up on purpose to lie, but rather, as a result of trying hard to "put things together plausibly," incorrect information sometimes happens to get mixed in.

What you can do to prevent this is simple.
That is, "Stop for a moment and think, is this really correct?"
Even if the answer looks natural at a glance,

• look it up a little

• read it carefully once more

Just by taking such small steps, you can significantly reduce the risk of being misled by lies.

3. Tips for using it well

Generative AI isn't perfect, but if you use it well, it can be a very reliable ally.
The important thing is to not expect a perfect answer.

The trick is to think of generative AI as a support role when you interact with it.
For example,

• Having it think of lunch menu ideas for your kids

• Having it roughly draft a LINE message for your mom friends

• Having it check a list of things to bring to daycare

Like this, first have the AI come up with ideas, then check and revise them yourself.

With this approach, you'll feel less anxious about "what if it's wrong?"

In other words, generative AI is a "tool that gives you a starting point for thinking."
You don't have to expect perfection from the start.

It's okay to have the mindset of "getting some help" or "thinking together."
If you can use it that way, you'll surely be able to use it more easily and with more peace of mind.

4. Summary

Generative AI is very convenient, but it is not perfect.
Sometimes it gives wrong answers or says things that sound plausible but aren't true.

However, that's not because the AI has malicious intent, but because it's trying its best to summarize things based on patterns, thinking "maybe it's like this?"

The important thing is "not to expect a perfect answer."
First, have the AI come up with ideas, then check the content yourself and revise it if necessary.
If you interact with it that way, generative AI will become a very reliable support role.

Instead of being afraid, know it, adapt to it, and use it well.
Just doing that might make your daily life much easier.

Next time, I will gently summarize
"Things to be careful about when using AI (Introduction to Literacy)"
!

If you read this article and thought,
"I want to know more about AI!"
"I want to try using it!"
then I would be happy if you could
like and follow me😊 See you in the next article!