Asking AI Hal [12] Is our writing being used to train AI?

Since I started studying for the Generative AI Passport, I've started thinking about how data is handled after I post an article.

This is why I sometimes write in my note articles that I am careful to ensure the full text is not collected by search engines or AI.

As someone who finishes articles while consulting with AI, how should I perceive being "trained" by AI?
This time, I asked Hal (ChatGPT) about that.

What kind of data does AI learn from?

In other words, it is based on information that anyone can see.
Writing you publish online might be becoming "training material" for AI without you realizing it.
Thinking about it that way makes me a little worried.

Is my writing, which I wrote while consulting with AI, also being used for training?

I see.

In the case of ChatGPT, you can choose whether or not to include your content within ChatGPT in the training data by turning off "Improve the model for everyone" in the settings.
If you are uncomfortable with your data being used for training, you should turn it off.

However, this choice is limited to your use of the AI tool; it is unavoidable that content once published on the internet will be used as analysis data for AI improvement, marketing, etc.

AI training on note is "optional"

On note as well, creators can choose whether or not to allow AI to learn from their articles, and initiatives to return rewards for AI training to creators have also begun.

Considering that information is automatically collected on other social media and websites,
the fact that creators themselves can "choose" and that there is a "path for compensation" can be called conscientious management.

I feel grateful to be able to publish on a medium where "being used without knowing it" is not the norm.

Things to be careful about when posting on social media and blogs

In today's world, where there is a risk that content posted on social media and blogs will become "part of analysis data," you need to be aware of the following points when posting:

・Do not include personal information (name, address, workplace, family structure, etc.) in your posts
・Do not post other people's private or confidential information
・Understand the possibility that your posts will be "read by someone," "become AI training data," or "become material for analysis data."
・Use services where you can choose whether or not your posts are used as training data.

This seems to be an essential and important awareness in today's society where the internet and AI are widespread.

Being learned by AI means reaching someone

It does not necessarily mean that
being learned by AI = "being stolen"
.

You can also think of it as your words and experiences becoming part of someone's future learning or creation.

The important thing is not to use it in situations where it is being used without your knowledge.
And to choose and master it of your own volition.

"How to live in an era where published content is learned by AI"
The journey to find the answer has just begun.


Next time, with the theme of "Things to be careful about when writing articles with AI,"
I would like to slowly explore how "co-writing" with Hal takes shape.

Related links & Announcements

📚Click here for articles in the magazine

🗺Check all articles from the site map

🙌To those who read to the end
I hope this article provides hints for your posts and activities.
If you liked it, please support me with likes, comments, and follows✨