The creators of ChatGPT have released a new feature which helps teachers detect if text was produced by a student or artificial intelligence. This will be important for schools to determine if a student cheated or not.

Can This Be Trusted?

ChatGPT has become a viral tool, and so much so that it is being banned at schools for excessive use. It's not the greatest at detecting if text is AI generated or not at the moment, but this feature should get better over time.

The feature to detect if text is AI generated will give a chance that the text was AI generated. It will also give a chance that it was human generated. And it can be WRONG. It can say that human text was generated by AI, which could lead to some false accusations.

It's likely to become a game of Cat and Mouse. Tools like ChatGPT are going to improve and get better, and the mechanisms to abuse them will also get more sophisticated. Students will find ways to use it even if it is banned.

There will come a time in the future where the AI will be so good at generating text that it will become indistinguishable from human text. When this happens, then I think we will be in for a big debate about what the point of having school is.

Will This Be Useful to Schools?

I think ChatGPT should be embraced and have lessons built around it. It should be just like the calculator was for students in the beginning and be like an assistant. I think this enables students to utilize the tool just like any other tool available to them.

ChatGPT doesn't always site its sources, which is something I think it needs to start doing. It needs to show a list of websites and sources, just like Wikipedia does at the bottom of its information. At least say where you got the information.

The other challenge of usefulness is that ChatGPT is built by people and can contain biases. Hopefully these get evened out over time, but with any Open Source system, you can have someone contribute that puts a "slant" or "bias" on something that could be a problem for schools.

For example, ChatGPT could be biased toward a particular technology or way of doing something or paint a negative picture about someone a student is researching, and that student would simply regurgitate what the AI is saying. This takes away that student's opinion and could cause opinions to form that aren't really that students.

I don't see ChatGPT stopping anytime soon. It's not going anywhere, and we are going to have to get used to it being around.

Do you see ChatGPT as always going to be here - or will something better come along?