Table of Contents
The European Union is moving towards new legislation that would require makers of AI tools like ChatGPT to disclose the materials used to build their systems.
This would give publishers and content creators a new weapon to capture a share of the profits when their businesses are bought for AI-generated content like ChatGPT.
However, the new legislation opens the door to questions about the source of the “GBT Chat” information, especially as it has sparked a race with “Google” and other companies to introduce so-called artificial intelligence tools, calling into question the accuracy and validity of your information.
Wikipedia
ChatGPT is an AI language model that is trained on a large body of text from a variety of sources such as Wikipedia, books, news articles, and academic journals.
It’s also important to understand that he’s not accessing a database of facts to answer your questions, instead his answers are based on patterns he saw in the training data, according to scribbr.
At the same time, and to find out about the sources used, a journalist from the ILLUMINATION site, specializing in technical matters, conducted an experiment with “Chat GPT” earlier this year, to ask him about his sources.
According to the intelligent robot, there are ten examples of publicly available sources from which it could draw its knowledge.
These include Wikipedia and news portals such as online platforms that publish news from around the world, as well as government and agency websites.
Scientific books and journals
And scientific journals such as professional journals in which scientific articles are published, as well as books such as those printed or electronically published which summarize knowledge and information on various topics.
GBT Chat also draws its information from textbooks, documentaries and audiovisual media.
and online courses such as online e-learning offerings, which impart knowledge and information on a variety of topics.
Communication platforms
Finally, among these sources are social media, such as Facebook, Twitter and Instagram platforms, where users can share and discuss information and knowledge.
Under new provisions added to the EU’s draft AI law, developers of AI models will have to publish a sufficiently detailed summary of the copyrighted material they used as part of their creation, says the project.
