Reasons for not using artificial intelligence in dream interpretation

Many current dream interpretation services show writings generated by inserting dream stories into generative models. Dreams-Link does not do this. There is no code that calls a model to create interpretations.

What we do is simple. We find the symbols in the text you provide and select the meanings that the dictionary has written about those symbols. There is no room for sentences that are not in the dictionary.

The dictionary itself is not created by a model. Each meaning is accompanied by which passage from the original dream interpretation text it comes from, and that passage is compared word-for-word with the original file.

Models do not say they do not know what they do not know. When asked about symbols without conveyed basis, they fabricate plausible origins. And whether it is fabricated or not is something the reader cannot discern. If creation is inserted in the place of conveying tradition, the premise of the service collapses.

We did try letting a model create symbols to expand the dictionary. Out of sixty-six examples selected as worthy of adoption, fifty-five could not provide any conveyed basis, and there were also examples like subway and highway that cannot exist in traditional dream interpretation. Therefore, none were included.

• If it is the same dream, the same interpretation will come out. The words do not change each time you look at it.
• It is fast. There is no waiting for the model's response, so the results are immediately available.
• The dream you wrote does not go outside. There is no need to send it to an external company's server — please read this together with the method that does not save.
• It can be offered for free. Dreams are something we have every day, so there are many inquiries. If a model is called for each inquiry, the cost must be covered from somewhere.

We cannot interpret what is not in the dictionary. If we had used a model, there would have been a plausible answer for whatever you wrote. We chose to say that we could not find it when we could not find it. What we show at that time is documented in when a symbol cannot be found.