No AI
No dream interpreter. On purpose.
Nothing on this site is generated. Every meaning was written and sourced, every pattern is counted from your own journal, and when we do not know something the page says so instead of writing you a paragraph. Here is how to check all three.
Three claims, and how to test each one
It says when it has nothing
The search box looks things up. When it finds no entry, it tells you so and stops. No generated site can do this, because producing an answer is the only thing it knows how to do, and a fluent answer about a symbol nobody has studied is indistinguishable from a fluent answer about one that has been.
Check it: Search for something obscure. Watch it decline rather than invent.
Every claim says what kind of claim it is
Sections carry a label: research, theory, tradition or folklore. A finding from a sleep laboratory and a piece of Victorian dream lore are both worth having on the page and they are not the same kind of thing. Flattening them into one voice is precisely what generated text does, because it has no way to tell them apart.
Check it: Open any symbol and read the labels above each section.
The patterns are your own data, counted
Recurring symbols, mood correlations and the dream map are arithmetic over entries you wrote yourself. That is why they can be shown as counts and dates rather than as a paragraph describing you. Nothing is inferred that is not already in what you recorded, and nothing about your dreams is sent anywhere to be processed.
Check it: Log a few dreams and look at the patterns page. Every number traces back to an entry.
Why this matters more for dreams than for anything else
Dream interpretation is the ideal habitat for generated text. The subject is private, so nobody can contradict you. It is emotive, so fluent writing feels like insight. And it is genuinely under-researched, so a confident paragraph is impossible to distinguish from a well-supported one unless somebody shows their working. Every incentive points at generating, and most of the category has.
The cost lands on the reader. Someone searching at three in the morning about a recurring nightmare deserves to know whether what they are reading comes from a sleep laboratory, from Jung, from folklore, or from nowhere at all. Those four are wildly different and a generator renders them in one identical, soothing voice. Telling them apart is the entire job, and it cannot be done at speed by something that does not know which is which.
What we are not claiming
That a dream predicts anything, that a symbol has one true meaning, or that writing something by hand makes it correct. A funeral in a dream does not foretell a death. Dreaming about an ex says nothing about what they feel. The claim on this page is narrower and it is one we can stand behind: what is here was written by somebody, it says where it came from, and when there is no answer you get told that rather than a sentence shaped like one.
See for yourself
The fastest test takes about ten seconds: search the dictionary for something we are unlikely to hold, and see what it does.
Every source the site draws on is listed in our sources, along with what each evidence label means. The mechanics are in how it works, your own record lives in the journal, and why the site exists is in about.
Questions about how this is built
The five that come up most.
None. No language model is called at any point. There is no AI package among this site’s dependencies and no request leaves our server to an AI service, because there is nothing to request. Every meaning, scenario and tradition entry was written and reviewed by a person against sources that are named on the page. Everything else is computed from your own journal entries.
Because an answer that is always available is an answer that never depended on knowing anything. A model will give you a fluent, affecting reading of any dream, including a dream nobody had, and it will give you a different one tomorrow. It has no way of telling you that the research on your particular question is thin, contested or absent, because it has no way of knowing which of those is true. Roughly half of what people want to know about dreams has no good answer yet, and saying so is the most useful thing this site does.
It tells you it has nothing, and it does not improvise. That is the single clearest difference between this and a generated dictionary, and it is the fastest way to test any of this: search for something obscure enough that we are unlikely to hold an entry. A generator would write you one on the spot.
No. It is counting and comparison over entries you wrote: which symbols recur, how often, alongside which moods, against your logged sleep. The result is arithmetic on your own data, which is why it can be shown to you as numbers rather than as a paragraph about yourself. Nothing is inferred about you that is not in what you recorded.
Every section carries a label saying what kind of claim it is making: research, theory, tradition or folklore. Those are different things and the site refuses to flatten them into one confident voice. Where a claim rests on a study, the study is named and linked. Where something is folklore, it says folklore rather than dressing it as psychology.