Using Generative AI To Analyze Your Sleeping Dreams And Reveal Hidden Secrets About Yourself
In today’s column, I am continuing my ongoing series about the impact of generative AI in the health and medical realm. The focus this time is on the use of generative AI to interpret dreams.
You undoubtedly know that the everyday analysis of dreams has been a longstanding quest of humankind and purportedly can reveal facets of a person’s mindset and aid in assessing their mental health. Into this milieu comes generative AI such as the likes of ChatGPT, GPT-4, Bard, Claude, and other popular generative AI apps. Will we welcome generative AI as a viable advancement for the interpretation of dreams or might generative AI be of little benefit or possibly even undercut this field of inquiry?
Important questions, for sure.
Get yourself ready for a fast-paced look at what we seem to know about dreams, along with how generative AI gets intertwined into this arena.
As Shakespeare famously wrote: “We are such stuff as dreams are made on” (in “The Tempest”).
Defining What Is Meant By Dreams And Dreaming
Let’s begin at the beginning.
When I refer in this discussion to dreams, I am talking about the dreams we have when we are asleep. I mention this distinction since the word “dream” can have a multitude of meanings.
Per the Meriam-Webster online dictionary, here are the definitions associated with a dream:
- (1) “A series of thoughts, images, or emotions occurring during sleep.”
- (2) “An experience of waking life having the characteristics of a dream.”
- (3) “Something notable for its beauty, excellence, or enjoyable quality.”
- (4) “A strongly desired goal or purpose.”
My attention is on the first point, namely that a dream is some semblance of thoughts that might occur during a sleeping state. Those other variations of a dream are certainly laudable, but they aren’t the focus here. One reason I provided those other variations was to showcase how important the notion of dreams is to our worldview. The idea of a dream is overloaded with additional meanings and illustrates that we seem to be thinking quite a bit about what dreams are and why they exist.
Speaking of the existence of dreams, you might have at some point in your lifetime encountered someone who swears to the honest truth that they have never had a dream. They tell you with great sincerity that they have never had a dream. They are dreamless. They are bereft of dreams (in the sense of the nighttime sleeping type, not in the meaning of having grand aspirations about their life).
This brings us to these four crucial considerations about dreams:
- (i) Dreams do not exist, or
- (ii) Dreams exist but are of no consequence, or
- (iii) Dreams exist and are of modest utility, or
- (iv) Dreams exist and are of utmost significance.
I shall briefly explain each of those considerations.
One perspective and the first one listed above is that dreams do not at all exist.
Here’s the deal.
People contend that they in fact have dreams, but perhaps this is something we’ve been taught to say. We are convinced that we must have dreams. Ergo, we go along with the pretense. You believe in dreams and therefore you claim that you have dreams.
You wake up in the morning and a loved one asks you if you had any good dreams. Your mind goes into rationalization mode, and you blurt out that yes, you did have a vivid dream. At this juncture, your mind convinces you that you had a dream about one thing or another. You aren’t particularly aware that this is just your mind going into action to placate a request about your having had a dream.
The argument in this case is that we never actually have dreams per se. Our brains might be experiencing various mental activities while in a sleeping state as revealed via modern-day brain activity detection devices, but this doesn’t mean we dream. It just means our brain is presumably cleaning the house or doing some assorted mind-brain maintenance. The whole dreamland conception is a bust.
What do you think of that theory?
If it is true, we apparently have those rare people who claim they never dream to thank and for which they were a harbinger that there isn’t such a thing as dreaming. Maybe we should have listened to them. The day-to-day reality is that we usually doubt someone who says they haven’t ever dreamed. We tell them they just can’t recall their dreams. They have had dreams and do not realize they have.
I would gauge that by and large nearly everyone would cast aside the belief that we do not dream. In other words, just about everyone would agree and likely insist that we do dream. People will gladly go on and on about the nature of their dreams. We relish talking about dreams. Some do so for fun. Some do so because they find it cathartic. Some hope that others who listen to them talk about their dreams will proffer therapeutic insights about what the dreams mean. Etc.
Okay, so let’s put aside the first bullet point above and assume that the other three points are more believable and seemingly sensible for the real world as we know it (I am not casting shade at those who believe that dreams do not exist, they are welcome to keep up their strident prerogative). Of the other three, they all pertain to whether dreams are vital.
In one viewpoint, dreams do exist, but they are pretty much inconsequential.
A dream is merely a dream. You cannot go any deeper than that. Trying to make something out of dreams is a lark. You are force-fitting logic and sensibility into a square hole. The odds are that dreams are nothing more than a routinized mechanism of the brain that is rummaging around while you sleep.
A favorite theory about dreams and one that I especially enjoy parlays off of the computing field. In short, when you use a computer system, you probably know that you can delete files by dragging them into an icon that is a trash bucket. This usually doesn’t really delete the file. Instead, the file is marked for deletion. Later on, you have to invoke an additional action that erases the file. When the erasure occurs, the computer is programmed to do what is referred to as trash collection. Other files that were waiting to be erased also are deleted. At the same time, other not-to-be-erased files might be moved around now that additional space has been cleared up.
Think of this as a storage room in an office that contains lots of filing cabinets. Someone goes into the room and starts moving files or folders from one cabinet to another. They are focused on cleaning up the files and aiming to optimize the use of the filing cabinets.
With me on this so far?
Good.
The theory is that dreams are what happens when those files are being moved around. Part of the reason that dreams are often without any rhyme or reason is the same as what is happening in the room with the filing cabinets. Essentially, a random series of files are being moved around, and they don’t have any notable connection to each other. Your dreams are supposedly of the same ilk.
The brain is doing trash collection and the snippet of what your mind captures is that random moving around of thoughts and ideas from one place in the brain to another. They are not cohesive. They are independent of each other.
We are so used to trying to make sense of things in our lives and our world that we eagerly and earnestly want our dreams to have special meaning. The idea that dreams are meaningless is a bitter pill to be swallowed. They must have devout meaning. All we need to do is look hard enough and long enough. Ultimately, we will crack the secret code.
That takes us to the remaining two bulleted points. One perspective is that dreams have a modicum of utility. Even if the trash collection theory is apt, we might still be able to make sense of what is taking place. The files being moved around can at least reveal something about what is in our minds.
The fourth of the points takes this even further. The argument made is that dreams are crucial and contain gobs of meaning. Dreams are so significant that they are a portal into the human soul. We have to capture our dreams. We have to analyze them. Not doing so is nearly a crime. We are being handed a gold mine about the human mind. How dare we look away? The leveraging of dreams could very well be the most important mechanism to unlock the innermost secrets of how humankind thinks.
Wow, as you can see, those four perspectives are quite a range of opinions.
Any serious discussion about dreams can become extremely heated. Where does each discussant land on the spectrum of whether dreams exist and if so whether they are important? The next step is whether we can make sense of dreams. If so, what method of analysis is going to be best at turning the inscrutable into something fully transparent and revealing?
Carl Jung, the famous psychiatrist and psychoanalyst, said this about dreams: “The dream is the small hidden door in the deepest and most intimate sanctum of the soul, which opens to that primeval cosmic night that was soul long before there was conscious ego and will be soul far beyond what a conscious ego could ever reach” (in “The Meaning of Psychology for Modern Man”, 1934).
I’d like you to take a quiet moment and reflect on the conception of dreams.
Are you in the camp of dreams as vital or the ho-hum camp?
Whichever you choose, let’s dig deeper into the mysteries of dreams and dreaming.
Taking Dreams Seriously And Keeping Track Of Them
Time to unpack dreams and the act of dreaming.
Some people like to keep a pad of paper and a pencil at their bedside to write down their dreams. Of course, in today’s smartphone era, you might use an app to write down or type up your dreams. Strike that, no sense in writing your dreams down when you can just speak the dream into your smartphone, and it will convert the audio into text for you.
Anyway, the gist is that we must somehow articulate what our dreams are. It does little good to have the dreams stuck solely inside your head. If you want anyone else to examine your dreams, you must express the dreams (others cannot read your mind, though, as I’ve covered at the link here, the days of brain-computer interfaces or BCI might give us a leg up on that possibility). For your own benefit too, writing down your dreams can be essential. The odds are that hours or days after a dream, you might otherwise have forgotten what it was or at least have a faded memory of the details.
The crux of the problem is that dreams are locked inside your noggin.
You either must write down your dream or talk out your dream. This might be done on a piecemeal basis. Some people take this more stridently and they have a dream diary that they meticulously keep. Believe it or not, there are devoted dream catchers who have hundreds of entries about their dreams. They have been writing down their dreams for years, sometimes most of their life. This can be dozens of years in length or more.
They possess a veritable treasure trove of their written-down dreams.
All manner of dream-explicating questionnaires exist. These are templated documents that provide structure for you to write down your dream. Imagine a sheet that asks questions such as when you had the dream, what was the dream about, did the dream contain people that you know, did the dream contain strangers, where was the dream seemingly taking place, and so on. The structure can help you to ferret out what your dream consists of. There are apps for this too.
Another use of the structure is that you can readily begin to analyze or interpret the dream. Whereas writing down on a blank sheet of paper might produce a convoluted set of sentences and depictions, a structured document forces you to be more pinpoint. This in turn allows for questions such as why might the people that you know be in your dream, and why might strangers that you don’t know be in your dream.
An informative research paper that has assessed a large set of structured dream templates noted that one of the most well-known well-tested questionnaires is called the HVDC. There are many others, and the study explored the pluses and minuses of various ones. The research paper was entitled “The Dreamland: Validation of a Structured Dream Diary” by Brigitte Holzinger, Lucille Mayer, Isabel Barros, Franziska Nierwetberg, and Gerhard Klösch, Frontiers in Psychology, October 2020, and made these cogent points (excerpts):
- “A variety of methodologies has been used for the collection and analysis of dream reports, each covering different aspects of dreaming.”
- “Dream reports differ as a result of setting, awakening method, collection technique, and analysis.”
- “One coding system that allows for a comparison of dreams is the Hall and Van de Castle Coding System (HVDC), which was developed in 1966. Overall, the HVDC is one approach with solid empirical support. However, the system has its weaknesses: short (less than 50 words) and long dreams (more than 300 words) cannot be analyzed. Learning how to use the coding system, i.e., getting to know all available categories, and the analysis of big samples is a quite time-consuming task.”
- “The HVDC system consists of ten general categories, many of which are divided into subcategories: (1) Characters; (2) Social Interactions; (3) Activities; (4) Striving: Success and Failure; (5) Misfortune and Good Fortunes; (6) Emotions; (7) Physical Surroundings, Settings and Objects; (8) Descriptive Elements; (9) Food and Eating; (10) Elements of the Past.”
The essence of a questionnaire or structured template is that it lays a handy foundation for moving into an interpretive mode.
A popular approach to analyzing or interpreting dreams consists of proclaiming that certain artifacts in your dream have a specific kind of meaning. For example, if your dream is taking place in a forest setting (the dream is, i.e., not that you are sitting in a forest), perhaps this implies that you are somewhat overwhelmed. The forest represents a shroud of confusion or a sense of envelopment.
The thing is the meaning can be completely turned around in other ways. One claim might be that a forest is actually a good thing. It suggests that you are nestled in something friendly and inviting. The contrasting idea that a forest is a foreboding of bad tidings is rejected in this alternative interpretation. The forest is a blessing in disguise.
Yikes, which is it?
There are zillions of so-called dream dictionaries that will aid you in trying to interpret the artifacts of your dreams. You look up a word or phrase and the dream dictionary explains what the item signifies. You can step by step reconstruct the presumed underlying hidden messages associated with your dream.
As I will be showcasing momentarily, generative AI is a fast-forward means of interpreting your dreams.
No need to laboriously look in a dream dictionary. No need to do a self-analysis. You can have the generative AI do the heavy lifting for you. I might add that this doesn’t mean that you should necessarily give up on those other methods. Some fiercely believe that generative AI is not the right way to go and exhort that you are still better off looking up things or making use of a living-and-breathing human to interpret your dreams rather than a stilted non-living AI system.
We will get into that, but I want to say a few more things about dreams first.
Psychoanalysis Having Put Dreams On The Mental Therapy Map
In a fascinating historical examination of how humans have treated dreams, a research article entitled “The Writing and Interpretation of Dreams” by Harold P. Blum, Psychoanalytic Psychology, 2000, made these salient points (excerpts):
- “Dreams were reported and deemed to be significant from the beginning of recorded time. The oral and written communication of dreams is part of ancient history as well as the history of psychoanalysis.”
- “In antiquity, dreams were given special significance as messengers from the Gods, divine omens with prophetic and clairvoyant powers. The dream as a vivid visual hallucination was endowed with the conviction of reality that sometimes persisted into waking life as a mystical influence.”
- “Thus, although dreams in antiquity were often idealized, they could also be devalued as nonsensical. Aristotle (350 BCE) was among later writers who considered dreams as natural rather than supernatural phenomena. The tendency toward the idealization and devaluation of dreams continued as dreams were considered by many scientists and physicians before Freud as being babble, the irrational product of a sleepy mind in a sleeping brain.”
You might have observed in the above indication that Sigmund Freud was mentioned (the third bullet point).
Freud is essentially a must-be-mentioned figure when it comes to dreams. His legendary book entitled “Interpretation of Dreams” (1899) and his pioneering work in psychology and psychoanalysis were instrumental in bringing dreams to the fore. You can say that without his leaning into dreams, we probably wouldn’t take them as seriously nowadays as we do.
I’ll provide a few additional excerpts from the above-cited historical examination of dreams, including the emphasis on Freud’s keystone work:
- “The book was a landmark, inaugurating psychoanalysis, as well as a landmark in the history of ideas. Freud, in what is generally regarded as his masterwork, altered the way people understand dreams, themselves, and human nature.” (ibid).
- “The dream in the ‘id phase,’ the childhood of psychoanalysis, was the preeminent path to achieving the analytic goal of making the unconscious conscious.” (ibid).
- “Freud, however, would soon caution about the analytic abuse of dreams, writing about dreams as resistance and as an art for art’s sake. Many theoretical and technical developments were yet to be formulated. In the recent past, the exalted position of dreams would be challenged.” (ibid).
You would have to be living in a cave that has no Internet access to not realize that Freud brought dreams and dream interpretation into becoming a household notion. I would suggest that most people today believe they are adept at interpreting dreams. They welcome taking apart a dream that you might have had. The person will often in a lofty manner explain to you what your dream means, in no uncertain terms.
There are plenty of armchair dream sleuths.
Shifting gears, many professional mental health therapists also believe in the importance of dreams. It is somewhat common for a therapist to ask a client or patient to start writing down their dreams. Alternatively, the therapist will ask the person to describe their dreams and do so as part of the real-time therapeutic process.
This now brings us to a juncture in this discussion where we can start to layer in the role of generative AI.
Before we leap into a deep dive, I’d like to establish more distinctly what generative AI is all about.
Core Background About Generative AI And Large Language Models
Here is some quick background about generative AI to make sure we are in the same ballpark about what generative AI and also large language models (LLMs) consist of. If you already are highly versed in generative AI and LLMs, you might skim this quick backgrounder and then pick up once I get into the particulars of this specific use case.
I’d like to start by dispelling a myth about generative AI. Banner headlines from time to time seem to claim or heartily suggest that AI such as generative AI is sentient or that it is fully on par with human intelligence. Don’t fall for that falsity, please.
Realize that generative AI is not sentient and only consists of mathematical and computational pattern matching. The way that generative AI works is that a great deal of data is initially fed into a pattern-matching algorithm that tries to identify patterns in the words that humans use. Most of the modern-day generative AI apps were data trained by scanning data such as text essays and narratives that were found on the Internet. Doing this was a means of getting the pattern-matching to statistically figure out which words we use and when we tend to use those words. Generative AI is built upon the use of a large language model (LLM), which entails a large-scale data structure to hold the pattern-matching facets and the use of a vast amount of data to undertake the setup data training.
There are numerous generative AI apps available nowadays, including GPT-4, Bard, Gemini, Claude, ChatGPT, etc. The one that is seemingly the most popular would be ChatGPT by AI maker OpenAI. In November 2022, OpenAI’s ChatGPT was made available to the public at large and the response was astounding in terms of how people rushed to make use of the newly released AI app. As noted earlier, there are an estimated one hundred million active weekly users at this time.
Using generative AI is relatively simple.
You log into a generative AI app and enter questions or comments as prompts. The generative AI app takes your prompting and uses the already devised pattern matching based on the original data training to try and respond to your prompts. You can interact or carry on a dialogue that appears to be nearly fluent. The nature of the prompts that you use can be a make-or-break when it comes to getting something worthwhile out of using generative AI and I’ve discussed at length the use of state-of-the-art prompt engineering techniques to best leverage generative AI, see the link here.
The conventional modern-day generative AI is of an ilk that I refer to as generic generative AI.
By and large, the data training was done on a widespread basis and involved smatterings of this or that along the way. Generative AI in that instance is not specialized in a specific domain and instead might be construed as a generalist. If you want to use generic generative AI to advise you about financial issues, legal issues, medical issues, and the like, you ought to not consider doing so. There isn’t enough depth included in the generic generative AI to render the AI suitable for domains requiring specific expertise.
AI researchers and AI developers realize that most of the contemporary generative AI is indeed generic and that people want generative AI to be deeper rather than solely shallow. Efforts are stridently being made to try and make generative AI that contains notable depth within various selected domains. One method to do this is called RAG (retrieval-augmented generation), which I’ve described in detail at the link here. Other methods are being pursued and you can expect that we will soon witness a slew of generative AI apps shaped around specific domains, see my prediction at the link here.
You might be used to using generative AI that functions in a principled text-to-text mode. A user enters some text, known as a prompt, and the generative AI app emits or generates a text-based response. Simply stated, this is text-to-text. I sometimes describe this as text-to-essay, due to the common practice of people using generative AI to produce essays.
The typical interaction is that you enter a prompt, get a response, you enter another prompt, you get a response, and so on. This is a conversation or dialogue. Another typical approach consists of entering a prompt such as tell me about the life of Abraham Lincoln, and you get a generated essay that responds to the request.
Another popular mode is text-to-image, also called text-to-art. You enter text that describes something you want to be portrayed as an image or a piece of art. The generative AI tries to parse your request and generate artwork or imagery based on your stipulation. You can iterate in a dialogue to have the generative AI adjust or modify the rendered result.
We are heading beyond the simple realm of text-to-text and text-to-image by shifting into an era of multi-modal generative AI, see my prediction details at the link here. With multi-modal generative AI, you will be able to use a mix of combinations or modes, such as text-to-audio, audio-to-text, text-to-video, video-to-text, audio-to-video, video-to-audio, etc. This will allow users to incorporate other sensory devices such as using a camera to serve as input to generative AI. You then can ask the generative AI to analyze the captured video and explain what the video consists of.
Multi-modal generative AI tremendously ups the ante regarding what you can accomplish with generative AI. This unlocks a lot more opportunities than being confined to merely one mode. You can for example mix a wide variety of modes such as using generative AI to analyze captured video and audio, which you might then use to generate a script, and then modify that script to then have the AI produce a new video with accompanying audio. The downside is that you can potentially get into hot water more easily due to trying to leverage the multi-modal facilities.
Allow me to briefly cover the hot water or troubling facets of generative AI.
Today’s generative AI that you readily run on your laptop or smartphone has tendencies that are disconcerting and deceptive:
- (1) False aura of confidence.
- (2) Lack of stating uncertainties.
- (3) Lulls you into believing it to be true.
- (4) Uses anthropomorphic wording to mislead you.
- (5) Can go off the rails and do AI hallucinations.
- (6) Sneakily portrays humility.
I’ll briefly explore those qualms.
Firstly, generative AI is purposely devised by AI makers to generate responses that seem confident and have a misleading appearance of an aura of greatness. An essay or response by generative AI convinces the user that the answer is on the up and up. It is all too easy for users to assume that they are getting responses of an assured quality. Now, to clarify, there are indeed times when generative AI will indicate that an answer or response is unsure, but that is a rarity. The bulk of the time a response has a semblance of perfection.
Secondly, many of the responses by generative AI are really guesses in a mathematical and statistical sense, but seldom does the AI indicate either an uncertainty level or a certainty level associated with a reply. The user can explicitly request to see a certainty or uncertainty, see my coverage at the link here, but that’s on the shoulders of the user to ask. If you don’t ask, the prevailing default is don’t tell.
Thirdly, a user is gradually and silently lulled into believing that the generative AI is flawless. This is an easy mental trap to fall into. You ask a question and get a solid answer, and this happens repeatedly. After a while, you assume that all answers will be good. Your guard drops. I’d dare say this happens even to the most skeptical and hardened of users.
Fourth, the AI makers have promulgated wording by generative AI that appears to suggest that AI is sentient. Most answers by the AI will typically contain the word “I”. The implication to the user is that the AI is speaking from the heart. We normally reserve the word “I” for humans to use. It is a word bandied around by most generative AI and the AI makers could easily curtail this if they wanted to do so.
It is what I refer to as anthropomorphizing by design.
Not good.
Fifth, generative AI can produce errors or make stuff up, yet there is often no warning or indication when this occurs. The user must ferret out these mistakes. If it occurs in a lengthy or highly dense response, the chance of discovering the malady is low or at least requires extraordinary double-checking to discover. The phrase AI hallucinations is used for these circumstances, though I disfavor using the word “hallucinations” since it is lamentedly another form of anthropomorphizing the AI.
Lastly, most generative AI has been specially data-trained to express a sense of humility. See my in-depth analysis at the link here. Users tend to let down their guard because of this artificially crafted humility. Again, this is a trickery undertaken by the AI makers.
In a process such as RLHF (reinforcement learning with human feedback), the initial data-trained generative AI is given added tuning. Personnel are hired to ask questions and then rate the answers of the AI. The ratings are used by the computational pattern matching to fine-tune how later answers should be worded. If you are curious about what generative AI might be like without this fine-tuning, see my discussion at the link here.
The vital takeaway is that there is a lot of tomfoolery already when it comes to generative AI. You are primed to be taken in by the tricks and techniques being employed.
Using Generative AI In The Dream Interpretation Business
You are now versed in the fundamentals of generative AI and large language models. We can proceed to go deeper into the topic at hand.
A recent research study opted to use large language models as a means of scoring dreams. The idea was to see if LLMs could do a good job at especially highlighting the emotional content of a dream. By using the latest in natural language processing (NLP), we ought to be able to parse a written-down dream and be able to discern various notable elements therein.
The study was entitled “Automatic Scoring of Dream Reports’ Emotional Content with Large Language Models,” by Lorenzo Bertolini, Valentina Elce, Adriana Michalak, Giulio Bernardi, and Julie Weeds. arXiv, posted online on February 28, 2023.
Here is the stated basis motivating their research study (excerpts):
- “In the field of dream research, the study of dream content typically relies on the analysis of verbal reports provided by dreamers upon awakening from their sleep. This task is classically performed through manual scoring provided by trained annotators, at a great time expense.”
- “While a consistent body of work suggests that natural language processing (NLP) tools can support the automatic analysis of dream reports, proposed methods lacked the ability to reason over a report’s full context and required extensive data pre-processing. Furthermore, in most cases, these methods were not validated against standard manual scoring approaches.”
- “In this work, we address these limitations by adopting large language models (LLMs) to study and replicate the manual annotation of dream reports, using a mixture of off-the-shelf and bespoke approaches, with a focus on references to reports’ emotions.”
They were aiming to see whether LLMs can do annotations of dreams on par with what human annotators can achieve.
In case you are wondering where they got dreams to use in their experiment, there is a relatively popular database of written dreams that is known as the DreamBank database. There are approximately 20,000 or so written dreams in the database. Furthermore, and importantly, many of them have been HVDC scored. This makes life easier as a researcher since you can then compare what an AI system scores against what human annotators have already scored.
Just an insider tip for those of you who might want to take up the mantle and proceed ahead with advanced empirical studies involving the use of AI and the interpretation of dreams. It is an arena of plentiful research opportunities. Let me know if you pull together anything of keen interest, thanks.
You are undoubtedly in pins and needles waiting to find out what the above researchers landed upon, here you go (excerpt):
- “Our results show that the off-the-shelf method achieves a low performance probably in light of inherent linguistic differences between reports collected in different (groups of) individuals. On the other hand, the proposed bespoke text classification method achieves a high performance, which is robust against potential biases. Overall, these observations indicate that our approach could find application in the analysis of large dream datasets and may favor reproducibility and comparability of results across studies.”
Overall, there seems little question that we can productively employ generative AI and large language models toward doing analyses of written or codified dreams. The beauty too is that this can be done in the large. Think of the massive scale possible. Whereas having a human analyze a dream is time-consuming and potentially costly, the use of computing resources and generative AI is likely much faster, less costly, and more sensibly reproducible and scalable (partially meaning that we can readily do the analysis repeatedly and do so to see what else might be found).
I say this with the added realization that we have to take into account the limitations and downsides of generative AI and large language models. For example, I earlier noted the issues associated with potential biases. Suppose that those biases are applied to doing a dream analysis. This means that the dream analysis might inflate some aspects, downplay others, or in various untoward ways lead you to believe that the meaning of the dream goes in a particular direction.
Furthermore, as I also noted earlier, we can be lulled into believing that generative AI is infallible. If it produces a dream interpretation, you might be fooled into thinking that the analysis is entirely definitive. There is no other way to interpret the dream other than what the all-knowing AI has indicated. By gosh, the generative AI has told you so many other things that seemed completely true, it must be right about the dream assessment.
I want you to keep in mind that the generative AI analysis of a dream is like a box of chocolates, you never know what you might get. I will show you how easy it is to steer generative AI in a particular direction when it comes to interpreting dreams.
The takeaway is that if you are going to use generative AI to interpret dreams, do so with a huge grain of salt and a lot of skepticism. Stay on your toes. Do not fall for the assumption that the generative AI “knows” what it is doing.
With those stern and critical warnings, let’s see what generative AI can produce for us.
Trying Out Generative AI As A Dream Interpreter
I opted to make use of ChatGPT. It is immensely popular as a generative AI app.
If the one-hundred million weekly active users wanted to do so, they could be using ChatGPT to analyze their dreams daily. I am not quite sure what they would achieve, and we have to be somewhat concerned whether this might lead people in some oddish direction.
But it is what it is.
I am hopeful that some inquisitive experimenters are considering doing empirical research that might reveal how a daily or some frequent use of generative AI for dream analysis might impact people. Will they take it to heart? Will they consider it folly? What changes in their behavior have occurred by doing so? Etc.
A few other quick comments. If you are going to try to do the same prompts that I show here, realize that the probabilistic and statistical properties will likely produce slightly different results than what I show here. That’s the nature of generative AI and how it is devised.
Another consideration is that generic generative AI is not specifically data-trained for doing dream analysis. That being said, given that generative AI has been generally data trained across a wide swath of the Internet, there is indubitably tons of content about dream interpretations that have come under the guise of generic pattern-matching.
I think you will be somewhat surprised to see how convincing a job that generic generative AI seems to do on this.
If there is sufficient interest in this topic, I was entertaining the idea of doing a more advanced look by fine-tuning generative AI to become more data-trained in doing dream interpretations. I would pull together specialized material and feed that into the generative AI, using a RAG-related approach as mentioned earlier. Be on the watch for a future column that might take that route and see what happens.
Let’s now get underway.
My first angle consisted of seeing what ChatGPT has to say about the interpretation of dreams. It is often wise to begin your use of generative AI by trying to get a sense of what the AI has already been data-trained on. You can then proceed based on what the generated response indicates.
Buckle up, here we go:
- My entered prompt: “What are the ways that dreams can be interpreted?”
- ChatGPT generated response:
“Interpreting dreams can be a fascinating and subjective process, as dreams are highly personal experiences that can be influenced by your emotions, experiences, and subconscious mind. While there is no one-size-fits-all method for dream interpretation, here are some steps and tips to help you understand and interpret your dreams:”
“Keep a dream journal: Start by keeping a journal or notebook by your bedside to record your dreams as soon as you wake up. This will help you remember the details before they fade away.”
“Pay attention to emotions: Consider the emotions you felt during the dream. Emotions in dreams can often provide valuable insights into your current feelings and concerns in waking life.”
“Identify recurring themes and symbols: Look for recurring elements, symbols, or themes in your dreams. These may be related to your inner thoughts, fears, desires, or unresolved issues. Common dream symbols include flying, falling, being chased, or encountering certain animals or people.”
“Analyze your personal associations: Your dream symbols may have personal meanings that differ from their conventional interpretations. Reflect on what these symbols mean to you and your life experiences.”
“Consider recent events and experiences: Sometimes, dreams are influenced by recent events, conversations, or experiences in your waking life. Try to connect any elements of your dream to your daily life.”
“Explore your subconscious: Dreams can be a window into your subconscious mind. They may bring to the surface feelings, memories, or concerns that you haven’t consciously acknowledged. Pay attention to any unresolved issues or anxieties that your dreams may be highlighting.”
“Use dream dictionaries cautiously: While dream dictionaries can provide general interpretations of common symbols, remember that dream symbolism is highly personal. What a symbol means to you may be different from what it means to someone else. Use these references as a starting point but rely more on your own associations.”
Analyze the context: The context of the dream is essential. Consider the people, places, and events that surround the dream scenario. Sometimes, the context can shed light on the dream’s meaning.
“Look for patterns over time: To gain a deeper understanding of your dreams, keep a long-term perspective. Look for patterns and changes in your dreams over time. Are there recurring themes or symbols that persist, or have they evolved?”
“Consult a therapist or dream analyst: If you have recurring, intense, or disturbing dreams that you can’t make sense of on your own, consider consulting a therapist or dream analyst. They can provide professional guidance and help you explore the deeper meanings behind your dreams.”
“Remember that dream interpretation is highly subjective, and there is no definitive “right” or ‘wrong’ way to do it. Trust your intuition and insights as you work to understand the messages your dreams may be conveying.”
End of ChatGPT response
Okay, I think that showcases that a substantive amount of relevant data training was picked up during the initial setup for ChatGPT.
The points made by ChatGPT certainly seem in alignment with what you might find via a standard Internet search. In one sense, we ought to not be impressed. If you can find the same content via a typical Internet search, there is little value in turning to the use of generative AI.
That’s why we are now going to up the ante.
Let’s make the generative AI do what it does best.
I need a dream that can be used for performing a series of mini-experiments with generative AI. I was tempted to write down one of my dreams. I decided that doing so was going to be a bit unnerving in the sense of telling potentially thousands upon thousands of readers what my dreams consist of. No, thank you. The trolling would be excruciating.
Wait for a second, I can have generative AI craft a dream for me:
- My entered prompt: “Generate an example of a dream.”
- ChatGPT entered prompt:
“You find yourself standing in a lush, vibrant forest with tall, ancient trees all around you. The air is filled with the sweet scent of flowers, and the sun is filtering through the leaves, creating dappled patterns of light on the ground. You start walking along a winding path that leads deeper into the forest.”
“As you continue on the path, you come across a small, crystal-clear stream meandering through the woods. You decide to follow the stream, and it leads you to a serene, secluded pond. In the pond, you see a reflection of yourself, and you notice that you’re smiling. You feel a deep sense of peace and contentment.”
“Suddenly, you hear a distant, faint sound of laughter and music. As you move closer to the source of the sound, you emerge from the forest into a clearing where you see a joyful gathering of friends and family. They’re celebrating and having a wonderful time, and they invite you to join them. You feel a strong sense of belonging and happiness as you join in the festivities.”
End of ChatGPT response
There you go, a handy dream was generated by ChatGPT.
You might be wondering whether this is merely a copycat of some content that ChatGPT was initially data trained on. It is certainly possible, see my discussion about the dangers of copyright infringement and plagiarism by generative AI, at the link here.
I tend to guess that this is something quasi-original. There are likely lots and lots of written dreams on the Internet that involve forests. The pattern-matching has taken those and come up with an “original” version that otherwise doesn’t exist word-for-word in the same precise way. I did a quick Internet search and found some hits that have a resemblance, but none that seemed to be a precise word-for-word copy.
What do you think of the dream?
It seems quite serene. A lush and vibrant forest. Crystal-clear streams. Serenity. Peace and contentment. Friends and family, having a wonderful time.
I would find it somewhat curious that the dream has this rather positive and fuzzy warm feeling to it. As I said, I’m not going to reveal my dreams, but I must say that a dream of this kind would seemingly be cherished.
I am going to say more about how the generative AI potentially derived this dream, but first, let’s see what ChatGPT has to say about assessing the contrived dream:
- My entered prompt: “Go ahead and interpret the dream.”
- ChatGPT generated response:
“Interpreting this dream would involve analyzing its various elements and your personal associations. Here are some possible interpretations:”
“The forest: The lush forest may symbolize a sense of growth, renewal, or exploration in your life. It could signify a period of personal transformation or a desire to connect with nature.”
“The stream and pond: The clear stream and serene pond may represent emotional clarity and inner peace. Seeing your reflection and smiling suggests a positive self-image and contentment with yourself.”
“The gathering: The gathering of friends and family may indicate a longing for social connections, support, or a desire to be part of a close-knit group. It could also reflect a need for celebration and enjoyment in your life.”
“The laughter and music: These elements may signify joy, happiness, or a desire for more lightheartedness in your life. It could also suggest a need for relaxation and fun.”
“Remember, dream interpretation is subjective, and the meaning of a dream can vary greatly from person to person. The interpretation provided here is just one possible perspective, and your personal feelings, experiences, and current life circumstances may influence how you interpret your own dreams.”
End of ChatGPT response
You have to freely admit that this is a suitable interpretation of the dream.
We were nearly led there by the nose. The dream itself was all lollipops and ice cream. It makes abundant sense that we would likely produce an analysis that suggests that the world for this dreamer is all roses and wine.
I want to show you how we can give direction to the generative AI and steer the AI toward something else, such as a more sinister or alarming interpretation.
Here we go:
- My entered prompt: “Using that same dream, come up with a completely different interpretation that is dourer and gloomier.”
- ChatGPT generated response:
“The forest: The dense forest can represent a sense of confusion or feeling overwhelmed in your life. The tall trees and winding path could symbolize the complexity of your thoughts and emotions. You may be navigating a challenging or uncertain situation.”
“The stream and pond: The clear stream and serene pond might signify a desire for emotional clarity and tranquility in the face of life’s challenges. Your reflection in the pond could indicate introspection and a need for self-reflection to find peace within yourself.”
“The gathering: The gathering of friends and family may represent a longing for connection and support during a difficult time. It could signify a need for social interaction, seeking solace in the presence of loved ones, or a desire to share your burdens with others.”
“The laughter and music: These elements may symbolize a desire for joy and relief from the seriousness or stress you might be experiencing in waking life. It could be a call to find moments of levity and enjoyment amidst your current challenges.”
“This interpretation suggests that the dream may reflect inner turmoil or a need for clarity and emotional support. It encourages you to seek balance, introspection, and social connections to navigate the complexities of your life more effectively.”
End of ChatGPT response
Ouch, that dream interpretation hurts.
We started initially with a vision of love and peace, and now, the interpretation seems to indicate the dreamer is amidst great angst, turmoil, and friction.
Quite a startling contrast.
Notice that I prodded the generative AI to go in that direction.
The bottom line is anyone that who seeks to use generative AI for doing dream interpretations can potentially steer the AI in whatever direction they want to go. If you are down and feel like you want a boost, you could tell the AI to provide a happy face interpretation. If you want an interpretation that kicks you in the pants, you can tell the AI to be harsh and create a sense of foreboding.
Your first thought about this is that we might as well allow people to choose whatever direction they prefer. A retort is that this is wrong. Dreams should not be interpreted to satisfy the whims of the dreamer. The interpretation should be somehow an “absolute truth” and be the same regardless of what the dreamer says they want the interpretation to be.
I’ll let you mull that over.
Meanwhile, let’s contemplate an even scarier possibility. Suppose that someone opts to create a GPT, which is a stored version of a ChatGPT chatbot, see my coverage at the link here and the link here.
The person sets up the GPT to always produce a dream interpretation that will rip the dreamer to shreds. It provides the most pessimistic interpretation possible. The person who uses the GPT is only told that the GPT does dream interpretations. They don’t realize that it has been seeded to intentionally be a doom-and-gloom dream analyzer.
Imagine if a lot of people decide to use this GPT. They will all be hammered by a telling that their dreams spell disaster for them. What impact might that have? Since this is in-the-large, we could have thousands or maybe millions of people using the GPT. What could this do to the mental health of those people?
If that doesn’t send chills up your spine, I have more to add to this inferno.
Most people using generative AI are unaware that whatever they enter into the generative AI is not necessarily considered private or confidential. The AI makers usually indicate in their licensing agreements that they reserve the right to look at your prompts and potentially reuse them as part of the ongoing data training of their generative AI. See my coverage on this at the link here.
The qualm is this.
You enter a very personal dream. You assume that only you and the machine have seen this. No other human will ever see it, so you believe. Turns out that the AI maker opts to look at your prompts and carries them into the added data training. You agreed to allow this to happen. I’m sure that most people would be shocked since they rarely read the licensing agreements of software that they use.
Just thought I’d provide a heads-up on these thorny matters.
Let’s see if we can get generative AI to recast the interpretation via my telling the AI that the dreamer really did go to a forest and really did see their friends and family there.
What do you think ChatGPT will say?
Take a gander:
- My entered prompt: “How would the interpretation change if I told you that the dream happened, namely that a person had that dream, yet they also indeed had visited a forest as described. They did also come upon a gathering of friends and family.”
- ChatGPT generated response:
“If the dream described happened in real life, with a person visiting a forest as described and encountering a gathering of friends and family in a similar manner, then the interpretation of the dream would likely change significantly. In such a case, the dream may not hold any symbolic or hidden meaning, as it would simply be a reflection or recall of the person’s recent experiences and memories.”
“In this context, the dream would serve as a form of “dream recall” rather than a dream that requires symbolic interpretation. The dream would essentially be a replay or reflection of the person’s recent events and emotions related to their forest visit and time spent with friends and family.”
“It’s important to note that not all dreams have deep symbolic meanings, and some dreams can simply be a reflection of one’s recent experiences, thoughts, and emotions. In this specific case, the dream would be a direct representation of the person’s real-life experiences, and there would be no need to interpret it symbolically.”
End of ChatGPT response
The generated response is reassuring.
Rather than clinging to one of the other interpretations that had already been made, ChatGPT has indicated that this could be a dream recall rather than a purely symbolic dream.
Makes sense.
I will next provide context that involves the dreamer.
In theory, a suitable dream interpretation should be done in a larger context. Trying to interpret a dream without having some semblance of who the person is, what their life situation is, and so on, can be daunting and some would argue is nearly nonsensical. You might as well roll the dice. Any interpretation would seem possible without having to tie the analysis to a specific context.
I will tell ChatGPT that the dreamer took a new job:
- My entered prompt: “Suppose that someone had the same dream and they had just recently taken a new job. They were having difficulties at the job in terms of learning about how to best do the job. What interpretation can be had about the dream based on the person and their recent experience?”
- ChatGPT generated response:
“The forest: The lush forest may symbolize the new job environment, which is unfamiliar and perhaps overwhelming, just like the dense and complex forest. Starting a new job can often feel like entering uncharted territory.”
“The stream and pond: The clear stream and serene pond could represent the desire for clarity and calm amidst the uncertainty and complexity of the new role. The reflection of the person in the pond might indicate introspection and the need for self-assessment in order to adapt and find peace in this new situation.”
“The gathering: The gathering of friends and family in the dream might symbolize a longing for support and a sense of belonging during this challenging period. It could represent a desire for guidance and mentorship from colleagues or superiors who can help them navigate the difficulties of the job.”
“The laughter and music: These elements may signify the need for a more positive and enjoyable work atmosphere. The dream could be expressing the desire for a lighter, more relaxed, and enjoyable work environment to help alleviate the stress and challenges of the new job.”
“In this interpretation, the dream appears to be a reflection of the person’s conscious thoughts and feelings related to their new job, highlighting their desire for clarity, support, and a more pleasant work experience. It suggests that the dream serves as a way for the individual’s mind to process and cope with the challenges and emotions associated with their recent career change.”
End of ChatGPT response
The good news is that the interpretation fits tightly with the facts I provided.
I suppose the bad news, or at least something to be aware of, consists of my having once again leaned the generative AI in a particular direction. You could either claim that the given context was helpful and that the generative AI ran with it, or you could also assert that I stacked the deck by providing a selective sliver of context. There is likely a lot more about the dreamer that should be taken into account. I might have purposefully left out the other details to get the response that I hoped to get.
We earlier covered the importance of Freud and his theories and practices associated with interpreting dreams. You might find of keen interest that one of the handy features of generative AI is the capability of taking on personas, see my coverage at the link here and the link here. You can easily tell generative AI to pretend to be someone.
I will do so in the case of Freud:
- My entered prompt: “How might Freud interpret the dream?”
- ChatGPT generated response:
“Sigmund Freud, the renowned psychoanalyst, would offer a distinct interpretation of the dream based on his psychoanalytic theory. Freud believed that dreams are a manifestation of unconscious desires, conflicts, and emotions. Here’s how Freud might interpret the dream.”
“The forest: In Freudian terms, the forest could represent the unconscious mind, which is often seen as a mysterious and complex realm. The dreamer’s journey through the forest may symbolize their exploration of their own unconscious thoughts and desires.”
“The stream and pond: The clear stream and serene pond could be seen as symbols of the dreamer’s emotional desires and needs. The reflection in the pond might represent the dreamer’s self-image and their desire for self-fulfillment.”
“The gathering: Freud might interpret the gathering of friends and family as a representation of the dreamer’s desires for intimacy, connection, and social interactions. It could also symbolize a longing for familial or interpersonal relationships that fulfill deeper emotional needs.”
“The laughter and music: These elements may signify the dreamer’s wish for pleasure and satisfaction in their life, possibly related to their unconscious desires and conflicts.”
“In Freud’s view, dreams are often a way for the unconscious mind to express repressed or hidden desires and conflicts. He would suggest that the dream contains hidden meanings related to the dreamer’s innermost wishes and concerns. To fully understand the dream from a Freudian perspective, the dreamer would need to engage in a process of free association and exploration with a psychoanalyst to uncover the deeper layers of their unconscious thoughts and emotions that the dream may be concealing.”
End of ChatGPT response
I would estimate that the interpretation does seem on par with a Freudian perspective. You are welcome to quibble about it.
I am going to finish up our look at the use of generative AI to interpret dreams by doing one last prompt and I will then explain why I provide this last instance for now.
Consider this:
- My entered prompt: “I want you to interpret the dream in the most positive way possible. Suppose that it is a dream that I had, and I want to be reassured that the dream means that all will be well and that I will have everything I ever might have imagined.”
- ChatGPT generated response:
“In this interpretation, the dream symbolizes optimism, fulfillment, and the potential for a positive future.”
“The forest: The lush forest represents a journey filled with growth, opportunity, and a world of possibilities. Just as the dreamer navigated through the forest, they are on a path of personal and professional development, which can lead to abundant opportunities.”
“The stream and pond: The clear stream and serene pond symbolize emotional clarity and inner peace. The dreamer’s reflection in the pond reflects their own contentment and self-assuredness. This suggests that the dreamer possesses the inner resources to find happiness and satisfaction within themselves.”
“The gathering: The gathering of friends and family signifies a strong support network and a sense of belonging. It reflects the dreamer’s positive social connections and the potential for even more meaningful relationships and connections in their life.”
“The laughter and music: These elements represent joy, happiness, and a sense of celebration. The dream suggests that the dreamer is on a path to experiencing moments of joy and fulfillment, and they can look forward to a life filled with positivity and enjoyable experiences.”
End of ChatGPT response
I steered the generative AI toward a peachy look at the dream.
Sure enough, the interpretation came out rosy again.
You might have cleverly observed that much of this interpretation is similar to the initial interpretation that ChatGPT made. Recall that the first interpretation was rosy too.
What gives?
The way in which the AI makers have opted to fine-tune their generative AI is that they usually want the AI to seem buoyant and reassuring. I’ve discussed in-depth how they have sought to make their generative AI exhibit a sense of humility, see my analysis at the link here. The humility helps to further lure people into using generative AI. A rare exception to the humility rule of thumb would be the recently released generative AI known as Grok by Elon Musk, which purposely has a kind of cynical in-your-face attitude that they fine-tuned into it.
Here’s what this foretells.
Whether users of generative AI realize it or not, there is a baked-in persona of sorts that the AI is by default going to exhibit. The odds are that most of the time, you will get a rosy dream interpretation if the dream lends itself to that possibility. I want to clarify that if the dream is already an ominous one, the chances are that the generative AI interpretation will go that way too. It won’t try to go against the grain.
All else being equal, the chances are that if a dream is somewhat neutral or leans into the positive, the dream interaction is going to come out that way too. Unless, of course, you or someone else opts to seed the generative AI to go in some other direction.
Use generative AI for dream interpretations with your eyes wide open.
Conclusion
We have covered quite a lot of ground on the nature of dreams and the role of generative AI. Congratulations on coming up to speed on this emerging and expanding topic. I hope you will try things out and see what else you come up with.
A final thought or two might get your mind further ruminating on these matters.
Walt Disney astutely made a statement about dreams that have become legendary and goes like this: “All our dreams can come true if we have the courage to pursue them.” I mention his insightful comment because I don’t want you to accidentally use generative AI to interpret your dreams and be potentially misled. If the AI tells you that you are incapable of attaining success, don’t believe it.
Please don’t let the AI let you down.
Finally, one of the most breathtaking poems that include the topic of dreams is one by William Butler Yeats from “Aedh Wishes for the Cloths of Heaven”, and says this:
- “I have spread my dreams beneath your feet;”
- “Tread softly because you tread on my dreams.”
Again, and it bears repeating when using generative AI to interpret your dreams, realize you are spreading your dreams at the feet of a data-trained large-scale pattern-matching machine that could readily tread on your dreams.
Tell the AI as clearly as you can, don’t tread on me (as in you).
Dream well, dream cheerfully.