We have dreamed up what would it be to put together all the important formulations for who we are as humans. Yet, you can’t ask AI to do the impossible, right? Here we guide a large language model (LLM) to use math in the best possible way to show us something we haven’t seen before.
The idea is not to make things up, the idea is to let AI put together things we already understand in new ways we might not have thought about before.
What you are about to see, is a lengthy prompt, no doubt. It provides a table of propositions and hypotheses, with self-criticism, using a selected set of 20 equations.
This is our favorite prompt so far.
Some astounding suggestions by this prompt (tilt sideways if reading on mobile):
| Title | Insight (hypothetically) |
|---|---|
| Persuasion = f(Central_Route_Processing) | If you want to persuade someone, try to make your arguments consistent with their existing beliefs. |
| Learning Acceleration Total Cognitive Load = Intrinsic Load x Self-Efficacy | Greater belief in one’s abilities can potentially increase the efficiency of learning complex material. |
| Optimal arousal leads to peak performance | If you want to perform your best, you should find the level of arousal that is right for you. |
| Impact of Cognitive Load on Detection | Minimize unnecessary cognitive load when trying to make important decisions. |
| Strength of New Associations | To optimize learning, one should consider both the learner’s readiness (potential for learning) and the environment. |
| Motivation = f(Autonomy, Competence, Relatedness) | If you want to motivate someone, try to make the task interesting to them. |
| Habits | We can change our habits by changing our environment or our past experiences (i.e., our interpretation of these). |
| Thinking, Fast and Slow T = f(S1, S2) | We have two systems of thinking: System 1 is fast and intuitive, while System 2 is slow and deliberate. We can improve our thinking by using both systems of thinking effectively. |
| The Power of Positive Thinking E = f(T) | Emotions (E) are a function of Thoughts (T). We can improve our emotions by changing our thoughts. |
Prompt (Copy-paste to ChatGPT – 1141 word length):
Please consider the following well-known formulas: 1. Classical Conditioning (Pavlov's Equation): ΔV = αβ(λ - ΣV). This equation describes the process of classical conditioning, where an organism learns to associate a neutral stimulus (conditioned stimulus) with a significant stimulus (unconditioned stimulus). ΔV represents the change in associative strength, α and β are learning rate parameters, λ is the maximum possible associative strength, and ΣV represents the sum of the strengths of all stimuli present. 2. Operant Conditioning (Rescorla-Wagner Model): ΔV = α(λ - V). The Rescorla-Wagner Model is used to describe the process of operant conditioning, which involves learning through the consequences of voluntary behaviors. Here, ΔV represents the change in associative strength, α is the learning rate parameter, λ is the maximum possible associative strength, and V is the current associative strength. 3. Yerkes-Dodson Law: Performance = f(Arousal). This law describes the relationship between arousal and performance, stating that there is an optimal level of arousal for peak performance. If arousal is too low or too high, performance will be compromised. 4. Signal Detection Theory: d’ = Z(H) - Z(F). Signal Detection Theory is used to model the process of decision making under uncertainty. d' represents the sensitivity index, Z(H) is the z-score for the hit rate (correct detection of a signal), and Z(F) is the z-score for the false alarm rate (incorrect detection of a signal). 5. Cognitive Load Theory: Total Cognitive Load = Intrinsic Load + Extraneous Load + Germane Load. Cognitive Load Theory describes the impact of cognitive load on learning and problem-solving. Intrinsic load refers to the inherent difficulty of the material, extraneous load is the load imposed by the way the material is presented, and germane load is the load associated with the construction of mental schemas. 6. Forgetting Curve (Ebbinghaus’ Formula): R = e^(-t/S). This equation describes the decrease in ability of the brain to retain memory over time. R is memory retention, S is the relative strength of memory, t is time, and e is Euler's number. 7. Sensory Processing Sensitivity (Aron & Aron): SPS = f(DOE). This model suggests that Sensory Processing Sensitivity (SPS) is a function (f) of the degree of environmental stimulation (DOE), where more sensitive individuals are more affected by sensory stimuli. 8. Dunning-Kruger Effect: Confidence = f(1/Knowledge). This effect describes a cognitive bias in which low-ability individuals suffer from illusory superiority, mistaking their ability as greater than it is. The less knowledge (or competence) a person has, the more confident they tend to be. 9. Prospect Theory (Kahneman & Tversky): Utility = f(Wealth, Loss_Aversion). This theory suggests that people make decisions based on the potential value of losses and gains rather than the final outcome. Wealth is the current status of the individual, and Loss_Aversion is the tendency to prefer avoiding losses over acquiring equivalent gains. 10. Flow State (Csikszentmihalyi): Flow = f(Skill, Challenge). Flow state occurs when a person's skills are fully involved in overcoming a challenge that is just about manageable. If the challenge exceeds the skill, anxiety may occur; if the skill exceeds the challenge, boredom may result. 11. Dual-Coding Theory (Paivio): Learning = f(Verbal, Visual). This theory suggests that both verbal and visual information are processed separately and in parallel. Information is easier to learn if it is presented in both formats, because it is encoded twice. 12. Drive Theory (Hull): Behavior = f(Drive x Habit). This theory posits that a behavior is the product of the drive (an aroused state of psychological tension that motivates an organism to satisfy a need) and habit (learned patterns of behavior). 13. Self-Efficacy Theory (Bandura): Performance = f(Self-Efficacy). This theory posits that an individual's belief in their ability to execute behaviors necessary to produce specific performance attainments greatly influences their performance. 14. Goal Setting Theory (Locke & Latham): Performance = f(Goal Difficulty, Goal Specificity). This theory suggests that specific and challenging goals lead to higher performance. 15. Cognitive Dissonance Theory (Festinger): Dissonance = f(Conflict, Importance). This theory suggests that we feel a tension (dissonance) when we hold two conflicting beliefs, or when our beliefs don't match our actions. The amount of dissonance depends on the importance of the subject and the degree of conflict. 16. Maslow's Hierarchy of Needs: Fulfillment = f(Physiological, Safety, Love/Belonging, Esteem, Self-Actualization). Maslow's theory suggests that we must satisfy lower level deficit needs before progressing on to meet higher level growth needs. 17. Elaboration Likelihood Model (Petty & Cacioppo): Persuasion = f(Central_Route_Processing, Peripheral_Route_Processing). This model proposes two routes to persuasion: the central route (involving thoughtful consideration of the arguments) and the peripheral route (involving other cues outside of the argument itself). 18. Mere Exposure Effect (Zajonc): Preference = f(Frequency_of_Exposure). This effect describes a psychological phenomenon by which people tend to develop a preference for things merely because they are familiar with them. 19. Attribution Theory (Weiner): Behavior = f(Ability, Effort, Task Difficulty, Luck). This theory suggests that we attribute causes to behavior based on internal factors like ability and effort, and external factors like task difficulty and luck. 20. Deci & Ryan's Self-Determination Theory: Motivation = f(Autonomy, Competence, Relatedness). This theory suggests that these three innate needs motivate self-determined behavior. Autonomy is the desire to be self-directed, competence is the need to be effective in dealing with the environment, and relatedness is the desire to have a connection with others.
Considering the meaning of all the variables and concepts involved in these 20 equations, please find ways to combine them meaningfully and creatively (e.g., using the transitive property and other algebraic operations) yet with strict mathematical rigor through a step-by-step execution to obtain a short list of the simplest and most interesting (based on the scope of potential implications or the depth of the links and observations) deductions and interpretations. Please provide hypothetical interpretations of each of these, all summarized in table-like format with a column with a tag line about the implications (let's call this column 'Title'), a column with 'Guidance' for applying what is implied if it were true (adding markings in bold to highlight key concepts), a column for the simplest possible simplification of the equation derived, a column for the theories involved and the simplest definitions of each variable, a column for the hypothetical interpretation and a final column to provide some confirmation of the validity of the operations used to derive the equations with a critical (could be positive considering discoveries are sometimes made this way) statement with solid reasoning about the likelihood of any of the suggestions being factual. Please provide this table going thoughtfully deeper on the implications taglines and the perspectives placed under modified quote column. To highlight key elements, underline text covering the most potentially insightful findings and observations based on their potential applicability in human endeavors and facilitating ways or views that could simplify and improve life.
********************End prompt*********************
This “mega-prompt” runs best on ChatGPT GPT-4, but GPT3.5 may handle the request. Google’s Bard also gives very interesting results with it. You can also get some result with Bing Chat (in our hands we had to choose the Conversation Style to “More Creative”, otherwise it wouldn’t work).
The most important part of this prompt is the part after the equations are provided. You can change the equations to obtain completely different suggestions and all sorts of intriguing hypotheses. For example, you could make it focus on mathematical concepts in sociology (see example here) and get deep insights that could optimize your networking capabilities and perhaps show different perspectives on how to grow your Twitter account (?! 🤫) Now, imagine the possibilities…
The tremendous power of something like ChatGPT is revealed further when after it shows you the first result from this prompt (the initial training), then, within the same chat you can ask it “Now do some more using a set of equations from physics”, for example. This is called Chain of thought prompting (CoT), where you basically provide a series of intermediate natural language reasoning steps. Notice how we achieve this in the following shared ChatGPT conversation: https://shareg.pt/V61qGQO.
| The Ingredients of Love | Love = (Listening / Noise) x Patience | Love can be seen as the product of meaningful listening (minus the noise or unwanted disturbances) and patience. The more patient we are and the better we listen, the more love we can feel and express. |
Try out our new quick & easy prompt optimizer WhatAPrompt.ai !
Power to the people!
Join Premium because you’ve just realized we know what we’re doing here and you want more than just ‘a taste’. This is not another AI newsletter; we will provide the occasional (bi-weekly 🤞🏽) mind blowing update exclusive to our Premium Members in our quest to open up extreme possibilities. Enjoy unlimited AI-tool pass, premium content on prompt engineering, thousands of optimized prompts and access to our exclusive hand-curated and highly sophisticated ‘prompt vault’: Prompt Museum. We explore to open doors for you and for all of us.