Shamans

Foundation: Wisdom, intuition, subconscious Content: High utility, low relevance

Arguably the earliest form of future forecasting comes from tribal shamans. Archaeological evidence shows that this practice was prevalent among many ancient cultures, from Native American medicine men to Rishis in prehistoric India. These societies depended heavily on a priestly class or individuals with secretive rituals aimed at accessing ancestral wisdom.

Often, these rituals incorporated mental techniques such as meditation, ascetic practices like fasting, and even the use of hallucinogens from plants or fungi. Recent studies on meditation and psychedelics indicate that shamans may have tapped into genuine latent information through altered states of consciousness.

However, the subjective nature of these predictions often led to high variability in interpretation. Whether interpreting animal entrails or recounting psychedelic experiences, these forecasts were shrouded in ambiguity, leaving critical decisions—like the timing of a harvest or the decision to engage in conflict—based on dubious information.

Oracles

Foundation: Wisdom, intuition, history Content: High utility, medium relevance

The Oracle can be viewed as a more structured and institutionalized evolution of the Shaman. Originating from Ancient Greece, we might also consider Egyptian priests in this category due to their similarities. Though documentation of their specific practices is scarce, accounts in Greek literature indicate that interpretation played a significant role, evolving from miraculous predictions to sage counsel for rulers. In this sense, Oracles can be seen as precursors to the specialized advisory teams that surround contemporary world leaders, albeit with a touch of mystique.

Astrology

Foundation: Positions of celestial bodies Content: Medium utility, no relevance

I won’t linger on this topic, as astrology’s inefficacy has been thoroughly debunked by science. The existence of twins with divergent life paths challenges the validity of astrological predictions, which often resemble gossip-column speculation more than credible forecasts.

Fortune Tellers

Foundation: Psychology, common sense Content: Medium utility, low relevance

The distinction between fortune telling and astrology lies in the former’s basis in psychological observation rather than celestial positions. A fortune teller gauges your responses, choice of words, emotional state, body language, and overall demeanor.

In some respects, a skilled fortune teller may offer more insight than a poor therapist. They lack new information, but can uncover truths you may be ignoring due to personal biases and emotions. Conversely, a fortune teller might exploit these biases to reinforce misguided decisions, suggesting that maintaining regular therapy sessions could still be beneficial.

Modern Methods of Future Prediction

Transitioning away from ineffective methods, we enter the realm of contemporary science and mathematics. What can we achieve with today’s best technologies?

The Human Brain

Foundation: Sensory data, memory, hormones Content: Low utility, high relevance

Surprising, isn’t it? Our very existence hinges on our ability to predict future outcomes. There are various mechanisms within our brains that facilitate this capability.

For example, consider catching a ball. You succeed because your eyes and hands work together, guided by your brain, to anticipate where the ball will be at a given moment. The complexity of this task becomes apparent when we attempt to program robots to mimic such actions.

A key mechanism in our brain is anticipatory timing. Without it, walking would be impossible, as you must subconsciously predict a myriad of factors, such as body mechanics, friction, and gravity, to avoid falling. This explains why infants often stumble as they refine their motor skills.

The second type of prediction can feel counterproductive, as it involves imagining various scenarios, often leading to overthinking. For instance, you might lie awake pondering the consequences of missing the bus or regretting a past decision. Our brains have the ability to construct intricate narratives involving real or imagined characters.

Other predictive processes involve how our brains interpret the world based on our perceptions. Overall, prediction appears to be an intrinsic human function, and we may have only begun to grasp its complexities, especially regarding consciousness.

CIA “Super Forecasters”

Foundation: Data, intuition, common sense Content: High utility, medium relevance

Building on our capacity to rationalize and envision different scenarios, it stands to reason that some individuals possess innate skills that enhance their forecasting abilities. In 2011, the CIA sought to explore this question by hosting a competition that pitted everyday people against their seasoned analysts.

To provide context, intelligence analysts typically construct models to predict outcomes, such as electoral results in Venezuela or the likelihood of conflict in the Middle East. These models rely heavily on existing data, either public or acquired through CIA operations, and aim to derive statistical insights based on rational hypotheses to assess potential outcomes. However, human judgement plays a considerable role in selecting reference data, framing hypotheses, and determining relevant metrics.

Surprisingly, the amateur “Super Forecasters” outperformed the analysts by a wide margin. There’s a wealth of material and podcasts available on this subject, but the crux is that these individuals utilized simpler models than the analysts, who had access to advanced technology and numerous mathematicians.

Instead, the hobbyists demonstrated superior intuition in selecting appropriate reference data and were dedicated to refining their forecasting skills, often resulting in better outcomes. Though the accuracy was marginal, with improvements from 51% to 55%, this still surpassed mere chance.

This trend has continued, as hobbyists have also outperformed government agencies in forecasting the spread of COVID-19.

Insurance Actuaries

Foundation: Data, logic, common sense Content: Medium utility, high relevance

Interestingly, the most mundane industry may be the most adept at predicting the future. However, the scope of these predictions is notably narrow, as insurers focus on risk factors that could lead to significant financial losses.

Insurance predictions originated with maritime voyages, where it was common for ships to fail to return. Funds were pooled as a bet against successful voyages, with payouts to the families of sailors and ship owners. Today, actuaries utilize complex models to assess probabilities, including those related to everyday events such as cracked smartphone screens or flight delays.

At its core, insurance relies on probability as a form of prediction—essentially betting on the likelihood of future events.

Hedge Fund Quants

Foundation: Data, logic, common sense Content: Medium utility, medium relevance

Hedge funds rely on predictions as their primary business model, aiming to maximize returns on investments by any lawful means. Their methods vary, but they continually seek information advantages.

For instance, hedge funds previously acquired satellite imagery to assess crop yields or count vehicles in shopping mall parking lots. This strategy aimed to gain insights into commodity performance or business operations before official data releases. Knowledge of upcoming trends allows for informed investment decisions when data becomes public.

While profitable, this approach has limited applicability beyond financial gain.

Artificial Intelligence

Foundation: Data alone Content: Medium utility, high relevance

The Artificial Neural Networks that drive modern A.I. technologies were inspired by the human brain. While our understanding of the brain is still evolving, we have made significant strides in employing basic principles to train computers for useful tasks. This evolution began with handwriting recognition in the 1990s and has advanced to A.I. systems that excel in games like Chess and Go, as well as autonomous driving in complex environments.

As with the human brain, certain A.I. applications involve predicting future events. For example, an A.I. must interpret various vehicle systems in real time, forecasting potential scenarios, such as emergency braking on icy roads. Even more complex is the need to anticipate human behavior on the road, given the unpredictable nature of drivers and pedestrians.

Tesla's Autopilot exemplifies a leading A.I. system, utilizing eight cameras for 360-degree visibility and sonar capabilities. Tesla has developed its own computer hardware, but the most critical factor is the vast amount of data collected over years from numerous vehicles. Without data, A.I. cannot learn or predict effectively.

Board games also serve as a strong example, as successful gameplay hinges on anticipating opponents' moves. The latest generation of A.I., AlphaZero, boasts a 99% win rate in Chess and Go against human players, demonstrating a profound understanding of strategy.

It's essential to note that A.I. models rely solely on data quality. Inaccurate or biased data leads to flawed predictions.

Academic Approaches to Future Prediction

While modern methods yield various real-world applications, academic disciplines also study human populations to enhance understanding. In a sense, the pinnacle of knowledge involves prediction, as it allows for functional data models.

Social Studies

Foundation: Data, logic, common sense Content: High utility, high relevance

As someone who hasn’t delved into the humanities, I’ll keep this brief. We have explored human behavior through psychology and biology, as well as through studies of group dynamics. Although we perceive ourselves as individualistic and free-willed, our routines dictate much of our lives, such as eating, sleeping, and working.

When analyzing large populations, we can begin to model and predict how individuals allocate their time. The accuracy of these models increases when examining specific groups, such as shift workers. Ultimately, one might determine with surprising precision what someone is doing at any given hour—between chores and leisure activities. The question remains: is this knowledge beneficial or merely invasive?

Population Studies

Foundation: Data, logic, common sense Content: High utility, high relevance

On a broader scale, we can assess population behaviors over time. While predicting an individual’s lifespan may be challenging, we can estimate the decade in which someone will likely pass away using the same models. Identifying the right parameters is crucial for effective forecasting.

Sociological research examines how populations evolve and the causal factors involved. For instance, one might investigate whether children from divorced families are more likely to experience divorce themselves or the likelihood of children from low-income households attending college.

This inquiry often employs techniques similar to those used in A.I., such as Hidden Markov Models, to analyze sequential datasets and establish relationships between events. These relationships are represented in complex graphs that indicate probabilities for future actions or life events.

In the context of A.I., these graphs map out the rules of the game of life, detailing various milestones such as marriage, divorce, parenthood, and retirement. Importantly, these life events are interconnected and can be analyzed to predict future scenarios.

Global Studies

Foundation: Data, logic, common sense Content: High utility, high relevance

Examining global phenomena necessitates broader perspectives. Climate change is a prime example. Traditional data gathering and analysis techniques are often employed, but the complexities of climate science can lead to misinterpretations.

The spread of COVID-19 serves as a more immediate case study, as demonstrated by an animation from the World Economic Forum showcasing the disease's rapid global proliferation.

Global predictions are instrumental for policymakers, although they provide limited insights for individual circumstances.

Theoretical Approaches to Future Prediction

Stepping beyond current methodologies, we can explore speculative methods, including potential advancements in A.I. that could redefine our predictive capabilities. It’s fascinating to consider how many modern technologies were anticipated decades ago.

Mathematics (“Foundation”)

Foundation: Data, logic, common sense Content: High utility, high relevance

If you’re unfamiliar with "Foundation," prepare to learn more as a high-profile adaptation streams on Apple TV. This seminal sci-fi series by Isaac Asimov introduces the concept of “psychohistory,” a mathematical framework for predicting human behavior on a large scale.

> “Since emotions are few and reasons many, the behavior of a crowd can be more easily predicted than the behavior of one person can. And that, in turn, means that if laws are to be developed that enable the current of history to be predicted, then one must deal with large populations, the larger the better.” > — Isaac Asimov, Robots and Empire

An intriguing aspect of this theory is that predictions remain valid only if they are kept confidential. If individuals are aware of a prediction, they may alter their behavior, thus skewing historical outcomes.

Although psychohistory is not a recognized field of study, recent global events underscore the necessity for improved models to understand complex dynamics. These models could inform more effective strategies for future pandemics. Interestingly, Elon Musk has cited Foundation as a significant influence, suggesting he is pursuing similar concepts in his work.

Quantum Mechanics (“Devs”)

Foundation: Data, logic, common sense Content: High utility, high relevance

Few scientific concepts generate as much fascination and confusion as quantum mechanics. Unlike relativity, which can be grasped through analogies, quantum phenomena remain elusive, making them difficult to comprehend.

One fundamental question remains whether the universe is inherently random. This debate has birthed several perspectives: determinists who believe randomness is an illusion, proponents of true randomness, and theorists suggesting that every random event results in a parallel universe.

The series Devs explores these themes, teasing the possibility of omniscience. However, it’s widely accepted among scientists that complete information about any system is unattainable, as measurement itself alters the system, a concept illustrated by Heisenberg’s uncertainty principle.

Superintelligence (“Westworld”)

Foundation: Data, logic, common sense Content: High utility, high relevance

This final section synthesizes our discussion. Current A.I. systems exhibit remarkable accuracy in predicting specific scenarios. However, we have yet to achieve General Artificial Intelligence, akin to human cognitive abilities, which allow for the application of learned skills across various domains.

What if we could unlock such capabilities? This would lead to advancements beyond our current comprehension. The potential for a system capable of human-level reasoning implies superhuman cognitive abilities, facilitated by near-infinite computational resources. The portrayal of such superintelligence in Westworld effectively captures this concept.

Without revealing too much, the A.I. in the show possesses extensive knowledge and intelligence, enabling it to predict outcomes on both individual and global scales, thereby influencing events.

Advancements in A.I. research promise to enhance the depth and breadth of predictions based on existing data. Researchers suggest that true superintelligence may be just decades away, potentially leading to a future where our decisions are guided by an entity with profound predictive capabilities.

Can We Truly Predict the Future?

So, what insights can we glean regarding our ability to forecast the future? The central takeaway is that successful prediction methods share a commonality—while the universe may not adhere to quantum principles or pure mathematics, probabilistic models based on extensive data have proven effective in numerous contexts.

Simple Problems: MAYBE

Recent years have witnessed a surge in the availability and open-source nature of powerful A.I. models applicable to diverse data types and challenges. The accuracy of predictions hinges on the quality of the underlying data, which cannot be overstated. Some technical skills may be required, but the increasing complexity of models necessitates greater datasets and hardware, presenting financial barriers.

Individual Life Events: YES

Complexity increases when addressing individual life events. This category of issues can be framed as gameplay scenarios, solvable through A.I. methods such as Hidden Markov Models and Reinforcement Learning, which have been developed to outsmart human opponents in games like Chess and Go.

Population Trends: YES

The predictive methods applicable to individuals extend effectively to populations and are already utilized in academia, government, and policy-making. Insurance models similarly apply at this level, allowing for predictions of health and life outcomes based on group characteristics.

Global Statistics: YES

While weather forecasting remains elusive due to the chaotic nature of climate systems, our understanding of numerous phenomena has improved. The COVID-19 pandemic highlighted both the challenges and advancements in technology necessary for effective prediction. Climate change models are likewise becoming increasingly accurate, guiding policymakers in managing quality of life and sustainability.

Stay tuned for our next installment, where we will delve deeper into individual life event predictions—an essential foundation for financial planning, which aligns with Bambu’s research objectives.

The author, Aki Ranin, is the founder of two startups based in Singapore. Bambu specializes in providing digital wealth solutions for financial services, while Healthzilla develops the Healthzilla health analytics app for iOS and Android devices.