Trial Translation/Digital Divination Era

The age of digital divination

 An anthropologist asked: "What do algorithms and astrology have in common in the digital age of predictive technology?"

A man stands before a judge charged with theft. The judge checked a source that will decide whether to grant the man bail.

A woman stands before a judge charged with adultery. The judge consulted a source who will decide whether the charges are established.

In the first case, the source of information was a computer; in the second, a chicken.

Both cases are forms of divination: using something (or someone) to allude to the future or to discover hidden information about the world.

Anthropologists have long been fascinated by this phenomenon. As early as 1937, British anthropologist EE Evans-Pritchard recorded several divination methods of the Zand people in central Africa, one of which is the above-mentioned feeding poison to chickens by the divination. The death of a chicken tells anxious husbands and wives whether their spouse is cheating. Ethnographic and historical research has uncovered countless other forms of practice: the movements of animals such as spiders or snails, patterns burned on animals such as turtle shells or sheep bones, the shapes of leaves and branches, revealed by tarot cards and palm prints story, etc.

Likewise, digital divination through the use of some predictive analytics software promises to predict the future and reveal those underlying truths about the world. Forecasting tools based on machine learning and artificial intelligence are being applied to relatively mundane but complex phenomena, from predicting housing prices, to inferring movements in financial markets, to predicting customer satisfaction with a business. But machine predictions can also turn into serious areas: new computational tools that try to determine whether someone is likely to abscond while on bail, as we did in the opening paragraph, or predict that you will be born with Covid-19 What a serious disease.

Whatever its form, divination is often based on answering specific questions to provide a narrow view of the future—leaving little room for other possibilities to emerge.

A recent article from Harvard anthropologists Kevin Hong and Joe Henrich provides a comprehensive overview of divination. Such practices, they say, serve many purposes: from settling disputes, reducing anxiety about a situation, or clearing those indecisions and concerns, to inferring the meaning of the world, inquiring about fate, or deciding when and how to wage war.

Divination is a way of revealing information on important matters that were otherwise unknowable and inaccessible. But interestingly, fortune-tellers are often not considered to have this information. They are people who know the art - can read secrets hidden elsewhere: in cards, in tea, in guts, or in algorithms.

Calling soothsayers "skilled technicians" may sound odd among contemporary Westernized, post-Enlightenment people. But there is a long history of linking divination to mundane activities. In the fourth century BC, Plato wrote of divination as a highly valued professional skill (mantikē technē). About four hundred years ago, in the ancient Greek epic Homer's Odyssey, divination was listed as a skill alongside medicine and woodworking (though Cicero scoffed at it centuries later). Early Chinese thinkers also included fortune-tellers in the category of craftsmen. In fact, divination and science were once partners: in the 17th century, when the lines between science, religion and art were increasingly blurred, scientist Johannes Kepler worked on both astronomy and astrology, and wrote more than 800 papers in his lifetime. A map of the Temple of Heaven.

The idea of skilled technicians predicting the future or revealing hidden information about the world through some obscure, seemingly magical operation sounds similar to today's developers of digital prediction technology.

The similarities go far beyond that. In many societies around the world, fortune-tellers charge high fees for their services, and not everyone has the opportunity to consult them. Likewise, techies who are well versed in AI and machine learning also charge high fees for their obscure knowledge and the right to use the software they produce.

It's worth mentioning that there is one major difference between digital divination and more traditional forms: the potential scale of its negative effects. Much divination affects only individuals or family units, but algorithms employed by large corporations or governments, such as "predictive policing," perpetuate systemic inequality and racial profiling. This is because their predictions are based on data sets that are often already biased against specific groups of people.

Given their unreliability, why do many of us still trust machine or divination predictions so much?

To try to answer this question, anthropologists Hong and Henrich ran a simulation. They found several conditions that lead people to overestimate the effects of divination. One is having a strong prior belief that divination sounds convincing and attractive. I think the same applies to machine predictions. Another reason is that failed forecasts tend to be underreported or explained; companies doing AI forecasting projects have similar mechanisms to downplay their mistakes.

A third factor is that the potential benefit from prophecy is considered to be higher than the cost of going to a fortune teller. Considering that these predictions have the potential to predict whether you'll have a heart attack or whether a person will commit a crime again, paying for machine predictions, or tying a phone or smartwatch to your phone, is a small but worthwhile price.

Some types of divination are considered trustworthy and reliable because they claim the ability to communicate and communicate with supernatural beings - such as ghosts announcing the results of important events, such as harvests. Although there is nothing supernatural about digital divination, it also seems indescribable and indescribable.

In both cases, non-professionals cannot test for themselves whether the divination is actually valid.

This is where the danger comes in: Divination may not be based on causal links, such as the relationship between traces of snail mucus and heart disease risk. Likewise, much of AI today is not based on causal models of the world. Machine predictions currently rely on finding statistical patterns in big data, but correlations are not causation. Without a solid, causal model of the world, digital divination can end up triggering costly or dire false alarms, such as persuading people to seek medical care they don't really need, or setting bail money too high or too high. Low.

As deeply social creatures, humans tend to learn from others. Anthropologists have found that people tend to judge the effects of divination through indirect rather than direct experience.

In other words, all of these intellectual techniques—from divination to astrology to artificial intelligence (no matter how firmly grounded in our physical reality)—are augmenting and justifying our choices, not replacing or changing them , the best effect. Similar to when technologies are built on our expertise and experience, they become truly useful.

Whether digital, physical, or paranormal, all forms of divination seem to offer narrow visions of the future: predicting, for example, a single event that may or may not happen. But in believing these predictions, we may be ruling out other possibilities we have.

What if we rely on a more tedious approach - controlling the world's most powerful prediction engine: our brains?

Imagining a possible future, and all participating in the act of imagining, may well lead us to a future we prefer. It also means taking over the responsibility for designing these possibilities from those few techies and other professionals, and involving as many people as possible in mapping the future — especially those on the fringes, and those who are often underserved Groups that experience negative effects in the decisions they participate in. This will mean pushing the shift from "what will happen?" to "what if...?".

Matthew Gwynfryn Thomas: Anthropologist, working in the non-profit sector in London. His current work combines data science, social science and future research to address the needs of people in crisis. Matthew's doctoral and postdoctoral research is on the evolution of collaboration in marginalized communities.

By Matthew Gwynfryn Thomas

Original: https://www.sapiens.org/culture/digital-divination/

Translation: statement

The translation is for study and exchange only. Criticism and guidance are welcome. Please indicate the source when reprinting.