“False news” is built-in?

From “Algorithms to Live By – The Computer Science of Human Decisions”, by Brian Christian and Tom Griffith (Henry Holt and Company, 2016)

p 147-148 (in Chapter 6 “Bayes’s Rule” ~ the statistics of causation and prediction)

Subheading: “Priors in the Age of Mechanical Reproduction

Introductory quotation: “As if someone were to buy several copies of the morning paper to assure himself that what it said was true.” – Ludwig Wittgenstein

“The best way to make good predictions … is to be accurately informed about the things you’re predicting. … … representing the world in the correct proportions – having good priors, appropriated calibrated. By and large, for humans and other animals this happens naturally; as a rule, when something surprises us, it ought to surprise us, and when it doesn’t, it ought not to. Even when we accumulate biases that aren’t objectively correct, they still usually do a reasonable job of reflecting the specific part of the world we live in. …

Everything starts to break down, however, when a species gains language. [emphasis added] What we talk about isn’t what we experience – we speak chiefly of interesting things, and those tend to be things that are uncommon. More or less by definition, events are always experienced at their proper frequencies, but this isn’t at all true of language. Anyone who has experienced a snake bite or a lightning strike will tend to retell those singular stories for the rest of their lives. And those stories will be so salient that they will be picked up and retold by others.

“There’s a curious tension, then, between communicating with others and maintaining accurate priors about the world. [“Priors” is statistical jargon for experiential background relevant to the subject matter when attempting to make predictions.] When people talk about what interests them – and offer stories they think their listeners will find interesting – it skews the statistics of our experience. That makes it hard to maintain appropriate prior distributions. And the challenge has only increased with the development of the printing press, the nightly news, and social media – innovations that allow our species to spread language mechanically.

“Simply put, the representation of events in the media does not track their frequency in the world.

“If you want to be a good intuitive Bayesian – if you want to naturally make good predictions, without having to think about what kind of prediction rule is appropriate – you need to protect your priors. Counterintuitively, that might mean turning off the news.” [end of quotation]

(The next chapter is titled “Overfitting – When to Think Less”…)

Views, interpretations of the above might be meta-discussion, relative to political reporting, perhaps even to the nature of internet forums. It also might reflect aspects of dhamma – i.e. dealing with conditioned defilements plaguing the human
mind.

3 Likes

I feel in future general public will know more about “bugs” of our mind once we can build AI that almost resembles human mind.

I would disagree with that statement about ‘proper frequencies’. I would rather say that language helps us to extend our ability of making predictions. Supposing that it is true that we mainly talk about unusual things (I have no idea if that statement has been quantitatively proven, but let’s go with it) - well, the common experiences we can learn to predict by our own experiences. But the unusual ones - since they are unusual, it is very difficult to gain sufficient data about them on which to build an accurate working model of what is going on. But if those are the very things which are spoken about, then we can process the information collecitvely. And we can even inherit it through the generations.

It is by such means, that medicine can develop, for example. A culture can gain knowledge of how to treat a wide variety of diseases or types of physical or even emotional damage, by having the sparse data recorded with speech, and passed on through stories and discussions. This can both warn of uncommon dangers, and preserve successful strategies for escaping or remedying those dangers.

The Buddha’s teachings are a good example of that.

[quote=“Senryu, post:3, topic:5801, full:true”]

…I would rather say that language helps us to extend our ability of making predictions. …[/quote]

I find your argument well-taken. It points out, perhaps, the limitations of computer science vis-a-vis the much broader perspective of cultural history. The analogy in medicine I can especially attest to.

The overall point by the authors of that book, however, I think does raise important perspective as to the more narrow scope of “truth” in the context of contemporary “media”, and the jungle of viewpoints and methods of persuasion, particularly in view of “globalization”, with which we are confronted, and perhaps occasionally distressed with, today.

(The whole issue of mathematical prediction is admittedly a bit abstruse to me, and seems more a matter of formalism that pertains to exploitation in specific areas of functional application – e.g. the current push towards saturating the internet with AI mechanisms focused, largely for commercial and political purposes, on deciphering and manipulating human behavior.)

1 Like