,
14 tweets,
4 min read
Read on Twitter
As I wrote this, I thought "Enough! You have gone far too far. Lies about lies? How can anyone believe anything anyone says?" "Exactly" my other self said, "that's the problem, you see". Then I remembered a paradox: the internet works by assuming that parts of it can & do fail.
The internet was designed to deal with the disaster, destruction & loss that a nuclear attack could impose on any widely distributed system of communication. Large parts of a network might vanish. How could the network enable any surviving parts to communicate with each other?
"The network infrastructure is considered inherently unreliable at any single network element or transmission medium & is dynamic in terms of availability of links & nodes. No central monitoring or...measurement facility exists that tracks or maintains the state of the network."
The trick was a system of packets. Each carries all or part of a message. Each knows where it's from, where it's going & how many more times it can try a new route. When it runs out, it just disappears. That's the Internet Protocol (see Wikipedia). It's what makes the net work.
We get reliable internet communication by sending unreliable packets and waiting for unreliable answers that tell us the packets arrived intact. The life of each packet is the number of routes it can try. When it dies, it just disappears, so that damage doesn't cause gridlock.
In the internet we get reliable communication by assuming that the system cannot be relied on. In real life news & opinion we do not & will never trust any central authority to tell us what is or is not a fact. So we need a way to decide what's sufficiently true to be useful.
The #IntegrityInitiative claim to tell us what's true. They say we're in a new war in which 'everything is a weapon' & in which our 'enemies' want to undermine everything: "our very concept of democracy...the very understanding of truth...our values.”
The #IntegrityInitiative & others tell us by their own words exactly what they don't want us to think: that their mark on a story acts as a warning to all to distrust it. They will tell some truth. We need to know what they want us to think & why they want us to think it.
We start by accepting that we cannot rely 100% on any news that we get from any possible source. We need an 'unreliable news protocol' to get us news we think we can trust. Maybe Professor Paul McKeigue can help us. Here he is, on @tim_hayward_'s blog: timhayward.wordpress.com/2017/08/11/how…
Correction. We have an 'unreliable news protocol' once we decide that we cannot in general rely on the news. What we need is a decentralised means of extracting reliable news from the unreliable noise of raw news. It has to be tamper-proof, widely distributed, subject to none.
Those 2 pieces by Paul McKeigue are not easy. He refers to a paper by Jack Good (Turing's assistant) who notes: "... people are by no means perfect Bayesians. But even dogs are fairly good Bayesians otherwise they wouldn’t survive as long as they do." Are we there with the dogs?
(Jack Good: "This surprising theorem can be expressed in the...form that the standard deviation is close to 3 times the square root of the expectation...this is perhaps the most terrifying theorem in mathematics...it shows how easily evidence can point in the wrong direction.")
What (I think) we get is the sense that the method is powerful, goes back a long way, was used & developed by Alan Turing at Bletchley & so helped us survive WW2. It corresponds to the way we really should think & exposes serious mistakes that most of us make. How can we use it?
@threadreaderapp unroll please
