Hooking the following into this. https://twitter.com/TACJ/status/1265023819917545473

Politics isn't about building a bridge, it's precisely about the things that aren't reducible to an objective outcome. It's about values and preferences and culture.

This is fascinating. It's not *coherent*, but each individual sentence and paragraph feels... plausible.

It's that classic mid-oughts blogging style of assuming everything can be searched, so you're under no obligation to explain what it is you're talking about.

Bret Victor.

I *think* (and this isn't snark) that Cummings might have been impressed by the guy who's trying to make the Iron Man HUD/Minority Report interface.

Cummings:

And here's the stuff, and it's cool stuff! https://dynamicland.org/ 

I absolutely agree with Cummings about the importance of immediacy and so forth. Clear and clean feedback is everything in any context. (NNT wouldn't admit it, but this is a big part of what 'skin in the game' actually means).

Cummings is confused about the distinction between programming languages and user experience. He thinks that software UE is bad because the programming languages are bad.

Yeah yeah

Our government is being run by a DataViz guy.

What's interesting is the mix of reasonable points (academic publishing is terrible) and insane notions (if we had cool visualisations politicians would make better decisions).

'My incompetence is someone else's fault' klaxon

I totally get it. These are the frustrations of someone who wants to speak maths, but doesn't

This is about the fact that academia is terrible, and also about how hard it is to interpret quantitive information if you don't have a maths A-level.

Academia is hyper-specialised, and engineering academia is hopelessly remote from practice.

That's *my* take on what Cummings is complaining about.

I'm a thoroughly mediocre 'engineer'; but I'm only one or two steps removed from people who have to put their hands on things.

I've had conversations with esteemed academics who, quite frankly, should not be advising government departments (despite, apparently, doing so).

But, and here's where I depart from Cummings, I don't draw any broad conclusions from this. It's just that people are people and 'competence' is a complex and difficult concept once you're out of the world of 'if you screw up, someone dies'.

'I'm annoyed because I can't just Google something and get a completely valid and robust answer'

The answer is because 'serious newspapers' are in the same business as the other kind of newspapers.

(ffs, I didn't even manage to finish readin this thing. I think I now understand by Boris Johnson trusts the guy so much.)

Viz: if you can write lots of words and it sort of seems as if you know what you're talking about, I'm going to give up and accept that there's a good chance you know what you're talking about.

An alternate reality where Cummings was capable of studying maths and/or decided to study philosophy.

Cummings is, if he is anything, a historian. He doesn't even have the economist's excuse for thinking the gold standard was anything other than a bad idea

He's an historian and he can't do maths. That's it. That's literally all there is

Yes. DataViz is great; but also maths is a language for talking about shape and form. It could be tought better, but that's what's interesting and worthwhile about it.

Viz is great. But it's not a substitute for actually learning the language.

I've been to the control centre for the distribution network for London, and I can confirm it does not look like this

He wants to be an engineer so much he's complaining that 10 Downing Street doesn't look sufficiently like the National Grid control centre

This is so dull. What is it the PM is supposed to be looking at that's as important as an SLD for a control engineer?

Running a country isn't the same things as running a rocket launch, or running an electricity system.

Has he ever been to, like, a normal meeting?

As someone once said, the purpose of a system is what it does.

A few months ago, I was stuck in a muddle trying to pay a company for doing some research work for us. Because I'm a fool, I didn't succeed. It didn't matter, they did the work gratis, because they were professionals, and they understood how the system worked.

It only occurred to me afterwards that it made absolute sense for the organisation to make it maximally hard for individuals to make it difficult to spend money. Whilst also making it easy for trusted partners to take their wage.