Jonathan Blow's talk, "Preventing the Collapse of ...
# thinking-together
d
Jonathan Blow's talk, "Preventing the Collapse of Civilization", is about the ever increasing complexity and unreliability of software: why this is not sustainable, and what we must do.

https://www.youtube.com/watch?v=pW-SOdj4Kkkβ–Ύ

πŸ‘ 2
πŸ€” 1
n
getting Blow on the podcast would be great cc @stevekrouse πŸ˜„
πŸ‘ 2
w
I would support this.
n
I imagine he would be very critical of our movement... he tends to prefer systems programming rather than higher abstraction
w
Not a bad thing. (Haven't seen this talk yet.) Previously Blow's main goal is has been to talk to hardware so most abstractions as such are not going to help.
c
yes @stevekrouse lets get him on the podcast, and lets prepare some good questions
πŸ‘ 1
I think I have some good questions
Its very interesting to see him making this diagnosis's and yet come up with quite different conclusions. Its not that he is the first to note the systems
I find it also a bit surprising that nobody here talks about the root cause he is referring to: (largely) self inflicted complexity.
We have such bad measures of complexity , its almost frighting. Everybody seems busy inventing his own shiny new tool. Which will then magically solve all the complexity problems we have?
πŸ€” 1
You can not seriously believe this.
πŸ’― 1
"we" the IT/techsector/ computer academia / computer science have strong biases, as all humans do. But I would argue that many complexity issues exist because of our own bias/problems, as well as societies.
Is all hope lost? No there are things we can do, which could lead to better relationships with technology.
I have a longer post in the pipeline where I try to sketch some thoughts and possible approaches to solutions. I hope I can finish it tomorrow
Also interestingly Jonathan Blow is not only arguing against complexity but basically he is noting that information it self is growing
w
Listened to the talk β€” while Mac was installing updates coincidentally. @stevekrouse I think the most fruitful direction of inquiry would be to help pin down what an esthetics for things that "aren't a giant horrible mess" actually looks like. How do you tell whether things are improving? How do simplification and abstraction differ? The Jai language is certainly Blow's specific answer in his specific problems. He notes though that his own thinking that has been trained by his tools (C++ and Game Programming) and therefore will miss what lies outside.
c
nice Observations @wtaysom it lends nicely to what I tried describing in terms of formulating measures of complexity
You actually see this kind of transformation going on in a lot of places in modern society
Today most countries measure their "performance"/wealth using the GDP metric.
It is this metric that their economy and politics are optimizing for.
But should this really be the main focus of politics or society?
The real struggle comes from defining https://en.wikipedia.org/wiki/Eudaimonia in terms of a social contract
Since population is exploding and technology is increasing, we still lack a good medium to discuss Eudamonia on larger level (going beyond the nation state)
πŸ€” 1
i
Here’s some money where his mouth is: https://twitter.com/Jonathan_Blow/status/1131554763038973953
s
Interesting.. I didn't know that Blow subscribed to the idea that the Unix philosophy is the real problem.
πŸ‘ 1
i
He explains that thought better elsewhere in the twitter discussion β€” that the problems with Unix are that everything communicates via text (no structure, no types), and that all the micro-binaries should just (or also) be library functions.
s
Ah I see. His talk also mentions that LSP etc should be libraries not binaries. In Unix, libraries communicate via the C ABI, so it's better than text, I suppose.