4 ms·

Intelligence Augmentation and the Myth of the “Golden Lost Age”

49 points by jf 11y ago

amirouche 11y ago
To be honest this is a "me too IA vs AI" kind of post. I upvoted to continue the discussion Previous posts: - AI vs IA https://news.ycombinator.com/item?id=10373180 https://news.ycombinator.com/item?id=10373180 - Machines for thinking: Computers will get smarter, but with humans in charge https://news.ycombinator.com/item?id=10324648 https://news.ycombinator.com/item?id=10324648
- joe_the_user 11y ago
  I disagree. The article both does a good job of outlining the issues and makes some interesting and reasonably substantial comments on these issues. An article doesn't have to be earth-shattering to be useful.
willholloway 11y ago
> “pop software” as Alan Kay calls it. There was a natural trend in the PC era to make computing devices available to everyone, which I think was a positive thing; this necessitated moving away from software that needed training or methodologies, for example. It also required software that many people would use, as opposed to more niche applications, in order to make the profit necessary to support a company. Unfortunately this has made it harder to create the more expert-oriented applications that could co-evolve with users via training and methodology that would be akin to calculus or music notation, or “bicycles of the mind” as Steve Jobs said; The greatest productivity boost I have yet found, and the closest I've been able to come to a mind-machine interface is debian linux + tiling window manager + command line proficiency + customization of text files for creating shortcuts, tailoring work environment. It has a learning curve, but the payoff is immense. I think we should be teaching linux and the command line in schools, even if we don't teach kids to code. The computer is the most powerful and important tool we have, and we all use it daily. Not teaching expert level proficiency is the same as sending kids into the world handicapped.
- vezzy-fnord 11y ago
  Linux is actually a large part of the "computing pop culture" that Kay laments, FWIW.
- robbintt 11y ago
  > customization of text files for creating shortcuts what do you mean by this?
fr0styMatt2 11y ago
There's a bit of a 'reboot' as such (I forget what it's generally referred to) that contributes to this. By that I'm referring to the way that microcomputers reset progress in software. When low-power microcomputers hit the masses, the advances that had happened in software in the minicomputer / mainframe world couldn't follow along. So in the first instance you had computers that could only barely run BASIC programs and where you were programming in assembly close to the bare metal if you really wanted to do much more advanced things with the machine. Now, the kids growing up on these computers in the 70s and 80s (I'm an 80s kid myself) had no idea about things like GUIs or the latest in virtualization technology on minicomputers or all the many problems that had been solved in big-mainframe world but were yet to hit consumer devices. One factor is that consumer hardware of the time wouldn't have run the software efficiently anyway; another one is that there was no Internet where you could just Google anything and find out. So many of us then grew up in this era and became software engineers that would go on to write operating systems and software of the 90s and beyond. We'd never seen or heard of, for example, Lisp Machines or what they could do. Which is I think why you end up with this weird generational gap, almost like a chasm of knowledge born in the 80s and 90s. Whenever I watch an Alan Kay video I'm blown away by how much was possible 'back then'. My mental map of technological progress starts with 8-bit micros in the 1980s, which we all thought were 'cutting edge technology' except that they weren't, in the broadest scheme of things. It's this amazement that I think leads to the feeling of a 'Golden Lost Age'. Remember when Windows 95 touted preemptive multitasking as a groundbreaking new feature? Or when DMA for hard disks was a 'new' thing? (except that the Mother Of All Demos basically showed the concept, from what I vaguely remember from watching it many years back). We see the same cycle in mobile computing - phones were once close-to-the-metal devices; now they're running full mutli-tasking basically desktop-class operating systems. The difference these days is that we have the Internet and we have the lessons of history actually available to us.
- agumonkey 11y ago
  I remember that awkward thrill to see a student thesis from the 60s being more thoughtful than the latest vector drawing program from A. And I've toyed with a long list of advanced packages, none of them has that tiny geometry solver in them. Time is surely not an arrow in progress. Lot's of colateral and inherited sub cultures are sucking energy. As you mentioned, a new 'market' is also a potentially huge drawback, but that's such a common thing. People will see the world their way, not as a PhD knowing the history and state of the art (even researchers don't know all). It happens in programming languages too. All the web started a freeform joyful environments, unlike c++ or java with their heavyweight specs and standards. But then complexity hits and they're suddenly bringing a lot of structure, types, conventions, etc etc. It's like a child, who cannot enjoy his parents universe, he needs something compatible with his new mind. ps: my latest 'the newest is less than the old' moment was realizing Haskell was specified in 1990. At that time mainstream users were given Windows 3.0 and DOS :)

49 points by jf 11y ago

To be honest this is a "me too IA vs AI" kind of post. I upvoted to continue the discussion Previous posts: - AI vs IA https://news.ycombinator.com/item?id=10373180 https://news.ycombinator.com/item?id=10373180 - Machines for thinking: Computers will get smarter, but with humans in charge https://news.ycombinator.com/item?id=10324648 https://news.ycombinator.com/item?id=10324648

joe_the_user 11y ago
I disagree. The article both does a good job of outlining the issues and makes some interesting and reasonably substantial comments on these issues. An article doesn't have to be earth-shattering to be useful.

willholloway 11y ago

> “pop software” as Alan Kay calls it. There was a natural trend in the PC era to make computing devices available to everyone, which I think was a positive thing; this necessitated moving away from software that needed training or methodologies, for example. It also required software that many people would use, as opposed to more niche applications, in order to make the profit necessary to support a company. Unfortunately this has made it harder to create the more expert-oriented applications that could co-evolve with users via training and methodology that would be akin to calculus or music notation, or “bicycles of the mind” as Steve Jobs said; The greatest productivity boost I have yet found, and the closest I've been able to come to a mind-machine interface is debian linux + tiling window manager + command line proficiency + customization of text files for creating shortcuts, tailoring work environment. It has a learning curve, but the payoff is immense. I think we should be teaching linux and the command line in schools, even if we don't teach kids to code. The computer is the most powerful and important tool we have, and we all use it daily. Not teaching expert level proficiency is the same as sending kids into the world handicapped.

vezzy-fnord 11y ago
Linux is actually a large part of the "computing pop culture" that Kay laments, FWIW.
robbintt 11y ago
> customization of text files for creating shortcuts what do you mean by this?

fr0styMatt2 11y ago

There's a bit of a 'reboot' as such (I forget what it's generally referred to) that contributes to this. By that I'm referring to the way that microcomputers reset progress in software. When low-power microcomputers hit the masses, the advances that had happened in software in the minicomputer / mainframe world couldn't follow along. So in the first instance you had computers that could only barely run BASIC programs and where you were programming in assembly close to the bare metal if you really wanted to do much more advanced things with the machine. Now, the kids growing up on these computers in the 70s and 80s (I'm an 80s kid myself) had no idea about things like GUIs or the latest in virtualization technology on minicomputers or all the many problems that had been solved in big-mainframe world but were yet to hit consumer devices. One factor is that consumer hardware of the time wouldn't have run the software efficiently anyway; another one is that there was no Internet where you could just Google anything and find out. So many of us then grew up in this era and became software engineers that would go on to write operating systems and software of the 90s and beyond. We'd never seen or heard of, for example, Lisp Machines or what they could do. Which is I think why you end up with this weird generational gap, almost like a chasm of knowledge born in the 80s and 90s. Whenever I watch an Alan Kay video I'm blown away by how much was possible 'back then'. My mental map of technological progress starts with 8-bit micros in the 1980s, which we all thought were 'cutting edge technology' except that they weren't, in the broadest scheme of things. It's this amazement that I think leads to the feeling of a 'Golden Lost Age'. Remember when Windows 95 touted preemptive multitasking as a groundbreaking new feature? Or when DMA for hard disks was a 'new' thing? (except that the Mother Of All Demos basically showed the concept, from what I vaguely remember from watching it many years back). We see the same cycle in mobile computing - phones were once close-to-the-metal devices; now they're running full mutli-tasking basically desktop-class operating systems. The difference these days is that we have the Internet and we have the lessons of history actually available to us.

agumonkey 11y ago
I remember that awkward thrill to see a student thesis from the 60s being more thoughtful than the latest vector drawing program from A. And I've toyed with a long list of advanced packages, none of them has that tiny geometry solver in them. Time is surely not an arrow in progress. Lot's of colateral and inherited sub cultures are sucking energy. As you mentioned, a new 'market' is also a potentially huge drawback, but that's such a common thing. People will see the world their way, not as a PhD knowing the history and state of the art (even researchers don't know all). It happens in programming languages too. All the web started a freeform joyful environments, unlike c++ or java with their heavyweight specs and standards. But then complexity hits and they're suddenly bringing a lot of structure, types, conventions, etc etc. It's like a child, who cannot enjoy his parents universe, he needs something compatible with his new mind. ps: my latest 'the newest is less than the old' moment was realizing Haskell was specified in 1990. At that time mainstream users were given Windows 3.0 and DOS :)