Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

We Built It with Slide Rules. Then We Forgot How

unmitigatedrisk.com

Post reply on HN
Print view View original

21–30 of 41 posts

Re: We Built It with Slide Rules. Then We Forgot How

#21

Post by bluGill »

WillAdams wrote: post #10

Ages ago, I worked at a flexographic print manufacturer, once, when a new hire had made a large plot on Kraft paper (which was moderately expensive/difficult to source and a nuisance to switch to/from), it turned out a circle was on a non-printing layer (why Adobe Illustrator allows that is a separate discussion --- Freehand's printing everything which is visible and not printing anything invisible or on the backgrou…

Great that you knew that - but that doesn't mean it was worth it for the kid to learn.

There is more interesting/useful things in life to learn than you will live. Just becoming a brain surgeon, heart surgeon, anesthesiologist, and other somewhat related medical specialties will take you to retirement age without ever leaving school. That is despite the overlap, we haven't even start to make you any form of engineer, musician, or any other the other interest fields there are out there.

We as a society have to look at things like manually drawing as hobbies you can learn if you want that should be put in a book just in case someone wants - but otherwise not taught. There is nothing wrong with what you knew how to do, but there are more important things to teach kids and we need them to move on to the real world not learning everything.

Re: We Built It with Slide Rules. Then We Forgot How

#23

Post by rawgabbit »

oersted wrote: post #20

I don’t quite understand this alarmist argument about AI making us forget how to build software. We are software engineers, we are used to this! The whole history of computing has been about creating higher abstractions to make it easier to build software. Who has thought recently about instruction sets, memory layouts, gotos, pointers, system calls? Some still do, but not everyone has to anymore. From day one I had…

React, Spark, Elasticsearch, AWS or Rust were deterministic programming languages; they did exactly what the developer specified.

With Claude Code, they are semi-independent non-deterministic agents; they are more like consultants that the developer manages. The fact that they tend to generate verbose code which overwhelms the developer's ability to review is also troubling.

Re: We Built It with Slide Rules. Then We Forgot How

#24

Post by commandlinefan »

The core problem is the quixotic quest for efficiency. Right now I'll blame JIRA because that's the latest incarnation of this beast, but it's the mindset behind thinking that's a good idea in the first place. As long as I've been working I've been under artificial, meaningless time constraints that seem to only exist to catch cheaters, but that actually serve to make experimentation impossible.

Re: We Built It with Slide Rules. Then We Forgot How

#25

Post by sifar »

oersted wrote: post #20

I don’t quite understand this alarmist argument about AI making us forget how to build software. We are software engineers, we are used to this! The whole history of computing has been about creating higher abstractions to make it easier to build software. Who has thought recently about instruction sets, memory layouts, gotos, pointers, system calls? Some still do, but not everyone has to anymore. From day one I had…

>> The whole history of computing has been about creating higher abstractions to make it easier to build software

These abstractions are understandable and predictable. There is no mental model for the current LLMs(in entirety, even though the parts are known), the output might as well come from a genie.

Re: We Built It with Slide Rules. Then We Forgot How

#26

Post by stephc_int13 »

Generational knowledge loss is often either discarded as irrelevant, illusory or misunderstood.

It is not a new phenomenon and can easily be traced back to antiquity.

Because _reality has a surprising amount of details_ the entire humanity knowledge at any given time is living in our memories, not written, and even if we had the time and will to try and formalize it, language is not complete enough and we lack the ability to fully introspect what we know.

You can ask a professional Tennis or Chess player to formalize his expert knowledge and it may contains some useful insights, but far from enough to replicate his skills.

So learning is re-discovering many things, a Sysphean task, and the majority is lost, we managed to keep just enough thanks to the invention of writing and books to reach a kind of slow escape velocity.

Because technology is constantly evolving, what is lost is not systematically relevant, like writing poetry in ancient Greek.

But there is the risk of losing too much, too quickly. As a veteran of the videogame industry I can attest that many mistakes that are made today were solved before, but the good designs and principles were largely lost.

Newcomers are not inherently less smart than their parents, quite often they just don't learn because the incentives changed.

I am not entirely convinced the emergence of "vibe coding" and other assistants will be a net gain.

Re: We Built It with Slide Rules. Then We Forgot How

#30

Post by nh23423fefe »

sifar wrote: post #25
oersted wrote: post #20

I don’t quite understand this alarmist argument about AI making us forget how to build software. We are software engineers, we are used to this! The whole history of computing has been about creating higher abstractions to make it easier to build software. Who has thought recently about instruction sets, memory layouts, gotos, pointers, system calls? Some still do, but not everyone has to anymore. From day one I had…

>> The whole history of computing has been about creating higher abstractions to make it easier to build software These abstractions are understandable and predictable. There is no mental model for the current LLMs(in entirety, even though the parts are known), the output might as well come from a genie.

That's not true. The mental model is a distribution over completions. Weird you say LLMs are like genies, genies being magical beings that can accomplish anything.
Post reply on HN