Hacker News Reader

It's not just speculative. It's a retrospective on the early mainframe era.

The nitty gritty punch card programming was something stereotypically done by middle aged/older women. Many early computers were massive and housed in what could be considered temples in big universities surrounded by gothic architecture. They were expensive to replace and kept around for long periods of time because of how difficult it was to upgrade (some large mainframe systems still exist today for that reason). Mechanical computers like the Babbage difference engine would last a lifetime. The idea of an average Joe spending all day using expensive compute time was insane. Those who focused on them did so for academic reasons, and spent most of their time meditating on computation, math, and the structure of language and meaning.

It brings in those aesthetics, but then gives individuals as much computer time as they want. Then zero people make spreadsheets.

Imagine a world where you can just assert any fact as the starting point of speculative fiction. Then imagine that someone would write a piece where computers were inherently old.

Expert prompt crafting!

A sci-fi magazine has cut off submissions after a flood of AI-generated stories - https://www.npr.org/2023/02/24/1159286436/ai-chatbot-chatgpt...

> The science fiction and fantasy magazine Clarkesworld has been forced to stop accepting any new submissions from writers after it was bombarded with what it says were AI-generated stories.

> The magazine officially shut off submissions on February 20 after a surge in stories that publisher and editor-in-chief Neil Clarke says were clearly machine-written.

I got some pretty interesting sci-fi type results for that in Stable Diffusion. It helped to negative-prompt titles, text, words, etc

I'm so mad you made this comment before me.

This breaks immersion for a lot of people, me included. It's one of the biggest issues I have with lots of sci-fi (moreso than most other types of speculative fiction) novels. Just asserting a fact this fundamental and building your world around it only works if societies, culture, technologies, and history are build around it. Humanity's history with tools and automation are millenia old. To alter this would require a lot of changes. We've been automating, forgetting, copying, and maintaining technologies for as long as our history.

Alien technology encounter stories satisfy some of this for me, as the injection of a foreign, advanced technology makes sense in this framework, much like how ancient peoples would discover technologies built by neighbors or rivals they didn't communicate regularly with.

To some degree I agree with your last point. Before the idea of aliens, writers like Plato and Thomas Moore would posit some undiscovered island where they could imagine a culture unlike their own existing, which like with aliens, gave plausibility as to why that culture would have a different history and traditions than their own. Of course they, like many speculative writers today, were really writing about how they wished their own society worked.

Hmmm. Good observations. I hope this doesn't break my interest in the Warhammer universe, lol, which I believe probably suffer from a lacking "origin story"

It's like asking "would you believe in homeopathy if doctors all told you that it was correct?"

To which the answer is "Doctors believe things for reasons. You can't just flip a switch and change what doctors think without changing the entire world so that homeopathy is actually true, and that would be such a weird world that science in it would be unrecognizeable."

(And since I'm not ChatGPT, I can answer your entire question: "Such a story would have an incoherent world. I could imagine someone writing a story that takes place in an incoherent world, but you wouldn't be able to get useful insights from it.")

Step 1: for some reason, the world stops making computers, but keeps the ones that already exist

Step 2: 100 years pass

Step 3: computers are now an old (and dying) tradition, kept going only by a small group of people who understand how to care for the machines.

Was that so hard?

Step 1 is really hard if computers are useful (why do you stop making something that's useful?), but step 3 is simply not believable: a society that has machines that could, at the minimum, help solving complex calculations and engineering problems much better than humans decides to completely ignore them. That's not how societies work.

Unless they have other means of solving those calculations better than computers; in that case, call those other means "computers" and the study of older computers "retro-computing".

something like the amish, relying on 100 year old technology developed by humans instead of those newfangled tools that aliens brought that no human can understand.

1 is easy to imagine - a large enough collapse that making new ones is very very hard, especially compared to scrounging and repurposing the huge amount of existing ones. Imagine a huge population decline, and search parties looking through the ruins for useful tech, like computers, decades later.

Such a society could probably make new computers, but the cost of making new, much worse specced computers, may not be cost effective for a long time as long as old computers can be found and repaired.

Edit: such a movie or novel would make for great Fallout-ish world-building. Imagine for instance a large hotel in a big city running the booking system on an old laptop in a central secure location, and new 8-bit computers used as terminals at the front desk. Public access 8-bit computer terminals at public libraries connected to copies of WikiPedia running on some old gaming rig and such things. :-)

No it isn't (No1 easy to imagine). Computers are only useful because we have a stable society that we can build on. Post apocalyptic societies have zero practical use for computers.

Post-apocalyptic doesn't necessarily mean unstable. We weren't post-apocalyptic back in the 1930s or 1950s, and computing machines had plenty of uses already (and even earlier).

> Post apocalyptic societies have zero practical use for computers.

Post global nuclear war reconstruction would really benefit from having even a simple spreadsheet program to aid with resource allocation, wouldn’t it?

> Whatever you do with them [computers] is automatically seen as practice of an ancient and unchanging tradition.

During a nuclear war reconstruction scenario, computing wouldn't be seen as ancient or unchanging. We'd be in a transitory period where we try and recreate computing infrastructure with what we have and the power sources we have access to. But we wouldn't develop an ancient tradition or anything.

At lot can be done with pen and paper but if you were rebuilding civilisation from scratch I'd be be happy to have a solar powered computer holding a copy of wikipedia and heaps of documentation to help me put to use any artifacts I'd find from the old world.

Step 1 and 3 are, arguably, quite probable to happen in the future. I don't think you're appreciating just how many circular dependencies we have in the global supply chains. We're one big cataclysm or global war to reverting back to pre-industrial times, and if that were to happen, we'd likely be stuck there for quite a long time. Consider the following observations:

A. For obvious reasons, we continuously improve tooling and processes. This creates circular dependencies that make our civilization non-restartable.

As a toy example, imagine we made tool A, which let us create tool B, which let us, among other things, upgrade A to A'. A' can still produce B, but can also produce upgraded B'. B' obviously lets us make A'', but also C', and A'' + C' let us make B''. Rinse, repeat for a hundred years. The result is that the whole industry relies on tools A'''''''''' and B''''''''''. The hobbyists still play with A and B, maybe even A' and B', but there's nothing between A' and A'''''''''' in actual use or production, nor anyone who remembers their design, because they've all been made obsolete long ago. If the calamity hits, and we can no longer make new A'''''''''' and B'''''''''', we don't get to downgrade a step or two - we go back all the way to A' and B'. And all the other tools and technologies that also relied on our advanced capacity - they all stop working and go "back to beginning" in lock step.

B. Making and operating tools involves ever increasing energy use.

C. We've used up all easily-available high-density energy sources long ago. Both our renewable tech and our fossil fuel mining and combustion processes are A''''''''''-level tech.

With that in mind, it's not that big of a stretch to write a plausible backstory for a culture as described in the article.

Even if we're ignoring wind and water power, boosted by charcoal, computers would be quite useful for lots of business purposes.

And if things are so low tech, the long-distance communication would be far more impressive than the story suggests!

I can't decide if I would enjoy reading real examples of these cycles. Just knowing how precarious the supply of natural rubber is, and how inadequate artifical solutions are (including prospective ones from continental), makes me uneasy, but I'll be damned if there isn't some kind of allure there as well.

It seems almost certain that climate change is going to severely disrupt several of these cycles in decade or two, and a slightly different failure mode of what you allude to in C. is that our tech and processes become capped before we adequately scale up renewables, and overshoot kicks in and it's actually a negative feedback loop that certainly doesn't just stop with computers.

I can imagine it, but it would require a lot of worldbuilding to explain.

I reckon the primary means of computing for the lay person becomes interacting with prompts to make the computer do what you want, but the actual knowledge of software engineering and computer science is gone. Even the prompts, due to linguistic evolution, are in a language that regular people don't understand and people's mastery of the language is akin to moderns trying to reconstruct something like Akkadian. We sort of know what specific words mean and the general grammar, but we simply don't know 90% of the vocabulary to express most ideas. Whatever the language model was populated on training data from an archaic language and is also full of weird idioms and phrases and misspellings that we can't reconstruct in an organized way.

In this context, the computers might be able to solve some basic problems for us, but everything we do with them is basically just received incantations from other crusty old people. Innovations keep happening, but they're small scale and mostly revolve around finding out new ways to tweak the incantations to do novel things. But the most we can do is say, like, "draw a picture" or "sort this list."

As I was reading the post I assumed (1) happened because something better came and completely replaced computers: magic, raw energy manipulation, biomachines, hextech... something like that.

1 happens every fucking day, hell wasn't there an article today about people who have older equipment that depend on floppy disks and how hard it is to find reliable disks these days? The article discussed an older embroidery machine, the only way to supply it with patterns to embroider was 3.5 inch floppy. His are starting to fail. But this is also true of older samplers, drum machines, synthesizers, etc.

Sometimes someone manages to spin up new manufacturing of retro components or things to replace the retro components (there are a few pcbs running around that can replace floppy drives that conform to certain interfaces with usb sticks or sd cards or even a network): Vacuum tubes, parts for vintage cars, parts for vintage watches, vinyl records, etc. but often before that happens the entire known supply needs to be scavenged.

So I can imagine a world where we manage to start doing the things we are doing with some new technology that is not digital silicon based computers that run on electricity. But there will be edges where the new technology is not compatible with some system, and in that edge lives this kind of speculation.

We stop making specific models, and lose how to make spare parts for those models. We don't lose how to make the machines in general.

This story would be completely different if the old silicon computers were replaced with neo-computers.

Also for anything that uses a standard floppy drive there are cheap and plentiful adapters to use flash storage instead of disks.

I have yet to see a thing I can insert into a floppy drive that will pretend to be a disk. I have seen plenty of things that can replace the drive itself with something that pretends to be a floppy drive.

The problem with those are that not all equipment that uses floppy drives conforms to standards. If you have a 50 year old embroidery machine from a company that no longer even exists you'd need someone pretty fucking technical to pull that shit apart and figure out if any of those drive replacements will even work in your weirdass machine. If you are lucky your machine used some off the shelf drive from a reputable company (I've seen some old weird drum machines machines that had Apple Disk ][ drives hanging off them) but otherwise...

hell the living computer museum in seattle, before it's at least physical closure by the sister of paul allen, is/was a bit like this. They have a bunch of vintage big iron that they have restored to working condition and allow members of the public to use. Some of these are still available on the internet.

https://www.livingcomputers.org/

https://www.livingcomputers.org/Computer-Collection/Online-S...

Hell deep in the chip shortage I talked with more than one person that were buying up electronics at thrift shops and ebay that they knew contained chips they could not source at the time.

That's just 40k with extra steps

I'd say it's 40k without a whole lot of extra steps.

The issue is with Step 1 - there is some reason why the world chooses to stop using computers despite (as far as I can see in the 'thought experiment') them being available and the homo sapiens needs/desires because of which we use(d) computers seemingly being the same. That reason isn't provided, isn't even implied or hinted to, however it's something major with far-reaching consequences to human behavior.

So it's kind of frustrating - in effect, the author describes a thought experiment of how society would change if X happened, but without saying anything at all about that X other than asserting that one of the things X causes is a mysterious lack of interest in using computers to fill needs/desires which (at least according to the original article) aren't filled by some better alternative.

You're basically missing the whole point of speculative fiction in this style, which is "we start with imagining a world which has these conclusions. Trying to imagine it forces us to try to make sense of what could lead to those conclusions."

You assert it's incoherent. That's part of the point: it's incoherent with your current understanding of our current world. There might be other circumstances that lead to a such world without being incoherent. If the conclusions in a piece of speculative fiction are appealing, it may be worth thinking of what it takes to reach those conclusions.

The thought exercise of trying to make it coherent is part of the point.

No, the point is it can't be done. It's like cleaning your house with an 100 year old tortoise: you absolutely would do that, if tortoises were plentiful and do a good job cleaning. Why would anyone clean their houses themselves, when there were all these old tortoises willing and able to do the job for them? Now if the tortoise moves all your crap around so you can't find it, leaves streaks on your mirrors and windows, fire his scaly butt!

The parent is describing an inductive approach to worldbuilding, while you are describing a deductive approach. Two sides of the same coin.

Good speculative fiction at least tries to give some reason why something might come to be the case. This article just says "pretend computers are old; old things are like X today; therefore, computers are going to be just like X," while ignoring all the things that are different.

>I can't conceive of such a world.

One word: Terminals.

Computers are inherently old at this point, the technology harkens back at least the better part of a century.

The old guard in particular cannot and will not let go of CLIs and esoteric flags and arguments, and will berate anyone who would dare suggest anything concerning a GUI.

To them, the only human interfaces that exist are the keyboard and the monochrome monitor, computing is still grey-on-black (or green/orange-on-black) monospace text and running emacs or vim. The rest of the world disregards them as living relics of ancient history.

Imagine a world where everybody simply has an AI assistant in their pocket, that simply tells them anything they might need to know. Design programs for you. Draw you any image. Perhaps even construct virtual worlds. We are almost there already.

Only weird old people would bother to learn how to program dumb desktop computers.

There's a lot of people who feel disquieted by the pace of software systems and aspire to do work in software that feels similar to the more methodological, slower paced work of more mature fields like HVAC engineering or structural engineering. These folks find solace in this kind of speculative fiction.

I feel their disquiet is misplaced. Any young, fast-moving field is going to be full of the same issues that software is in now. You can look back at the history of mechanical engineering during the industrial revolution to see many similar problems we had with unsafe projects, hyped-up snake oil, or iteration for the sake of iteration. The history of automobiles and aviation was also marked by similar issues. Slower paced engineering fields are more mature and have gone through decades or centuries of iteration before coming up with tried and true solutions. But fundamentally fiction speaks to the soul more than it speaks to any measurable outcome. Truly the only fix for this disquiet is to search inside rather than look out. Fiction can be a great tool for that.

This is not unlike the speculative future of Anathem.