Recent searches
Search options
@chrisg@fosstodon.orgThe year is 2030.
Computers boot directly into the browser. IDEs are just a web app now, running in the GPU. No one knows why. Or how.
All programs run in 4 nested containers on top of a hypervisor abstracting over the 5 major computational clouds. The last time a branch was predicted correctly, in any CPU anywhere, was 4 years ago.
Cloud costs are withdrawn directly from your retirement fund.
Ext7 just came out, it's written in Javascript and uses AI to guess what the file may contain.
@chrisg
>The last time a branch was predicted correctly, in any CPU anywhere, was 4 years ago.
LMAO
@chrisg@fosstodon.org in 2035 a breakaway movement to rescue human sanity uses a kind of FORTH directly on the machine to propel a browser that doesn’t understand html or JavaScript or css, but instead, only uses the ancient forgotten wisdom of WML from the early days of the Nokia 7110, and very little of it connects to the ‘normal’ internet at all, it’s an ox-bow-lake bypassed ‘other’ internet – nobody promotes it, nobody talks about it, it’s like fight blub, you don’t talk about it
Dagnabbit, Pascaline! 
i know its wrong to ask ChatGPT for a prospect but so did I and it even gave me some smiles 
The Year 2040:
Computing and Society:
1. The Hyper-Browser OS: By 2040, the browser is the computer. Layers of abstraction have reached absurd heights, where every interaction is mediated through predictive AI models embedded within the browser's neural framework. Offline computing is virtually extinct. "Kernel" now refers to the latest browser patch.
2. AI Drives Everything: IDEs, operating systems, games—everything is a predictive machine-learning black box. Programmers are more "prompt architects" than coders, working in layers of probabilistic inference where "bugs" are just emergent features.
3. Energy Collapse: The computational demand of multi-cloud, containerized infrastructure now consumes a measurable percentage of planetary resources. Renewable energies are monopolized by computation. Rolling brownouts are considered acceptable trade-offs for the privilege of using Excel 2040.
4. Financial Systems: Retirement funds are depleted in real-time to cover computational overheads. A market for "offline years" has emerged, where users sell their digital presence to spend time disconnected.
Technology:
1. Ext9 & ExtX: Filesystems are now interpreted as "suggestions." Ext9 doesn't store data at all—it merely indexes AI's best guess of what your files "should" contain. Users live in perpetual Schrödingerian doubt of their data integrity.
2. CPU Branch Prediction Paradox: By 2040, CPU design abandons traditional logic altogether. Computing is probabilistic at its core, with hardware optimized for running Monte Carlo simulations instead of sequential logic. Quantum computing promises salvation but remains locked in academic experimentation.
Cultural Backlash:
1. Neo-Luddite Revival: The FORTH-based internet of 2035 evolves into a full-fledged movement. These "Neo-Mechanists" build systems using only direct machine instructions and low-level network protocols like X.25. They shun anything predictive or abstracted, viewing JavaScript as the downfall of civilization.
2. Digital Feudalism: The breakaway Neo-Mechanists form isolated communities with strict rules: no cloud, no JavaScript, no containers. In these havens, simple, efficient devices run single-tasking FORTH-based OSes on 1980s CPUs. Their philosophy spreads underground, infiltrating corporate systems through "Trojan simplicity."
The Paradoxical Web:
1. Two Internets: The traditional internet continues as a bloated, cloud-driven monstrosity, accessible only via hyper-browser. Meanwhile, the WML-inspired "Othernet" flourishes quietly, powered by ad-hoc networks of long-abandoned devices. These separate realities rarely intersect.
2. Tech Collapse Warning: By 2040, mainstream systems are so complex and fragile that minor bugs cascade into global outages. Calls to "reboot the stack" grow louder, as the weight of abstraction becomes unsustainable.
Final Notes: In 2040, humanity stands on the precipice: either to simplify and reclaim control over technology, or to dive deeper into the abyss of predictive computation and abstraction.
@chrisg 2030? This just sounds like a Chromebook in a few months
@engravecavedave@mastodon.social @chrisg@fosstodon.org I came here to say this, except “already happened” re: chromeos.
@chrisg Windows XP is still used on boarding terminals at airports around the world.
@rysiek That's only because the marauding Rectifier(TM) teams from Microsoft prioritize hunting down Windows 98 SE installs.
@rysiek @chrisg Siemens was hiring Windows 3.11 Admins last year. 3.11 will be definitely around, too.
https://www.heise.de/news/Deutsche-Bahn-sucht-Admin-fuer-Windows-3-11-for-Workgroups-9611543.html
️Arter 
@chrisg JavaScript has 1400 package managers
@chrisarter That's probably the magic number.
@chrisarter …and 382 package manager managers.
@chrisarter @chrisg No no, OP it talking about the future not the present 
@chrisg Alllrightythen...
@chrisg new ext version every two years??
Copilot has replaced all programmers by 2028 and keeps spitting out ext implementations. No one can stop it.
IDEs are just a web app now
@chrisg I use code-server locally and load it in Firefox so I get vscode without an electron instance. The future is now.
@em That's a good point. I'll try to do better for my 2035 predictions.
@nshephard
I was not aware of #nyxt , but it sure looks interesting. Thanks for the hint.
@Reiddragon Yeah, this post is doing numbers and I'm not sure if that's because people think it's funny or because it's true.
Probably both?
I mean, it'll probably be called something different than ext7 - openAI will put their marketing team on it.
@chrisg@fosstodon.org This fills me with as much despair as climate change.
@neia That's ok, they are basically the same thing.
Or, if that's too hot a take, they share common causes.
@chrisg@fosstodon.org Some days, I feel like we should go back like 15 years in computer hardware tech and just stay there for a few decades. Force developers to care about performance again. Force us to stop using so many layers of frameworks and indirection, move toward simplicity. Or at least stop the bleed a bit.
@neia Completely agree, and (shameless plug) I wrote as much in my small dbms manifesto
https://radiki.dev/posts/a-small-dbms/
TL;DR I hate the cloud and what it represents. I just want us to stop wasting all that energy, creativity and resources just to sell ads.
@chrisg @neia Great post, thanks! A question, though. In my MediaWiki world "scaling" means that our wikis have millions of articles nobody cares about, but a few that get hit millions of times. Per second. And you don't know which.
I think this is similar in many software systems. Shops for example have a few bestsellers, or people buying the same stuff all at the same time. You don't want the system to crash exactly when you can make the most sales. Isn't this what scaling is about?
I agree that this can be one definition of scale. In my post, I talk about database scaling as processing large amounts of data. Not as the ability of a system to manage a large number of requests. That is a different challenge, that spans a larger part of the software stack. Databases are part of it, but doing that and managing petabytes of user data are different problems with different solutions and different implications.
@chrisg Funny thing: I was working on a filesystem using AI* to guess where to put file on a heterogenic disk matrix in most efficient way for long term storage cost.
[as always - AI was a simple model that could be just a bunch of ifs and elifs instead, but wouldn't sound so fancy].
@kbielnik Please tell me you named it ext5
@chrisg much to realistic for my taste!
@chrisg Hello world webpages requires a 7 node, 160 CPU cluster running Kubernetes and a pile of 67kB of IASC yaml. The webpage weighs 1.2MB.
(IASC stands for Intrastructure as Shitty Code, everyone does this now)
@chrisg that future is one I wouldn't want to even contemplate, and am trying to think, as much as possible given the evidence, that it and other such similar scenarios never come to pass. All this free software work, all the cultural shift, people slowly realising what's happening, it can't all be for nothing in the end, hopefully
@bgtlover Hopefully. The small web "movement" and other similar efforts give me hope, but who knows how things will go.
@chrisg this is far too real
@chrisg congratulations on reinventing chromeOS
[witchzard] 
@chrisg ehhh no it's 2015 and it's exactly why Google released Chrome books. Computers with basic OS just to run Chrome browser.
@chrisg Oh please no! Nightmare fuel :)
@aral It could suck, wouldn't it?
How hard can it be to name it something else than ext7?
@chrisg Hahaha, yep, umm, that’s definitely the scariest bit ;)
@chrisg This sounds awfully dystopian!
@chrisg Haha, reminds me a bit of https://www.destroyallsoftware.com/talks/the-birth-and-death-of-javascript (have you seen it?)
@tito Oldie but goldie. "Wat" is a classic from them too.
@chrisg @Lunaphied I see your dystopia and raise you a “ads have a 1% chance of popping up on every keystroke and only have a buy or remind me in 5 minutes option”
@chrisg won't going to happen, but too funny