alt Hacker NewsNot sure the dial-up analogy fits, instead I tend to think we are in the mainframe period of AI, with large centralised computing models that are so big and expensive to host, only a few corporations can afford to do so. We rent a computing timeshare from them (tokens = punch cards).
I look forward to the "personal computing" period, with small models distributed everywhere...
Replies
Funny you would pick this analogy. I feel like we’re back in the mainframe era. A lot of software can’t operate without an internet connection. Even if in practice they execute some of the code on your device, a lot of the data and the heavyweight processing is already happening on the server. Even basic services designed from the ground up to be distributed and local first - like email (“downloading”) - are used in this fashion - like gmail. Maps apps added offline support years after they launched and still cripple the search. Even git has GitHub sitting in the middle and most people don’t or can’t use git any other way. SaaS, Electron, …etc. have brought us back to the mainframe era.
Why would companies sell you the golden goose when they can instead sell you an egg every day?
We have a ton of good, small models. The issues are:
1. Most people don't have machines that can run even midsized local models well
2. The local models are nearly as good as the frontier models for a lot of use cases
3. There are technical hurdles to running local models that will block 99% of people. Even if the steps are: download LM Studio and download a model
Maybe local models will get so good that they cover 99% of normal user use cases and it'll be like using your phone/computer to edit a photo. But you'll still need something to make it automatic enough that regular people use it by default.
That said, anyone reading this is almost certainly technical enough to run a local model. I would highly recommend trying some. Very neat to know it's entirely run from your machine and seeing what it can do. LM Studio is the most brainless way to dip your toes in.
I like to think of it more broadly, and that we are currently in the era of the first automobile. [0]
LLMs are the internal combustion engine, and chatbot UIs are at the "horseless carriage" phase.
My personal theory is that even if models stopped making major advancements, we would find cheaper and more useful ways to use them. In the end, our current implementations will look like the automobile pictured below.
[0] https://group.mercedes-benz.com/company/tradition/company-hi...
I'm not a big google fan, but I really like the "Google AI Edge Gallery" android app [0]. In particular, I've been chatting with the "Gemma-3n-E2B-it" model when I don't have an internet connection, and it's really decent!
[0] https://play.google.com/store/apps/details?id=com.google.ai....
Dial-up + mainframe. Mainframe from POV as silos, dial-up internet as the speed we have now when looking back to 2025 in 2035.
Mainframes still exist, and they actually make a lot of sense from physics perspective. It's good idea to run transactions in a big machine rather than distributed, the latter is less energy efficient.
I think the misconception is that things cannot be overpriced for reasons other than inefficiency.
We are also in the mainframe period of computing, with large centralised cloud services.
I actually think we are much closer to the sneaker era of shoes, or the monorail era of public transit.
this -- chips are getting fast enough both arm n x86. unified memory architecture means we can get more ram on devices at faster throughput. we're already seeing local models - just that their capability is limited by ram.
I think we are in the dotcom boom era where investment is circular and the cash investments all depend on the idea that growth is infinite.
Just a bunch of billionaires jockeying for not being poor.
ollama and other peojects already make this possible
> "personal computing" period
The period when you couldn't use Linux as your main OS because your organization asked for .doc files?
No thanks.
I actually don’t look forward to this period. I have always been for open source software and distributism — until AI.
Because if there’s one thing worse than governments having nuclear weapons, it’s everyone having them.
It would be chaos. And with physical drones and robots coming, it woukd be even worse. Think “shitcoins and memecoins” but unlike those, you don’t just lose the money you put in and you can’t opt out. They’d affect everyone, and you can never escape the chaos ever again. They’d be posting around the whole Internet (including here, YouTube deepfakes, extortion, annoyance, constantly trying to rewrite history, get published, reputational destruction at scale etc etc), and constant armies of bots fighting. A dark forest.
And if AI can pay for its own propagation via decentralized hosting and inference, then the chance of a runaway advanced persistent threat compounds. It just takes a few bad apples, or even practical jokers, to unleash crazy stuff. And it will never be shut down, just build and build like some kind of kessler syndrome. And I’m talking about with just CURRENT AI agent and drone technology.
I mean, people can self-host plenty off of a 5090, heck even Macs with enough RAM can run larger models that I can't run on a 5090.
Imagine small models on a cheap chip that can be added to anything (alarm clock, electric toothbrush, car keys...)
> I look forward to the "personal computing" period, with small models distributed everywhere...
One could argue that this period was just a brief fluke. Personal computers really took off only in the 1990s, web 2.0 happened in the mid-2000s. Now, for the average person, 95%+ of screen time boils down to using the computer as a dumb terminal to access centralized services "in the cloud".