• 2 Posts
  • 79 Comments
Joined 3 months ago
cake
Cake day: May 14th, 2026

help-circle







  • Actually, I’ve been toying around with doing just that.

    https://github.com/karpathy/nanochat

    Given the surplus of power from solar I have, my set up (discussed elsewhere), I can run the training process end to end over about 50 days. It could be fun to make something truly bespoke from scratch.

    Returning to my prior example - it was only overkill for this task (and only because the problem turned out to be relatively simple), but I wanted to answer against the claim that people were using LLMs and agents for just programming.

    The example I cited IS silly, but it happened - IRL - 3 days ago. I can cite several much more “serious” use cases if you’d like.

    On the broader topic of costs: none of us can really comprehend the embedded costs in the things we use everyday.

    However unlike (say) the recurring embedded costs of buying bottled water, a cup of coffee or streaming a movie, the costs of my set up are paid once and pay off multiple times.

    Ultimately, I get why there’s a FuckAI sentiment… but really we should expand that to “Fuck Corpo AI”. My feeling is that it’s not AI that people are against per se, it’s the corpos and what they’re doing with them. Too much of that gets flattened and tribal.

    The “true cost” argument only holds if you assume the only valid infrastructure is someone else’s. I don’t make that assumption, and I’d say that neither do others in the enthusiast community.



  • Using a sledgehammer to crack a wallnut can be satisfying :)

    Besides, the kids are always finding fun and inventive ways to break the TV, so I figured set it up once and then let the bot handle it.

    Twelfth Doctor: This is Clara, not my assistant. She’s a… Some other word.

    Clara: I’m his carer.

    Twelfth Doctor: Yeah, my carer. She cares so I don’t have to.






  • OK, here’s a real life use case, small time as it is, that has nothing to do with anything rarefied.

    I have a TCL Android TV that, after some “unapproved measures”, runs without a Google account - FLauncher as the front end, telemetry blocked, home media streaming, YouTube ad-free. The yar har har works.

    Two days ago the kids come in. TV’s not working.

    Jellyfin: fine. Nova Player: fine. Stremio, YouTube, anything hitting the net: dead.

    We’d had a brownout that day, so I figured the router was scrambled. Reset it, re-added the network… and Smarttube now launches, plays for 30 seconds, then hard-locks. Just weird JSON vomit and then dead.

    At that point I’m thinking the TV is fucked. Options:

    1. factory reset and reinstall the workarounds - 2-3hrs of pain-in-the-ass fuckery, but it wouldn’t be the first time.

    2. crawl the net decoding cryptic error codes.

    3. take it in for repair (good luck with that).

    4. fall down a TV diagnostics rabbit hole.

    Instead, I chose option 5)

    I told my local LLM: "This TV is on this IP. This laptop has ADB access. I’ve given you MCP tools with ADB libraries. Here’s the issue, here’s what I tried. Check prior history of what I did for this project, figure out what’s wrong now and fix it, or tell me it’s broken.

    If it looks like the brownout caused corruption, factory reset, stream-install the backup APKs directly, sign in my accounts, get it back up and running.

    Failing that, if it’s looking like a hardware issue, push a diagnostic tool, test the TV and tell me which parts I might need, as well as likely costs using eBay or Amazon. I’m happy to spend $50 on this shitbox and not one cent more. If the fix looks like it’s more than $50, find me a TV of this size under $200. Prefer local shop X or Y.

    Present everything as markdown in the usual folder, using house style. Ping me on my phone if you need me to authorise anything or when you’re done with the TL;DR."

    The agent ADBs in, pokes around, tests, checks the WiFi, and figures out the brownout had corrupted the date and time.

    Turns out this p.o.s doesn’t have a RTC back up battery.

    Wrong clock = certs flagged as invalid = no net.

    It syncs the date and time with my location. Boom, everything comes back.

    Now, could I have twigged it myself?

    Yes - eventually. But the agent already had the tools, knew the TV’s prior state, and had a record of what I’d already tried.

    It could’ve reset the TV, sideloaded all my apps, re-signed into my accounts - the whole lot - if needed.

    This is not the only time I’ve had real world use of my local, and it had very little to do with “writing functions”.

    I had a project - jailbreak my tcl. Over time, I’d stored enough context of what I had done in the LLM and its tools, so that it could do what I’d do (and more), faster, based on the constraints I had put in place.

    People that call these things “spicy autocorrect” are demonstrating either a skill issue or a lack of imagination.


  • Just in defence of your point:

    I prefer to self host, and am in the process of setting up my entire system to run off solar (tiny 2x250 watt solar array, 3 deep cycle 100Ah marine batteries and sundry equipment).

    End to end, my AI rig - a BC-250 (a cut down PS5, basically) consumes 50w standby / 200w on inference and can run Qwen 3.6-27b (Ternary Bonsai specifically, about par with Sonnet 4.6 on the AAII) at good speeds (~30 tok/s)

    With a smart power plug and Raspberry pi (uses about 5w) as the pilot fish, I can run my entire system “off grid”, to say nothing of the excess solar that accumulates (several days battery back up to run home router, pi server, hard drive array and 2-3hrs daily of inference).

    For $1500 AUD, “Poor man’s Sonnet 4.6 at home” ain’t nothing to sneeze at.

    Outside of folks such as myself, this blog post may be of interest -

    https://blog.andymasley.com/p/individual-ai-use-is-not-bad-for




  • Could be. No more or less than a very large library of books.

    Plus, you wouldn’t have to be 100% Robinson Crusoe-ing . Peer-to-peer networks, mesh radio and sneakernet would still exist in that hellscape.

    Still, for nerd fun, I paper-napkin’ed the maths last night with the aid of clankers. Here’s how I’d make an “internet in a box”.

    Reference: 1–2TB

    • Wikipedia, Gutenberg and other Kiwix archives: 300–600GB.
    • Medical, repair, farming and technical manuals: 100–300GB.
    • Forums, software docs and selected papers: 100–400GB.
    • OpenStreetMap, routing and regional map tiles: 100–500GB.
    • Selected Zimit web archives: 100–500GB.

    Zimit might fit roughly 100,000 ordinary web pages into 100GB. Kiwix server could serve the ZIM archives (inc websites), while Recoll could provide Google-like full-text search across the wider library of websites, books, manuals and documents.

    Other Software: 1–2TB

    • “Linux images”: 100–200GB.
    • Open-source apps, drivers and firmware: 200–400GB.
    • Compilers, package mirrors and useful GitHub repos: 300GB–1TB.
    • Emulators, open games and preserved software: 300GB–1TB.

    Media and education: 5–10TB

    • Podcasts and radio: 300GB–1TB.
    • Music: 1–2TB.
    • Courses, documentaries and educational video: 2–4TB.
    • Films, television and selected YouTube channels: 3–5TB (at 720p, that’s anything between 5-10yrs of content, if watched 2hrs/day).

    AI and search: 100–500GB

    • A few local language models: 50–150GB.
    • Speech, image, OCR and embedding models: 20–100GB.
    • Search indexes and vector data: 20–200GB.
    • OpenWebUI, Kiwix, Jellyfin and other services: under 20GB.

    Community and nerd nonsense: 100–500GB

    • A local forum, wiki, IRC or Lemmy instance.
    • Reticulum or Meshtastic links for nearby networks.
    • Minetest, Doom, Quake and local game servers.
    • Old Reddit or BBS archives
    • OASIS or another Reddit-style simulator populated by argumentative LLM agents (yeah that’s a thing apparently).
    • Random GitHub projects.

    (That last category has little survival value, but probably considerable morale value)

    So, roughly:

    • Lean: 6–10TB.
    • Comfortable: 12–18TB.
    • Media-heavy: 20TB or more.

    Twelve to eighteen terabytes would give you a decent slice of “the internet”. Get a 3D printer while you’re at it for the ultimate off grid chic.