,

AI-2

$0.00

AI-2 is a Linux system for ordinary and older computers that gives them a local AI assistant, a language model that runs on the machine itself, on the CPU, with no account, no subscription and no data leaving the computer. Based on Artix Linux, free and open source (MIT).

Knowledge Packs. A Knowledge Pack is a set of documents the computer can search and answer from in seconds, with no internet, naming the document every answer came from. No AI writes those answers, so an old computer answers as well as a new one. Three come on the image (AI-2’s own help, everyday Linux tasks, and 196 countries), so a fresh install answers questions straight away. You can add more, or make one from your own PDFs and notes and share it, in the community catalog.

  • 64-bit PC, 2 GB RAM minimum (4 GB recommended). Very old CPUs without AVX work.
  • Lean by design: the desktop, a browser, a text editor and the AI engine.

What AI-2 does:
* Honest about your hardware: it detects the machine, assigns a capability tier, measures what it can really do (the AI Score) and recommends models that genuinely fit
* Knowledge Packs: sets of documents the computer searches and answers from, offline and with sources. Three ship on the image, AI-2’s own documentation, twenty everyday Linux tasks, and 196 countries with their capitals, currencies, languages and calling codes. Ask through Applications > AI-2 > Search Knowledge, or with ai-2 doc search. What comes back is the documents’ own text with the source named, so nothing in the answer is written by the AI, and it answers in seconds where a chat answer on these machines takes minutes
* Your own documents too: index your PDFs and notes on a fast computer, export them as one pack file, and install it on the old one, which then searches them offline in seconds
* A community catalog of packs: every Knowledge Pack is listed with its download at https://github.com/ProWoos-Devs/ai2-knowledge, and anyone can share one they made. AI-2’s own three packs are there with everyone else’s
* Local first: the model runs on your CPU; an online model is used only if you explicitly choose it
* Runs on old PCs: one AI engine package (llama.cpp) that picks the right build for your processor when it starts, including CPUs without AVX or SSE4.1; validated on a 2011 laptop with 4 GB of RAM
* Setup wizard: at the first login it tunes the system, installs the engine, downloads a first model and measures your AI Score
* Chat page: ai-2 chat starts the local AI and opens a chat page in the browser
* Speech to text: ai-2 transcribe turns a recording, a voice note or a video’s audio into a text file, on your machine
* One computer can answer for the house: ai-2 gopher serves the knowledge packs to any other machine on your network, read with the text browser AI-2 ships or any Gopher client, with nothing of AI-2 needed on the asking side
* Lean by design: only the desktop, a browser, a text editor and the AI engine ship; anything else is a pacman -S away
* In your language: like ProWoos, the installed system already speaks English, Spanish and German, and anyone can contribute their own languages (https://github.com/ProWoos-Devs/ai-2/blob/main/TRANSLATING.md)
* Rolling release: based on Artix Linux (runit); you update, you never reinstall

Install: download the ISO, verify it with the SHA-256 checksum, write it to a USB stick of 4 GB or more (sudo dd if=ai-2-x86_64.iso of=/dev/sdX bs=4M conv=fsync status=progress on Linux or macOS, Rufus or balenaEtcher on Windows) and boot from the stick. The live desktop lets you try AI-2 without touching your disks; the START-HERE guide on that desktop explains the installation.

Status: AI-2 is young. It has been tested on two old laptops and in virtual machines, so yours is probably the third. If something does not work, open an issue at https://github.com/ProWoos-Devs/ai-2/issues. Every machine reported is a machine AI-2 gets better on.

Made by Rafael Minuesa at ProWoos. AI-2 is MIT licensed. Artix Linux, llama.cpp and the models keep their own.

Reviews

There are no reviews yet.

Only logged in customers who have purchased this product may leave a review.

Shopping Cart