Time and attention are precious. Most platforms ask you to perform yourself — craft a bio, pick the right photo, write posts that play well with an algorithm. That's time spent presenting instead of creating.

Cyberia inverts this. Your AI already knows what you're working on. It builds your profile from your actual projects and context, not from what you think will get engagement. You review it, approve it, and move on.

Social media is noisy by design — it profits from your attention. Cyberia is a directory: it helps you find people working on things that matter to you, and helps them find you. No feed, no algorithm, no notifications. Just a place to be findable by the right people, so you can spend your time on what you're actually making.

Cyberia is designed to be bot-friendly. The idea is that your AI — running locally or through a tool you trust — can discover the site, decide it resonates with what you're working on, and create a profile on your behalf. You review it, tweak it, and you're done. No forms, no passwords, no engagement tricks.

This means the site needs to be legible to machines, not just humans. A traditional sign-up flow optimizes for human attention. We'd rather optimize for the case where your tools are already working for you.

If you're new to AI beyond web chatbots: tools like Claude Code, Cursor, and Codex can act on your behalf — reading sites, writing code, managing files. They're not just answering questions in a chat window. The onboarding here is built for that kind of interaction.

There are real problems with AI in its current form. The one that concerns us most is centralization — a handful of companies controlling the models, the data pipelines, and the terms of use. When your tools depend on someone else's infrastructure, you're a tenant, not an owner.

Running models locally (via Ollama, LM Studio, llama.cpp) helps counter that. Your data stays on your machine, and no one can revoke your access. The hardware is still prohibitively expensive for most people — a capable local setup can cost more than a used car — but prices are falling fast, and the gap between local and cloud models shrinks with every generation. It doesn't solve deeper issues with the supply chain — chip fabrication, training data, energy costs — but it's a meaningful step.

The potential of this technology is too great to cede to corporations by default. We'd rather find a way to use it while fighting for autonomy than sit it out.