# Global Situational Awareness Dashboard ## Strategic Mission Radical changes in available intelligence are likely within 1–10 years. The window for distributing that capability — and the resilience to absorb it — is now. This project exists to: 1. **Increase available intelligence in a decentralized way.** Make advanced AI capability accessible to individuals and communities, not only to centralized institutions. This means tracking what is possible, building deployment tooling, and lowering barriers to local/edge AI. 2. **Improve resilience by rebalancing offense and defense.** As powerful AI tools proliferate, the security surface expands. Improve infosec for edge and community devices, improve privacy tooling, strengthen the defensive side so that decentralized capability is not also decentralized vulnerability. 3. **Improve general well-being.** Broadly available intelligence should improve material conditions, health, education, and autonomy. This is primarily a downstream effect of (1) and (4), not a separate workstream — but it is the reason the other goals matter. 4. **Decentralize power in general.** Reduce single points of failure and control. Favor architectures, tools, and policies that distribute agency rather than concentrate it. These are not advocacy positions to argue for. They are engineering goals to build toward. ## Instruments ### The Dashboard (situational awareness) A continuously updated, evidence-based operational picture of AI progress and the global conditions surrounding it. It tells you what is changing, how fast, and what the leverage points are. ### Tooling & Recipes (direct empowerment) Deployment guides, edge-compute recipes, security hardening playbooks, and privacy tool evaluations that let individuals and communities act on what the dashboard reveals. *(This is nascent — the dashboard came first, tooling will grow as priorities become clear from the data.)* ## The Problem AI capabilities are advancing on timelines that outpace institutional, regulatory, and public response. Critical developments — new frontier models, open-weights releases that decentralize advanced capabilities, breakthroughs in autonomous agents, and shifts in geopolitical posture — occur weekly or faster. No single source aggregates these signals into a coherent, actionable picture, and almost no one is building the connective tissue between "what is possible" and "what individuals can actually deploy." This project closes both gaps: awareness and action. ## What We Track The categories below reflect the situation as of early 2026. They are not fixed. The landscape that matters is defined by what is actually happening, not by a schema decided in advance. Categories will be added, split, merged, or retired as developments warrant. Some of what matters most in six months may not have a name yet. The organizing questions are: 1. **How capable are AI systems, and how fast is that changing?** 2. **Who has access to those capabilities, and what are they doing with them?** 3. **What can an individual or small community actually deploy today?** 4. **What are the emerging threats and defensive tools available?** ### Current tracking categories **AI Capability & Decentralization (Primary Focus)** - Frontier model releases — closed and open-weights, with capability assessments - Open-weights ecosystem — the capability gap between open and closed models as it narrows - Benchmarks & evals — METR task horizon, reasoning evals, coding evals, and other standardized measures over time - Agent frameworks — new releases, capability demonstrations, autonomy levels *(category emerged ~2025; may merge into general capability tracking as agents become the default interaction mode)* - AI-accelerated discovery — AI systems contributing to scientific or engineering breakthroughs - Recursive self-improvement signals — AI systems improving AI development (code generation, architecture search, data curation) **Compute & Hardware** - Hardware releases — from data center accelerators down to consumer systems (Mac Mini, etc.) that can run capable models locally. Performance/price trajectories across the full spectrum. - Training and inference cost trends — cost per FLOP, cost per token, efficiency gains - Local inference feasibility — the trend line for what models can run on what hardware at what quality, driven by both hardware improvements and model efficiency gains (quantization, distillation, architecture) **Security & Resilience** - Offensive/defensive balance — are new AI capabilities making attack or defense easier? Track significant shifts. - Privacy-enhancing technologies — adoption and maturity of tools for private inference, federated learning, encrypted computation - Edge/community device security — hardening guides, vulnerability trends, and tooling for non-enterprise environments - AI-enabled security tools — defensive applications of AI (anomaly detection, automated patching, threat analysis) accessible to individuals and small orgs **Geopolitical & Regulatory** - Active conflicts with relevance to AI development or deployment - Regulatory actions — legislation, executive orders, international agreements, export controls - Chip and technology trade restrictions - Centralization pressure — policy moves that concentrate or distribute access to AI capability **Ecosystem & Influence** - Funding and investment trends — capital flows into AI capabilities vs. safety research, and into centralized vs. decentralized efforts - Open-source and community ecosystem health — contributor trends, project viability, toolchain maturity ## How It Works - **Data collection**: Automated feeds run hourly on a self-hosted VPS, pulling from APIs, structured data sources, and curated feeds. Results are written as static JSON files. - **Presentation**: A plain HTML/JS dashboard reads those JSON files and renders interactive charts, timelines, and summary views. No build tools, no framework dependencies. - **Trend emphasis**: The default view prioritizes trends over time — capability curves, benchmark progressions, release cadences — so that the rate and direction of change are visible at a glance. - **Self-hosting by design**: The entire stack runs on a single VPS or a laptop. No cloud dependencies, no accounts required. This is deliberate — a tool for decentralization should not depend on centralized infrastructure. ## Who This Is For Technologists, builders, and communities who understand the stakes and want to act — not by lobbying or writing papers, but by deploying capability and building resilience. People who want to run models locally, secure their infrastructure, and reduce dependence on centralized providers. Also useful for researchers and policy-aware individuals who want an honest operational picture, even if their mode of action is different. ## What This Is Not - Not a news feed or aggregator - Not an advocacy platform — we build, not persuade - Not an introduction to AI - Not a hype tracker - Not dependent on any centralized service to function It is a sober, continuously updated operational picture paired with practical tooling — designed for people who want situational awareness and the means to act on it.