EN 
Українська
Polski
Español
Deutsch
日本語
Italiano
Português (BR)
Français
العربية
Nederlands
中文
한국어
हिन्दी
Türkçe
Tiếng Việt
Русский
Bahasa Indonesia
فارسی
বাংলা
ไทย
Filipino
Kiswahili
اردو
Bahasa Melayu
தமிழ்
עברית
Ελληνικά
ਪੰਜਾਬੀ
አማርኛ
မြန်မာဘာသာ
ខ្មែរ
ქართული
नेपाली
Монгол хэл
Հայերեն
සිංහල
ລາວ
Čeština
Română
Magyar
Български
Svenska
Dansk
Suomi
Norsk
Slovenčina
Hrvatski
Српски
Slovenščina
Lietuvių
Latviešu
Eesti
Íslenska
Shqip
Македонски
Bosanski
Malti
Azərbaycanca
Қазақша
O'zbekcha
Тоҷикӣ
Кыргызча🧠 AGI Cognitive Network: From Narrow AI to Superintelligence
Activate six cognitive domains around a central intelligence core, wire them together, and watch a generality index climb from narrow AI toward AGI and superintelligence.
About the AGI Cognitive Network
Artificial General Intelligence (AGI) is usually described less as a single capability than as a threshold: the point at which a system stops being a specialist and starts being a generalist — able to understand, learn and apply knowledge across domains at or beyond human level, adapting to situations it was never explicitly trained for. Researchers commonly separate the trajectory into stages: narrow ("weak") AI that excels at one task, strong AI that reaches human-level competence broadly, and a hypothetical superintelligence that exceeds human capability across the board. What actually distinguishes these stages isn't raw processing power alone — it's breadth (how many distinct capabilities a system has) combined with integration (how well those capabilities work together on a single problem).
This simulation makes that distinction visible. Six cognitive domains — language, vision, reasoning, creativity, emotion and social interaction — sit around a central intelligence core. The "Intelligence level" slider activates domains one by one, the way capability expands from narrow to general; the "Connection density" slider wires activated domains together, standing in for the cross-domain integration that turns a bundle of specialists into something more general. A generality index G combines both factors, so a system with every domain switched on but nothing wired together scores low — breadth without integration isn't generality — matching how researchers actually argue about what AGI requires. The exact formula and control panel are covered in the "How it works" section of the control panel above the viewport.
Frequently Asked Questions
What is Artificial General Intelligence (AGI)?
AGI refers to a hypothetical AI system that can understand, learn and apply knowledge across many different domains at a level matching or exceeding human intelligence — as opposed to today's narrow AI systems, which are highly capable within one trained task (translation, image recognition, game-playing) but do not transfer that competence to unrelated problems without retraining.
What are the stages between narrow AI and superintelligence?
A common framing lists four stages: narrow ("weak") AI — specialised single-task systems, the current state of the art; strong AI — human-level competence generalised across domains, which is what "AGI" usually refers to; and superintelligence — a system that exceeds human capability across essentially all cognitively meaningful domains. None of the last two exist yet; they describe capability thresholds researchers debate the path toward, not shipped technology.
Why does the simulation combine breadth and connection density instead of using level alone?
Because activating many capabilities in isolation is not the same as generalizing. A system can have a language model, a vision model and a planning model bolted together and still fail at tasks that require them to inform each other in real time. The generality index G = (D/6) × (C/21) captures that: raising only the intelligence level (D) without also raising connection density (C) caps G well below 1, mirroring the argument that integration — not just capability count — is what "general" is supposed to mean.
What do the six cognitive domains represent?
Language (natural-language understanding and generation), vision (visual perception and scene understanding), reasoning (logical deduction, induction and planning), creativity (generating genuinely novel ideas or solutions), emotion (recognising and modelling affective states) and social interaction (communication and cooperation with other agents). These map to the capability list commonly cited in AGI discussions — no single one of them is sufficient on its own, which is the point of visualising all six together.
Why does a ring appear around the core only near maximum generality?
The faint ring is a deliberate, explicitly speculative visual cue, not a claim about real AI: it stands in for the open question of machine consciousness or self-modelling that researchers associate with hypothetical superintelligent systems. Its appearance threshold (G above roughly 0.75) is a simulation design choice to mark "beyond ordinary AGI," not a scientific claim that any measurable G value corresponds to actual machine self-awareness — that remains a contested philosophical and technical question with no agreed answer.
What are the main risks discussed around AGI development?
Commonly cited technical risks include uncontrolled capability growth outpacing safety research and misalignment — a system optimising for goals that diverge from human intent even while behaving competently. Commonly cited social risks include labor displacement, concentration of power among those who control advanced systems, and over-dependence on AI decision-making. These are active, unresolved areas of AI-safety research, not settled outcomes.
Is any part of this simulation a real trained AI model?
No. This is a conceptual, parameter-driven visualization — the domain activations, link density and generality index are deterministic functions of the two sliders, not the output of a trained neural network. It is designed to make the breadth-versus-integration argument about AGI intuitive, not to simulate an actual cognitive architecture.
Activate six cognitive domains — language, vision, reasoning, creativity, emotion, social — around a central intelligence core, wire them together, and watch a generality index climb from narrow AI toward AGI and superintelligence.
3D · Three.js / WebGL renderer · 60 FPS target · runs fully client-side, no install