{"slug": "an-interactive-map-of-ai", "title": "An Interactive Map of AI", "summary": "Artifipedia launched an interactive map visualizing 56 AI concepts and 150 connections between them, allowing users to drag nodes, trace links, and click for details. The map is embeddable and updates automatically as the encyclopedia grows.", "body_md": "Navigate\n\n# The map of AI.\n\nEvery concept on Artifipedia, and every link between them. Drag a node, hover to trace its connections, click to read.\n\n### Embed the map\n\nFree to use on any site. Paste this where you want it to appear — it stays current as the encyclopedia grows.\n\n56 concepts · 150 connections · this map is generated from the \"connects to\" links on every entry, so it grows as the encyclopedia does.\n\n## Every concept\n\n### Agent Memory\n\nGiving an AI a way to remember across conversations, since the model itself forgets everything the moment a session ends.\n\n### AGI (Artificial General Intelligence)\n\nA hypothetical system with broad human-level capability across domains — undefined enough that people can argue about whether it's arrived.\n\n### AI Agent\n\nSoftware that pursues a goal by taking its own steps — deciding, acting, and reacting — instead of answering once and stopping.\n\n### AI Alignment\n\nThe problem of making AI systems actually do what people intend — reliably pursuing the goals we want, not just the ones we accidentally specified.\n\n### Artificial Intelligence\n\nThe field of making machines do things that seem to require intelligence — a definition that has moved every time the machines succeed.\n\n### Attention\n\nThe mechanism that lets an AI decide which other words matter when interpreting each word — the core idea behind transformers.\n\n### Backpropagation\n\nThe algorithm that works out which weights caused a mistake and by how much — the reason neural networks can learn at all.\n\n### Bias & Fairness\n\nThe problem of AI systems producing unfair or discriminatory outcomes — usually by absorbing biases present in their training data.\n\n### Chain-of-Thought\n\nGetting a model to reason step by step before answering — which dramatically improves its performance on hard problems.\n\n### Clustering\n\nGrouping things that resemble each other — and the fact that the algorithm always returns groups, whether or not any exist.\n\n### CNN (Convolutional Neural Network)\n\nA network that slides small filters across an image to find local patterns — the architecture that made computer vision work.\n\n### Context Window\n\nThe maximum amount of text an AI can consider at once — its short-term working memory, measured in tokens.\n\n### Deep Learning\n\nMachine learning using neural networks with many layers — the approach behind nearly every recent AI breakthrough.\n\n### Diffusion Model\n\nHow most AI image tools work — starting from random noise and removing it step by step, guided by a prompt, until a picture appears.\n\n### Embeddings\n\nTurning words (or images, or anything) into lists of numbers, arranged so that similar meanings end up close together.\n\n### Explainability\n\nGetting a model to show its working — and the uncomfortable fact that most methods explain the explanation, not the decision.\n\n### Feature Engineering\n\nReshaping raw data into things a model can actually use — still where most of the accuracy comes from outside deep learning.\n\n### Fine-tuning\n\nContinuing a model's training on your own examples so its *behavior* changes — baked into the model, not supplied at answer time.\n\n### GAN (Generative Adversarial Network)\n\nTwo networks trained against each other — one faking, one detecting — until the fakes pass. The technique diffusion largely replaced.\n\n### GPU\n\nThe chip that made deep learning possible — thousands of small cores doing the same maths at once, which is exactly what neural networks need.\n\n### Gradient Descent\n\nWalking downhill on the error surface, one small step at a time — how a model's weights actually get updated.\n\n### Guardrails\n\nThe checks around a model that decide what it's allowed to receive, say, and do — the part that stops a demo becoming an incident.\n\n### Hallucination\n\nWhen an AI produces something fluent and confident that is simply false — fluency is not the same as accuracy.\n\n### Image Classification\n\nGetting an AI to look at an image and say what it is — the foundational task of computer vision.\n\n### Image Segmentation\n\nLabelling every pixel rather than drawing a box — what you need when the exact shape matters.\n\n### Inference API\n\nRenting a model by the request — how nearly everyone actually uses AI, and the dependency that comes with it.\n\n### Intelligence\n\nThe word underneath \"artificial intelligence\" — used constantly, defined by nobody, and the reason the field's biggest arguments never resolve.\n\n### Jailbreaking\n\nGetting a model to do what it was trained to refuse — and the structural reason it keeps working.\n\n### Large Language Model (LLM)\n\nAn AI trained on enormous amounts of text to predict the next piece of writing — the technology behind chatbots like ChatGPT and Claude.\n\n### Loss Function\n\nThe number that says how wrong the model is — and therefore the definition of what it's trying to become.\n\n### Machine Learning\n\nGetting computers to learn patterns from data and improve at a task, instead of being explicitly programmed with rules.\n\n### Multi-Agent Systems\n\nSeveral AI agents working together on one problem, each with a role — powerful in demos, awkward in production.\n\n### Multimodal AI\n\nModels that handle more than one kind of input — text and images, sometimes audio and video — in a single shared representation.\n\n### Neural Network\n\nA system of simple connected units that learns patterns from examples — the foundation underneath deep learning and modern AI.\n\n### Object Detection\n\nFinding *where* objects are in an image and *what* they are — drawing a labelled box around each one.\n\n### OCR (Optical Character Recognition)\n\nTurning pictures of text into text — solved for clean documents, still genuinely hard for everything else.\n\n### Open-Weight Models\n\nModels whose weights you can download and run yourself — often called open source, usually not quite.\n\n### Overfitting\n\nWhen a model memorizes its training data instead of learning the general pattern — so it looks great in training but fails on new data.\n\n### Privacy & PII\n\nPersonal data going into AI systems, coming back out of them, and the fact that a trained model is very hard to un-train.\n\n### Prompt Engineering\n\nThe craft of writing instructions that get the best, most reliable output from an AI model.\n\n### Quantization\n\nStoring a model's numbers with less precision so it fits in less memory and runs faster — usually at a surprisingly small cost in quality.\n\n### Red-teaming\n\nAttacking your own system on purpose, before someone else does it for free.\n\n### Reinforcement Learning\n\nLearning by trial and error through rewards — the way you'd train a pet, applied to software.\n\n### Retrieval-Augmented Generation (RAG)\n\nLetting an AI answer from a specific set of documents by looking them up as it responds — instead of relying only on what it memorized.\n\n### RLHF (Reinforcement Learning from Human Feedback)\n\nTraining a model on human preferences rather than correct answers — the step that turned text predictors into assistants.\n\n### Supervised Learning\n\nTeaching an AI by showing it labelled examples — inputs paired with the correct answers — so it can predict answers for new inputs.\n\n### System Prompt\n\nThe standing instructions a model gets before the conversation starts — influential, invisible to users, and not a security boundary.\n\n### Temperature\n\nA single setting that controls how random or predictable an AI's output is — low for focused, high for creative.\n\n### Token\n\nThe small piece of text an AI reads and writes — usually a chunk of a word, not a whole word.\n\n### Tool Use\n\nLetting a model call real software — a search, a calculator, your database — instead of trying to answer everything from memory.\n\n### Train/Test Split\n\nHolding back data the model never sees, so you can find out whether it learned anything or just memorised.\n\n### Training vs Inference\n\nBuilding the model versus using it — two completely different activities with different costs, hardware, and constraints.\n\n### Transfer Learning\n\nStarting from a model that already learned something general, instead of from random numbers — why small teams can build real AI.\n\n### Transformer\n\nThe neural-network architecture behind almost every modern AI model — built around attention, which lets it weigh every word against every other, all at once.\n\n### Unsupervised Learning\n\nFinding structure in data nobody labelled — useful, underrated, and much harder to know if you got right.\n\n### Vector Database\n\nA database built to store embeddings and find the most similar ones fast — the search engine behind meaning-based retrieval.", "url": "https://wpnews.pro/news/an-interactive-map-of-ai", "canonical_source": "https://artifipedia.com/map", "published_at": "2026-07-16 03:53:55+00:00", "updated_at": "2026-07-16 04:06:49.122217+00:00", "lang": "en", "topics": ["artificial-intelligence", "ai-tools", "ai-research"], "entities": ["Artifipedia"], "alternates": {"html": "https://wpnews.pro/news/an-interactive-map-of-ai", "markdown": "https://wpnews.pro/news/an-interactive-map-of-ai.md", "text": "https://wpnews.pro/news/an-interactive-map-of-ai.txt", "jsonld": "https://wpnews.pro/news/an-interactive-map-of-ai.jsonld"}}