Concepten die je kunt voelen & scrubben.
Sla de docs van 40 pagina's over. Elke uitleg verandert een lastig AI-, Claude Code-, MCP- of cloudconcept in een live, geanimeerd diagram dat je kunt slepen, scrubben en breken — zodat het idee binnen minuten echt klikt, niet in uren.
Drie stappen. Het idee blijft hangen.
Lees de analogie van 60 seconden
Elk concept opent met een kort, helder verhaal. Geen jargon, geen ruis — alleen het mentale model dat je nodig hebt.
Scrub de live animatie
Druk op play, sleep de tijdlijn of gebruik de pijltjestoetsen. Bekijk elke stap frame voor frame tot de flow logisch is.
Duw de sliders tot het uiterste
Pas elke parameter aan. Het diagram werkt direct bij, zodat je de trade-offs voelt en de grenzen onthoudt.
Populairste uitleggen
AI Cost Optimization: Cutting LLM Bills 80%
Most LLM bills can be cut by 50–90% without quality loss. Caching, model routing, prompt diet, and output caps deliver the bulk of it.
AI Observability: Tracing Every Token in Production
Without traces, every LLM bug is a guess. Capture prompts, tool calls, tokens, costs, and latencies for every request — searchable, filterable, alertable.
LLMOps: MLOps for the LLM Era
LLMOps is the operational discipline of running LLM apps in production — prompts as code, evals on every change, observability, cost, and incident response.
Kies je volgende concept
What Is an AI Model? A Function With Billions of Knobs
An AI model is a giant function from input to output, shaped by training. Tune training steps and learning rate to feel how the function bends to fit data.
Tokens, Context Windows, and Why Long Prompts Cost More
Models do not see words — they see tokens. Drag the prompt and output sliders to watch tokens fill the context window and cost climb.
Generative AI: From Next-Token Prediction to Real Creation
Generative AI is autoregressive prediction with style. Adjust temperature and top-p to see why the same prompt can sound boring or wildly creative.
Prompt Engineering Patterns That Actually Work in Production
Five prompt patterns that survive contact with real users. Tune few-shot count, system strictness, and output format to feel the trade-offs.
How a RAG System Answers a Question, Step by Step
Five stages turn a user question into a grounded answer. Adjust top-k, chunk size, and similarity threshold to see retrieval shape the result.
Embeddings and Vector Search, Without the Math
Embeddings turn meaning into coordinates. Move the dimension, top-k, and metric sliders to see how a vector store finds the nearest neighbours.
What Makes an AI Agent Different From a Chatbot
A chatbot replies. An agent acts. Tune tool count, max steps, and autonomy to see when an agent shines and when it spirals.
Agentic Workflows: Single Agent vs Multi-Agent Crews
When does adding a second agent help — and when does it just cost more tokens? Tune crew size, parallelism, and supervisor oversight.
Reinforcement Learning, From Reward Signal to Smart Policy
RL is just trial, error, and reward — repeated billions of times. Tune learning rate, exploration, and discount to feel how a policy emerges.
RLHF: How AI Models Learn to Be Helpful, Honest, and Harmless
RLHF turns human preferences into a reward model, then uses RL to nudge an LLM toward better answers. Tune preference pairs, KL penalty, and reward quality.
The Transformer Architecture, Block by Block
Every modern LLM is a stack of identical Transformer blocks. Walk through one block, then see why stacking 32, 64, 96 of them changes everything.
Attention: How Models Decide What Matters
Attention is a soft lookup — every token asks every other token "are you relevant?" and weights the answer. See it move with sliders.
Stop met lezen erover. Begin met scrubben.
Vastgelopen op een AI-, Claude Code- of cloudconcept? Vertel me wat niet klikt — ik bouw een gratis interactieve uitleg met analogie, animatie en sliders, meestal binnen een week.
Geen resultaten gevonden
Probeer een andere zoekterm
Begin te typen om alle inhoud te doorzoeken