What really happens inside ChatGPT, Gemini and Claude when you press Enter? This lesson explains next-token prediction, tokens and token IDs, embeddings, artificial neurons, weights, biases and parameters, mixture of experts, attention and transformer layers, temperature and top-p, context windows, and why AI can hallucinate.
In this lesson you will learn:
• One Simple Job: Predict the Next Token
• Tokens: How AI Reads Text
• Embeddings: Meaning as Numbers
• Neurons, Weights and Biases
• Billions of Parameters
• Attention: Which Words Matter?
• Probabilities, Temperature and Top-p
• The Context Window
• Why AI Can Be Wrong
Chapters
0: 00 Introduction
0: 07 One Simple Job: Predict the Next Token
0: 30 Tokens: How AI Reads Text
0: 58 Embeddings: Meaning as Numbers
1: 16 Neurons, Weights and Biases
1: 48 Billions of Parameters
2: 17 Attention: Which Words Matter?
2: 46 Probabilities, Temperature and Top-p
3: 16 The Context Window
3: 44 Why AI Can Be Wrong
4: 07 Keep learning
📄 Presentation (PDF): https://www.MyWebUniversity.com/presentations/index.html
The full AI Explained series
▶ #1 What Is AI? From Alan Turing to ChatGPT and Claude:
★ #2 How LLMs Work Under the Hood: Tokens, Neural Networks, Parameters & Attention: (this video)
▶ #3 How AI Learns: Training, Loss, Gradient Descent, Fine-Tuning & RLHF:
▶ #4 Hardware for AI: CPU vs GPU vs TPU, VRAM, RAM & Running AI on Your PC:
▶ #5 The Python AI Toolkit: NumPy, pandas, scikit-learn, TensorFlow, PyTorch & LLM APIs:
▶ #6 ChatGPT vs Gemini vs Claude vs DeepSeek vs Llama vs Ollama: The Big LLM Comparison (2026):
▶ #7 Prompt Engineering: How to Talk to ChatGPT, Gemini & Claude and Get Great Answers:
▶ #8 Prompt AI vs Agentic AI: What’s the Difference? (Works for Any LLM):
▶ #9 Claude AI Explained: Opus, Sonnet, Haiku, Projects, Artifacts & How to Use Claude Well:
▶ #10 Claude Code Explained: The AI Coding Agent in Your Terminal (Install, CLAUDE.md, MCP, Subagents):
▶ #11 Build Your Own AI Agent in Python with the Claude API (Tool Use, Agent Loop, MCP):
▶ #12 Using AI Responsibly & Your AI Career Roadmap:
Free technical education since 2000 — 2,600+ videos and 50,000+ searchable PDFs in English and Dari:
🌐 https://www.MyWebUniversity.com
📺 English: @MyWebUniversity | Dari: @MyWebUniversity-Dari
Facts are current as of October 2026; AI products change quickly, so check each provider’s site for the latest versions.
Sources and further reading:
• Vaswani et al., “Attention Is All You Need,” 2017
• Mikolov et al., “Efficient Estimation of Word Representations” (word2vec), 2013
• Brown et al., “Language Models are Few-Shot Learners” (GPT-3), 2020
• Anthropic, “Tokens and context windows” — platform.claude.com/docs
• 3Blue1Brown, “Neural Networks” video series — 3blue1brown.com
All illustrations are original artwork made for this video. No company logos are used; product names belong to their owners.
Narration: AI voice. Script, illustrations and slides produced with the help of Claude (Anthropic).
#LLM #NeuralNetworks #Transformer #AI #MachineLearning #MyWebUniversity