Endless UFrom Mythic Machines to Generative AI
A chronological history of the ideas, inventions, breakthroughs, and setbacks that produced modern artificial intelligence.
Imagining Artificial Minds
Explore humanity’s earliest stories and machines that anticipated artificial life and intelligence.
1.1 Ancient Myths of Artificial Beings
Examine artificial servants, animated statues, and created beings in Greek, Jewish, Chinese, and other traditions, while distinguishing mythology from technology.
1.2 Automata in Greece, China, and the Islamic World
Introduce documented mechanical automata and the engineering traditions that produced self-moving devices.
1.3 Medieval Mechanical Machines
Trace clocks, water-driven figures, and courtly automata through the medieval period.
1.4 What Early Automata Could—and Could Not—Do
Clarify the difference between programmed mechanical motion, apparent agency, and genuine adaptive intelligence.
Logic Becomes a System
Follow the development of formal reasoning, algorithms, probability, and dreams of mechanizing thought.
2.1 Aristotle and Formal Reasoning
Explain syllogisms and the attempt to represent valid reasoning through explicit formal rules.
2.2 Al-Khwarizmi and the Origin of Algorithms
Explore systematic calculation, algebra, and how Al-Khwarizmi’s name became associated with algorithms.
2.3 Ramon Llull’s Mechanical Logic
Examine Llull’s rotating logical figures as an early attempt to generate combinations of concepts mechanically.
2.4 Leibniz and Calculating Thought
Introduce Leibniz’s calculating machine, binary arithmetic, and ambition for disputes to be settled through formal calculation.
2.5 Probability, Statistics, and Bayes
Show how probability and Bayesian inference created mathematical tools for reasoning under uncertainty.
Every Endless U course is taught one-on-one. Ask questions, go deeper, or skip ahead, ChatGPT teaches at your pace, not a fixed script.
Open Endless U in ChatGPT →The First Programmable Machines
Connect industrial automation, symbolic logic, and programmable machines to the foundations of computing.
3.1 Industrial Automation
Explain how mechanized production transformed ideas about machines performing organized human tasks.
3.2 The Jacquard Loom and Punched Cards
Show how punched cards encoded instructions and influenced later programmable machines.
3.3 Charles Babbage’s Engines
Compare the Difference Engine and Analytical Engine and explain the latter’s programmable architecture.
3.4 Ada Lovelace’s Vision
Explore Lovelace’s program for the Analytical Engine and her insight that symbolic machines might manipulate more than numbers.
3.5 Boolean Algebra and Symbolic Logic
Explain how Boole and later logicians translated reasoning into operations suitable for machines.
Brains, Computation, and Information
See how mathematics, neuroscience, wartime computing, and information theory converged before AI was named.
4.1 Mechanical Calculation and Data Processing
Trace calculators, tabulators, punched-card systems, and their role in making information machine-processable.
4.2 Gödel and the Limits of Formal Systems
Introduce incompleteness and what it revealed about the powers and limits of formal reasoning.
4.3 Alan Turing and Universal Computation
Explain the Turing machine, computability, and the concept of a general-purpose programmable computer.
4.4 The Artificial Neuron
Examine McCulloch and Pitts’ mathematical model connecting neural activity with logical computation.
4.5 Cybernetics, Feedback, and Information
Introduce Wiener, Shannon, feedback control, and information theory as foundations for intelligent systems.
4.6 The Stored-Program Computer
Explain how electronic stored-program computers made increasingly sophisticated AI experiments practical.
Sitting down with a real tutor has always been the best way to learn, it's just been expensive and hard to access until now. Endless U gives you that experience for anything, for free.
Open Endless U in ChatGPT →Artificial Intelligence Becomes a Field
Study the moment machine intelligence became a named research program with competing approaches.
5.1 Can Machines Think?
Examine Turing’s 1950 paper, the imitation game, and his responses to major objections.
5.2 Early Games and Learning Programs
Explore early chess, checkers, and learning systems as experimental settings for machine intelligence.
5.3 The Dartmouth Workshop
Explain the 1956 proposal, participants, expectations, and why it became AI’s conventional founding event.
5.4 Symbolic AI
Introduce symbols, search, logic, and the physical-symbol-system hypothesis.
5.5 Perceptrons and Early Neural Networks
Explain the perceptron, its learning rule, initial achievements, and ambitious public claims.
Early Optimism and the First AI Winter
Follow AI’s first celebrated programs and the technical limitations that brought optimism crashing down.
6.1 Logic Theorist and General Problem Solver
Show how Newell, Simon, and Shaw used symbolic search to model problem solving.
6.2 ELIZA and Simulated Conversation
Explore ELIZA’s pattern matching, the ELIZA effect, and why people attributed understanding to it.
6.3 Shakey the Robot
Explain how Shakey combined perception, planning, and action in a simplified physical environment.
6.4 Machine Translation’s Early Failure
Trace early translation optimism, the ALPAC report, and the difficulty of language understanding.
6.5 The Perceptron Debate
Explain the important limitations identified in early perceptrons and the broader consequences for neural-network research.
6.6 The First AI Winter
Connect unmet promises, computational limits, critical reports, and funding reductions during the 1970s.
This isn't a fixed curriculum. Ask for a course on literally anything, and Endless U drafts the outline with you before saving it.
Open Endless U in ChatGPT →Expert Systems and the Knowledge Revolution
Examine AI’s revival through specialist knowledge and its second collapse under the weight of complexity.
7.1 Knowledge as Rules
Explain knowledge bases, inference engines, and the idea of reproducing expertise with if-then rules.
7.2 DENDRAL and MYCIN
Study influential scientific and medical expert systems, including their successes and practical limitations.
7.3 Commercial Expert Systems
Trace expert systems from laboratories into corporate use and the rise of the knowledge-engineering industry.
7.4 Japan’s Fifth Generation Project
Explore Japan’s ambitious computing initiative and the international response it provoked.
7.5 The Second AI Winter
Explain how maintenance costs, brittle systems, specialized hardware failures, and disappointed expectations caused another downturn.
7.6 Why Hand-Built Knowledge Did Not Scale
Examine the knowledge-acquisition bottleneck and the difficulty of encoding common sense.
Machines Learn from Data
Trace the shift from manually written rules toward statistical and learning-based systems.
8.1 The Return of Neural Networks
Explain why connectionist approaches regained attention and how they differed from symbolic AI.
8.2 Backpropagation
Teach the intuition behind gradient-based learning and the role of backpropagation in training multilayer networks.
8.3 Bayesian Networks and Probabilistic AI
Show how graphical models represented uncertainty and causal or conditional relationships.
8.4 Decision Trees, Support Vectors, and Ensembles
Survey influential classical machine-learning methods and the kinds of problems they solved.
8.5 Reinforcement Learning
Introduce agents, rewards, value functions, and learning through interaction.
8.6 Deep Blue Defeats Kasparov
Examine IBM Deep Blue’s 1997 victory and distinguish specialized search from general intelligence.
Close the chat, come back next week, whenever. Endless U remembers exactly where you left off, down to the lesson.
Open Endless U in ChatGPT →The Internet and Big-Data Era
Understand how digital data, web-scale services, GPUs, and benchmarks prepared the deep-learning revolution.
9.1 Search Engines as Intelligent Systems
Explain ranking, indexing, relevance, and how the web created an enormous applied machine-intelligence laboratory.
9.2 Statistical Machine Translation
Trace the replacement of hand-written language rules with models learned from large bilingual corpora.
9.3 Recommendation Algorithms
Explore collaborative filtering, personalization, and the growing influence of predictive systems on culture and commerce.
9.4 Speech Recognition Improves
Show how statistical methods, more data, and computing power made speech systems increasingly practical.
9.5 Data, GPUs, and Benchmark Culture
Explain why hardware, datasets, shared benchmarks, and competitions accelerated measurable progress.
9.6 ImageNet Changes Computer Vision
Introduce ImageNet and the large-scale classification challenge that set the stage for a major breakthrough.
Deep Learning Breaks Through
Follow the breakthroughs that moved neural networks from a specialized approach to the center of AI.
10.1 AlexNet and the 2012 Turning Point
Explain why AlexNet’s ImageNet victory mattered and how GPUs, data, and deep networks combined.
10.2 Convolutional Neural Networks
Teach how convolutional architectures learn visual features and transformed computer vision.
10.3 Word Embeddings and Machine Meaning
Explain how systems learned useful geometric representations of words from patterns in text.
10.4 Recurrent Networks and Sequences
Introduce recurrent networks, LSTMs, and their use in language, speech, and time-series tasks.
10.5 AlphaGo
Explore how deep networks, tree search, and reinforcement learning produced AlphaGo’s landmark victory.
10.6 The Limits of Specialized AI
Distinguish impressive task-specific competence from general, transferable intelligence.
Every lesson you finish earns points, and learning on consecutive days builds a streak, up to 2x your points at a 21-day streak.
Open Endless U in ChatGPT →Transformers and Foundation Models
Explain the architecture and training strategies that led directly to modern language and generative models.
11.1 Attention and the Transformer
Explain the attention mechanism and why the transformer enabled efficient learning from long sequences.
11.2 Pretraining and Transfer Learning
Show how models learn broad capabilities from large datasets and are then adapted to downstream tasks.
11.3 BERT and Contextual Language
Explain bidirectional pretraining and the leap in language-understanding benchmarks associated with BERT.
11.4 The GPT Lineage
Trace autoregressive transformer models from early GPT systems through increasingly capable large language models.
11.5 Scaling Laws and Emergent Capabilities
Explore observed relationships among compute, data, model size, performance, and debated claims of emergence.
11.6 Diffusion Models and Generative Images
Explain the basic denoising process behind modern image generation and its rapid adoption.
The Generative-AI Explosion
Study how generative systems became mass-market products and raised urgent practical and ethical questions.
12.1 ChatGPT and Conversational AI
Explain why conversational access to a capable language model changed public adoption and expectations.
12.2 Instruction Tuning and Human Feedback
Introduce instruction datasets, preference training, and reinforcement learning from human feedback.
12.3 Multimodal Models
Explore models that process and generate combinations of text, images, audio, and video.
12.4 AI Coding Systems
Trace code models from completion tools to systems capable of planning and executing larger software tasks.
12.5 Open and Closed Models
Compare proprietary systems, open-weight models, access models, and debates over innovation and safety.
12.6 Hallucinations, Bias, Copyright, and Safety
Examine major weaknesses, legal disputes, alignment problems, and attempts to make generative systems more reliable.
Proud of a course? Make it public and anyone can follow along or take it themselves, like you're doing with this one.
Open Endless U in ChatGPT →Agents, Reasoning, and AI Today
Survey the current frontier and the unresolved choices shaping AI’s next phase.
13.1 Tools, Retrieval, Memory, and Agents
Explain how models use external information, software tools, memory systems, and multi-step action loops.
13.2 Reasoning-Focused Models
Explore newer training and inference methods intended to improve multi-step reasoning, planning, and verification.
13.3 Robotics and Embodied Intelligence
Examine the effort to connect foundation models with perception and action in the physical world.
13.4 AI in Science, Medicine, Education, and Work
Survey real applications, measured benefits, limitations, and changing roles across major institutions.
13.5 Regulation, Geopolitics, and Energy
Discuss regulation, chip supply chains, national competition, data centers, and environmental costs.
13.6 AGI Arguments and Possible Futures
Compare major definitions and forecasts of general intelligence without treating uncertain predictions as settled facts.
New to Endless U?
Endless U turns anything you're curious about into a guided course, taught by ChatGPT, one focused lesson at a time. It's free, and takes seconds to start.
Open Endless U in ChatGPT