AI Fundamentals

What machine learning actually does, how models are built, and why "thinking machine" is the wrong metaphor

It's Pattern-Finding, Not Thinking

When people say "the AI understands you" or "the AI decided," it's easy to picture something like a mind at work. It helps to drop that picture. What's actually happening underneath a chatbot, a recommendation feed, or a photo-tagging app is statistics at enormous scale: a system that has looked at huge amounts of data and learned which patterns tend to go together.

Show a machine learning system millions of photos labeled "cat" and millions labeled "not cat," and it will gradually adjust itself until it can guess correctly on photos it has never seen. It hasn't learned what a cat is the way a child does — through touch, sound, and years of embodied experience. It has learned which pixel patterns statistically correlate with the label "cat." That distinction — statistical correlation versus lived understanding — is the single most useful thing to hold onto when reading any AI headline.

Key Insight

A large language model predicting the next word in a sentence and a spam filter predicting whether an email is junk are doing the same fundamental thing: finding patterns in data and using them to make a probabilistic guess. Scale and data are what make one feel like conversation and the other feel like plumbing.

Training vs. Inference: Two Very Different Phases

Every AI system you interact with went through two distinct stages, and confusing them is a common source of misunderstanding.

1

Data Collection

Engineers gather examples — text scraped from the web, labeled images, transcribed speech, gameplay logs. Quality and diversity here set a hard ceiling on everything that follows. A model can't learn a pattern it never saw examples of.

2

Training: Adjusting the Parameters

The model starts with random internal settings, called parameters or weights — a modern large language model can have hundreds of billions of them. During training, it repeatedly guesses an output, compares it to the correct answer, and nudges its parameters to be slightly less wrong next time. Do this billions of times across powerful GPU clusters, and the parameters slowly converge on values that produce useful outputs. Training a frontier model can take months and cost tens of millions of dollars in compute.

3

Evaluation

Before release, the model is tested on data it never saw during training, to check whether it actually generalized useful patterns or just memorized its training set. This is also where teams test for bias, safety failures, and factual reliability.

4

Deployment

The trained model — now a fixed, frozen set of parameters — gets packaged behind an app or an API so people can actually use it. This is the point where a research artifact becomes a product.

5

Inference: Actually Using It

Every time you type a prompt or upload a photo, you're triggering inference — the model applying its already-frozen parameters to your new input to produce an output. Inference is fast and comparatively cheap; it's training that's expensive. This is also why a deployed model doesn't "learn" from your conversation the way a person would — its parameters don't change just because you talked to it.

Why This Matters

Nearly every debate about AI — "why does it repeat the same biases," "why doesn't it know about recent events," "why is it so expensive to build but so cheap to use" — traces back to this training/inference split. A model's knowledge is frozen at whatever its training data contained, and any bias baked into that data gets baked into its parameters too.

Neural Networks, Without the Math

A neural network is loosely inspired by neurons in the brain, but the resemblance is more marketing than biology. Structurally, it's simpler: layers of small mathematical units, each one taking in numbers, weighing them, and passing a new number to the next layer.

Input Layer

Where raw data enters — pixel brightness values for an image, or numerical codes representing words for text. Nothing is "understood" yet, it's just numbers.

Analogy: Think of it as the network's eyes and ears — it converts the world into numbers it can work with.

Hidden Layers

Stacks of units in between, each combining and re-weighting signals from the previous layer. Early layers in an image model tend to detect simple things like edges and colors; deeper layers combine those into shapes, textures, and eventually whole objects.

Analogy: Like a factory assembly line — each station adds a bit more structure to the raw material passed down from the last one.

Output Layer

Produces the final answer — a probability that the image contains a cat, a predicted next word, a classification label. The "depth" in deep learning simply refers to having many hidden layers stacked between input and output.

Analogy: The finished product coming off the assembly line — a decision the rest of the system can act on.

Weights & Parameters

The numbers that determine how strongly each unit's input influences its output. Training is the process of adjusting these weights. There's nothing else "inside" the network to adjust — intelligence-seeming behavior emerges entirely from getting billions of these numbers right.

Scale check: GPT-2 (2019) had 1.5 billion parameters. Frontier 2026 models are estimated in the hundreds of billions to low trillions.

Narrow AI vs. General AI

Nearly everything called "AI" today, including chatbots, is narrow AI — extremely capable within a bounded domain, but without transferable general reasoning the way a human has it.

Narrow AI

What exists today, including every LLM

Trained for a task or a broad-but-bounded set of tasks — writing, coding, image generation, translation. A model can be extraordinarily fluent across many domains and still be "narrow" in the technical sense: it doesn't have goals, embodiment, or understanding outside what it was trained on.

Examples in 2026: GPT-5, Claude, Gemini 3, DALL-E, Sora — all narrow, however broad-feeling their outputs.

Agentic AI

The 2025-2026 frontier, still narrow — but with autonomy

Systems that can chain multiple steps together on their own — browsing the web, writing and running code, calling other tools — to complete a multi-step task with less human supervision. This is the industry's current growth edge, not a step toward general intelligence, but a step toward automating longer chains of narrow tasks.

Examples: Coding agents, computer-use agents, autonomous research assistants.

General AI (AGI)

Still hypothetical

A system that could learn and reason across arbitrary domains at human level, the way a person can go from cooking to accounting to consoling a friend without retraining. No current system does this, whatever the marketing language suggests, and there's genuine scientific disagreement about whether today's approaches lead there at all.

Status: Actively debated research goal, not an engineering roadmap with a fixed date.

LLMs and Generative AI

A Large Language Model (LLM) is a neural network trained on enormous amounts of text to predict the next chunk of text (called a "token") given everything before it. That's the entire mechanical task. What makes it feel like conversation is that, at large enough scale with enough training data, "predict a plausible next token" turns out to be enough to produce essay-writing, code, translation, and dialogue that reads as coherent and often useful.

Generative AI is the broader umbrella: any model that produces new content — text, images, audio, video, or code — rather than just classifying or predicting a single number.

Text

LLMs like GPT-5, Claude, and Gemini generate text token by token, drawing on patterns learned from books, code, articles, and conversation.

  • Built on the transformer architecture (see Pioneers below)
  • Now commonly paired with "reasoning" modes that generate intermediate steps before answering
  • Powers chat assistants, coding copilots, and summarization tools

Images

Diffusion models learn to reverse a process of adding random noise to an image, so they can start from pure noise and "denoise" their way to a picture matching a text prompt.

  • Tools: Midjourney, DALL-E, Stable Diffusion, Adobe Firefly
  • Trained on billions of image-caption pairs scraped from the web
  • Raises real questions about consent and compensation for source artists

Video

The newest and most compute-hungry frontier: models that generate coherent, moving video from a text prompt, extending image diffusion techniques across time.

  • Tools: Sora, Veo, Runway, Kling
  • Struggles most with physical consistency — objects that should stay put, lighting that should stay fixed
  • Improving fast enough that "is this real" is now a serious media-literacy question

Code

Code-generation models are trained on public repositories and treat programming languages the same way text models treat English — as sequences of tokens to predict.

  • Tools: GitHub Copilot, Claude Code, Cursor
  • Increasingly used as the "hands" of agentic AI systems
  • Still requires a human to verify correctness — fluent code isn't always correct code

A Word on Hallucination

Because LLMs generate text by predicting plausible next tokens rather than by looking facts up, they can produce confident, fluent, entirely false statements — fabricated citations, invented case law, wrong dates. This isn't a bug that gets fully patched; it's a direct consequence of how the technology works. Verifying important outputs stays the user's job.

Pioneers & Landmark Moments

The story of AI isn't one straight line — it's decades of quiet research punctuated by a handful of moments that suddenly changed what seemed possible.

1950

Alan Turing's "Computing Machinery and Intelligence"

Turing sidesteps the unanswerable question "can machines think?" and proposes a practical test instead: can a machine's conversation be mistaken for a human's? The Turing Test still frames public debate about AI capability today.

1958

The Perceptron

Frank Rosenblatt builds the Perceptron, the first trainable artificial neuron, at Cornell Aeronautical Laboratory. It's a single layer and can only learn simple patterns, but it introduces the core idea every modern neural network still uses: adjust weights based on error.

2012

AlexNet Wins ImageNet

A deep neural network built by Alex Krizhevsky, Ilya Sutskever, and Geoffrey Hinton crushes the ImageNet image-recognition competition by a stunning margin, trained on consumer GPUs. This is widely seen as the moment deep learning went from academic curiosity to the industry's default approach.

2017

"Attention Is All You Need"

Google researchers publish the transformer architecture, replacing older sequence models with one built entirely around "attention" — letting the model weigh the relevance of every other word when processing each word. Every major LLM since, GPT, Claude, Gemini and beyond, is a descendant of this paper.

2022

The ChatGPT Moment

OpenAI releases ChatGPT in November 2022. The underlying technology wasn't new, but a free, conversational interface made LLM capability tangible to hundreds of millions of ordinary people almost overnight — the fastest-growing consumer app in history at the time.

2024-2025

Reasoning Models Arrive

OpenAI's o1, and later open competitors like DeepSeek's R1, popularize models trained to generate an internal chain of reasoning steps before answering, substantially improving performance on math, logic, and coding tasks at the cost of slower, more expensive inference.

2025-2026

The Agentic Shift

The frontier moves from single-turn chat to agentic AI: systems that plan, browse, write and execute code, and use software tools across many steps with minimal supervision. Coding agents and computer-use agents become mainstream enterprise tools, and "can it act, not just answer" becomes the industry's defining benchmark.

Misconceptions Worth Retiring

MYTH: "It's just autocomplete, so it's not impressive"

REALITY: "Predict the next token" is the mechanism, not a measure of capability. The same mechanism, at sufficient scale, produces working code, coherent essays, and passable legal analysis. Simple mechanisms can produce complex, useful behavior.

MYTH: "A bigger model is always a better model"

REALITY: Since around 2023, gains have increasingly come from better training data, reasoning techniques, and post-training refinement rather than raw parameter count alone. Several smaller 2025-2026 models outperform older, larger ones on real tasks.

MYTH: "If it sounds confident, it's probably right"

REALITY: LLMs have no built-in mechanism to signal uncertainty by default — a hallucinated fact and a well-supported one can be delivered in an identical, confident tone. Confidence in phrasing is not evidence of accuracy.

MYTH: "The model remembers our past conversations because it's learning from us"

REALITY: A deployed model's parameters are frozen. Apparent "memory" across a conversation, or across sessions in some products, comes from separately engineered systems feeding past text back in as context — not from the model quietly retraining itself.

Continue Your Journey

This page is the front door. The AI section on this platform goes much deeper — from history and taxonomy, to ethics, to hands-on practice.

Menu