You now know enough to start

With variables, collections, control flow, functions, NumPy, and pandas, you have the actual toolkit working data scientists and AI engineers use every day. The remaining "AI" part is mostly calling libraries with the Python you already know. Let's connect the dots.

Calling an AI model is just a function call

Modern AI is astonishingly approachable from Python. Talking to a powerful model is a few lines:

from openai import OpenAI
client = OpenAI()
reply = client.chat.completions.create(
    model="gpt-4o-mini",
    messages=[{"role": "user", "content": "Explain embeddings in one sentence."}],
)
print(reply.choices[0].message.content)

Look closely and it's all things you know: importing a library, creating an object, calling a method, passing a list of dicts, and reading a value out of the result. No new language — just a new library.

Where each skill leads

• pandas / NumPy — loading and preparing data for any model.
• Functions and loops — building pipelines that process many items.
• Dicts — the format APIs speak (messages in, JSON out).
• The libraries ahead — scikit-learn for classic ML, PyTorch for deep learning, LangChain for LLM apps — all ordinary Python.

A sensible path forward

1. Get comfortable loading and exploring data with pandas.
2. Train a first model with scikit-learn (see the ML Foundations course).
3. Call an LLM API and build something tiny that uses its output.
4. Pick a direction — classic ML, deep learning, or LLM apps — and go deeper.

The leap from "learning Python" to "doing AI" is smaller than it looks. You've built the foundation; the AI libraries are just well-organized Python that sits on top of it.

Try this: Take one thing you learned here — say, looping over a list — and imagine the AI version: loop over a list of questions, send each to a model, collect the answers in a new list. That mental picture is a real program, and it's now within your reach.