Teach artificial intelligence online in English

People come to artificial intelligence from very different places. A computer science student wants to know why the loss on their neural network stopped falling after the third epoch. A developer wants to build on top of a language model without treating it as a black box. A doctor, a lawyer or a marketing manager wants to understand what these systems can really do, and where they fail. The ideas underneath are fewer than the noise around the field suggests: data, a model, a way to measure its errors and a way to make those errors smaller.

That is also why the subject is hard to learn alone. Libraries such as scikit-learn and PyTorch hide the linear algebra, probability and calculus underneath, so a student can train a model in ten lines of Python and still be lost the moment it misbehaves. A tutor who moves from the intuition to the equation to the code in a single lesson supplies the part the online tutorials skip.

Why Clasora is looking for artificial intelligence tutors who teach in English

Clasora shows students the tutors who teach in the language they use the site in. A student reading Clasora in English sees tutors who teach in English, so if English is your teaching language, those are the students who find you.

In AI, the words themselves cause a surprising share of the confusion. The field takes everyday English and gives it a narrow meaning: “loss”, “attention”, “regularisation”, “temperature”, “hallucination”. “Bias” alone can mean a learned parameter in a neuron, a systematic error in statistics or an unfair pattern in a model's decisions. A tutor who says plainly which one is meant, and points to the line of code where it lives, spares the student a lot of guesswork.

Clasora does not employ tutors, and this is an opening rather than a job. You work independently: you set your own prices and hours, and you teach as much or as little as you like. Tutoring can be a side income on a couple of evenings a week, or you can teach enough for it to stand in for a job.

What teaching artificial intelligence online on Clasora looks like

This opening is 100% remote. All lessons are online, and you teach from wherever you live.

Students book from your calendar and see only the times you have marked as free, in their own time zone. You decide how much notice you need and how far ahead they can book.

Lessons run in the browser, and nobody installs anything. Screen sharing does most of the teaching. You open a Jupyter notebook, change the learning rate, rerun the cell, and the student watches the loss curve change shape. You can also step through a training loop in VS Code, or draw a small network in a whiteboard app and follow the chain rule backwards, one weight at a time. Just as often the student shares their screen, and you read the stack trace together until the tensor with the wrong shape turns up. The in-call chat is handy for passing a line of code or an error message, and background blur keeps your room out of the picture.

Between lessons, each student has a conversation with you. They can send the notebook where the model refuses to learn, and you can reply with comments or a cleaner version of the code. Messages, files and shared links sit next to the sessions they belong to. Your own example notebooks can live in your personal file area, documents can come across from Google Drive, and Google Calendar can sync with your Clasora calendar.

What students usually want help with

Machine learning fundamentals: supervised and unsupervised learning, regression, classification and clustering, and which one a problem calls for.

The maths underneath: vectors and matrices, derivatives and the chain rule, probability, and why gradient descent finds its way downhill.

Neural networks: activation functions, loss functions, backpropagation, and what to do when training stalls or blows up.

Deep learning architectures: convolutional networks for images, recurrent networks for sequences, and transformers with their attention mechanism.

Language models: tokens and embeddings, predicting the next word, prompting, fine-tuning, and retrieval over a company's own documents.

The Python toolkit: NumPy, pandas, scikit-learn, and PyTorch or TensorFlow.

Working with data: cleaning it, choosing features, splitting it for training and testing, and catching data leakage before it produces a model too good to be true.

Evaluating a model: precision and recall, the confusion matrix, cross-validation, and choosing a metric that fits the actual problem.

Classical AI and reinforcement learning: search algorithms such as A*, minimax for games, and Q-learning worked through by hand.

Projects: planning a project for a course or a portfolio, debugging it, and reading the papers it builds on.

AI for non-specialists: professionals who want to know how these tools work and where they go wrong.

Who we are looking for

You know the areas of AI you offer well enough to explain why a method works as well as how to use it, and to find the bug in someone else's training loop.

You explain clearly in English, and you can put a technical idea into plain words for someone who has never seen the notation.

You can tell what current models genuinely do from what is claimed about them.

You might teach computer science already, hold a degree in computer science, mathematics or statistics, be a student with a firm grasp of machine learning, or use it every day at work.

You have a computer with a camera, a microphone and a steady internet connection.

You are 18 or older.

Teaching experience helps, but it is not required.

What you get

No commission. Clasora takes nothing from the price you set.

Free to join and to publish. There is no subscription and no listing fee.

Your own prices. You set a price for each session length, and you can offer a free first session.

Your own hours. You choose your weekly hours and days off, the notice you need, and how far ahead students can book.

A public listing. Your listing is a page of its own that search engines can find, shown in the language each visitor reads, with your photo, an intro video, what you teach, the languages you speak and your price for each session length.

Reviews that measure teaching. Students rate you on knowledge, clarity and engagement.

Checked qualifications. You can add your degrees and other qualifications to your profile, and a moderator checks them.

Lessons in the browser. Video with screen sharing, background blur and an in-call chat, with nothing to install.

How to start

Create an account with the button on this page.

Publish a listing for artificial intelligence, saying which areas you cover and who your lessons are for.

Open your calendar and mark the hours you are free.

It takes a few minutes. You accept the terms when you become a tutor, and a moderator checks your listing before it goes live.

Questions tutors ask

Do I need a doctorate or a job in AI research?

No. You need a solid command of the parts of AI you offer and the ability to explain them clearly. If you hold a degree or another qualification in the field, add it to your profile and a moderator will check it.

Can I teach people who have never written a line of code?

Yes. Your listing says what you teach and who it is for. You can offer an introduction to how AI works with no code at all, Python from the first line with machine learning as the goal, or advanced topics for experienced developers only.

Can I teach just one area, such as computer vision or language models?

Yes. You can offer the whole subject or only the parts you know best, such as deep learning or the maths behind it.

How do I get paid?

Card payments are not switched on yet. For now a booking is confirmed without a card, and you and the student settle payment directly between you. Clasora takes no commission on what you charge.