REVIEWED
Courses and sources
Courses, tools and sources the editors have looked at themselves. Each entry states the language, the prior knowledge needed and the price.
Course & Article
Adam
The optimiser practically every network today is trained with. Short and readable.
Short and readable; the optimiser practically every network today is trained with.
AI Act Explorer
A searchable, cross-linked edition of the EU AI Act. Jumps from an article straight to its recitals.
Far more pleasant than the PDF, but not the official version. When in doubt, use the full text.
AI Austria, network and events
An Austrian research-and-industry network with a regular events programme.
For the Austrian market, the obvious place to start when you are looking for people rather than documents.
AI for Everyone
Andrew Ng's non-technical introductory course. Around six hours, well suited to managers.
For leaders without a technical background who have to decide rather than build.
AI Index Report
An annual report with sourced figures on models, cost, adoption and regulation. Useful when a statement needs a source rather than an impression.
When a statement needs a source rather than an impression.
An Image is Worth 16x16 Words
Images as a sequence of patches, processed like text. The paper that brought transformers into vision.
For anyone processing image and text in one model.
Attention Is All You Need
The 2017 paper that introduced the transformer. Everything called a language model today rests on these eight pages.
The eight pages everything called a language model today rests on.
Austrian Bar
The representative body publishes what professional conduct rules require. For AI in client work that is the authoritative source, not the tool vendor.
For Austrian law firms: authoritative on professional conduct, not the tool vendor.
Austrian Data Protection Authority
The competent supervisory authority for Austria, with forms, decisions and guidance on reporting breaches.
For controllers in Austria: the competent authority for notifications and enquiries.
Batch Normalization
Why normalising intermediate values is what makes deep training stable in the first place.
For anyone asking why normalisation is what makes deep training possible.
BERT
Pre-training on masked text, then fine-tuning per task. The pattern that shaped language processing before the large models.
For understanding what shaped language processing before the large models.
BSI on the security of AI systems
Technical guidance on operating AI systems securely. Free, and unusually concrete.
For security and operations in German-speaking countries; unusually concrete for an official source.
Chain-of-Thought Prompting
Intermediate steps in the prompt markedly improve multi-step tasks. The basis of today's models with reasoning steps.
For anyone writing prompts: the paper behind the intermediate-steps pattern.
CNIL on artificial intelligence
The French supervisory authority publishes the most practical guidance on AI and data protection in Europe, in English as well as French.
For data protection leads who need a supervisory reading rather than the bare text of the law.
Common Crawl
The open crawl of the web that a large share of language model training data comes from. It shows concretely what a pre-training corpus actually contains.
For anyone asking where a model's knowledge comes from, and for the question of opt-out reservations.
Datasheets for Datasets
Documenting the provenance, composition and limitations of a dataset. The template today's documentation duties come from.
For anyone documenting datasets; the template behind today's duties.
Deep Learning
The standard work by Goodfellow, Bengio and Courville, free to read. Mathematically dense, complete, and in its foundational parts timeless.
For the systematic route. Strong as a reference, too dense as a first read.
Deep Learning Specialization
Five courses from the basics of neural networks to sequence models. Thorough, with programming exercises, and in places older than current practice.
For anyone who can program and wants to work through the field completely.
Deep Residual Learning
The shortcut across layers that made hundred-layer networks trainable. Present in every architecture today.
For understanding how networks were able to get deep at all.
Denoising Diffusion Probabilistic Models
Removing noise step by step instead of generating an image in one pass. The basis of every current image model.
For understanding every current image model, thought through from noise.
Distilling the Knowledge in a Neural Network
A large model teaches a small one what it knows. The basis of the small models running in production today.
For anyone deploying small models in production who wants the basis for it.
Dive into Deep Learning
A textbook with runnable code beside every derivation. Each chapter opens as a notebook you can recompute yourself.
For anyone who wants to compute along while reading; every chapter opens as a notebook.
Dropout
Randomly switching off neurons as regularisation. Simple, effective and still in use.
Simple, effective and still in use; readable without deep background.
Efficient Estimation of Word Representations
The paper that first established words as vectors with computable meaning. The origin of all embeddings.
The origin of all embeddings; short and still illuminating.
Elements of AI
A free introductory course from the University of Helsinki, no prior knowledge required. Around 30 hours, with a certificate.
If you want somewhere to go after our beginner path, this is the obvious next step. It overlaps our foundations in places but goes considerably deeper into how things work.
ENISA publications
Reports from the EU cybersecurity agency, including on securing AI systems and on the threat landscape. Free and in a European frame.
For security leads who need a European frame of reference rather than an American one.
Essence of Linear Algebra
Fifteen short films showing what a matrix does to space. If you have only ever seen vectors as lists of numbers, you will see something else afterwards.
For anyone who has only ever seen vectors as lists of numbers and needs an intuition.
EU AI Act, the official full text
Regulation (EU) 2024/1689 in full, in every official language. The primary source for any legal question.
Our governance articles summarise and contextualise. Where the exact wording matters, this text governs.
FlashAttention
Computing attention without ever materialising the quadratic matrix. The precondition for long contexts.
For anyone running long contexts who needs to know what memory hangs on.
Fraunhofer IAIS on artificial intelligence
German-language guidance on auditing, certifying and operating AI systems, from applied research.
For German-language audit and certification questions where English sources do not help.
Gender Shades
The study showing how far face recognition misses depending on skin tone and gender. The trigger for today's rules.
For any discussion of face recognition: the study that triggered today's rules.
Generative Adversarial Networks
Two networks learning against each other. The first approach that produced convincing images, now largely superseded.
Historically important; largely superseded today but useful for understanding how things developed.
Guidelines of the European Data Protection Board
The GDPR as interpreted by the body of supervisory authorities. In a dispute about a legal basis, the most solid source after the text of the law itself.
For legal teams and data protection officers when an interpretation has to hold up.
High-Resolution Image Synthesis
Diffusion in a smaller latent space rather than in pixels. The step that made image generation possible on ordinary hardware.
For anyone running image generation themselves; explains why it works on ordinary hardware.
Hugging Face agents course
Tool calls, planning, and the safeguards without which an agent is not viable in operation. Hands-on, with runnable code.
For anyone building an agent who needs to know which safeguards belong with it.
Hugging Face audio course
From the signal through spectrograms to recognition and synthesis, with code throughout. The practical companion to the speech and audio articles.
For anyone building speech technology themselves; assumes Python and some signal knowledge.
Hugging Face computer vision course
From image preprocessing through convolutional networks to vision transformers, with runnable examples for detection and segmentation.
For development with image data; assumes Python and delivers runnable examples in exchange.
Hugging Face datasets
Open datasets with description, licence and preview. Useful for evaluation sets, risky as training data without checking provenance.
Good for evaluation sets; as training data only with a provenance and licence check.
Hugging Face deep reinforcement learning course
Reward, policy and exploration in playable environments. Useful for understanding what actually happens when a language model is aligned.
For anyone wanting to understand what actually happens when a language model is aligned.
Hugging Face LLM Course
A free technical course on language models. Assumes Python knowledge.
For the step from using models to running open ones yourself.
Hugging Face model hub
Hundreds of thousands of open models with licence, model card and weights. The first place to look when checking whether a local model is enough for a task.
The licence is on the model card, and not every open model allows commercial use.
Hugging Face NLP course
Tokenisation, transformers, fine-tuning and deployment, with code throughout. Assumes Python, and in exchange you end up working with real models.
For development with language models once it has to go beyond calling an interface.
ImageNet
The dataset image processing measured itself against for a decade. Historically important and well documented in its biases.
For anyone reading benchmarks: almost every image recognition figure refers back to it.
Interpretable Machine Learning
What explainability methods deliver and where they get over-interpreted. The most sober treatment of the topic, freely available.
For anyone who has to promise explainability and should know what the methods actually deliver.
Jupyter
Notebooks where text, code and result sit side by side. The usual tool for recording a calculation so that others can follow it.
For your first calculations; keeps text, code and result traceable in one place.
Kaggle datasets
Real, untidy data to practise on. That is exactly what makes them valuable, because tidied examples hide the actual work.
For practising on untidy data, because tidied examples hide the actual work.
Kaggle Learn
Short units with runnable notebooks, from Python through pandas to a first model. No installation, straight in the browser.
For the very first steps with no installation, straight in the browser.
LangChain documentation
Building blocks for retrieval, tool calls and agents. Useful as a catalogue of the patterns, even if you end up building without the framework.
Useful as a catalogue of patterns, even if you end up building without the framework.
Language Models are Few-Shot Learners
The paper showing that examples in the prompt can replace fine-tuning. The origin of what is now called prompting.
For understanding why examples in the prompt can replace fine-tuning.
Learning Transferable Visual Models
Images and text in one shared space. The basis of searching images by description and of image generation.
For understanding image search by description and image generation.
llama.cpp
Running quantised models on ordinary hardware, down to single cards and small boards. The reference implementation for operating without a data centre.
For running without a data centre, down to single cards and small boards.
LoRA
Adaptation through a few additional parameters instead of full training. The reason fine-tuning is affordable today.
For anyone fine-tuning on a budget; the basis of affordable adaptation.
Machine Learning Specialization
Andrew Ng's course in its reworked form. Regression, classification, neural networks and the mistakes that actually happen in practice. The maths is included.
Free to audit; the certificate costs. If you only want to understand it, you do not need one.
Mathematics for Machine Learning
Exactly the mathematics machine learning needs and none of the rest. Linear algebra, calculus and probability in one volume, free as a PDF.
For anyone who wants exactly the mathematics machine learning needs and no more.
MIT 18.06 Linear Algebra
Gilbert Strang's lecture course, complete on video with problem sets. If you want to understand vectors, matrices and projections once and properly, this is the reference.
For anyone who wants to understand linear algebra properly once rather than look it up.
MIT 18.065 Matrix Methods
Singular value decomposition, principal components and optimisation applied to data. The bridge between linear algebra and what models actually compute.
For the step from pure mathematics to what models actually compute.
MIT 6.036 Introduction to Machine Learning
More formal than most online courses, with derivations rather than recipes. A good follow-up to a hands-on course when the question of why is still open.
For anyone left with the question of why after a hands-on course.
Model Cards for Model Reporting
The proposal to document purpose, limits and tested groups for every model. Today effectively a precondition for any audit.
For anyone preparing an audit; model cards have effectively become a precondition.
Model evaluation in scikit-learn
The complete overview of scoring metrics with their pitfalls. The shortest answer to why accuracy is usually the wrong number.
The shortest answer to why accuracy is usually the wrong number.
Netron
Opens a model file and draws its structure. The fastest way to see what a delivered model actually contains.
For a quick look inside a delivered model before putting it into operation.
Neural networks, explained visually
From a single weight through gradient descent to the attention mechanism. The best available intuition for what the formulas describe.
Watch it before your first textbook, not after. It saves weeks of confusion.
NIST AI Risk Management Framework
A structured frame for your own risk assessment, independent of the AI Act. Useful as an outline when none exists internally yet.
For building your own risk assessment, independent of the AI Act.
NIST AI RMF Playbook
The practical build-out of the risk framework: concrete suggestions per function on what to do and what to document.
For anyone building an internal risk assessment who wants an outline that has already been thought through.
NumPy
The foundation of all array computation in Python. If you want to recompute linear algebra yourself, you need nothing else.
For anyone who wants to recompute the formulas from the engineering articles.
OECD AI Principles
The international frame of reference that the European definition also draws on.
For placing national rules in an international frame.
Ollama
Run open models locally, one command per model. The simplest way to try running without a provider at all.
For a first try with a local model, with no provider and no account.
OpenCV tutorials
The official guides to filters, edges, features and calibration, each with runnable Python code.
For anyone writing image processing themselves; every example runs.
opencv-python
The package that brings OpenCV into a Python environment. One command, and classical image processing is available.
One command, and classical image processing is available in your own environment.
OWASP Top 10 for LLM applications
The list of weaknesses that actually occur in systems built on language models, from prompt injection to insecure tool integration.
For anyone building. The list replaces no review, but it is the best starting point for one.
Practical Deep Learning for Coders
Starts with a working model in the first hour and supplies the theory afterwards. The shortest route from basic Python to a model you trained yourself.
For impatient readers who know Python: your first model runs within the first hour.
Probabilistic Machine Learning
Kevin Murphy's volumes, building machine learning consistently out of probability theory. Extensive and freely available.
For anyone wanting machine learning built consistently out of probability theory.
PyTorch documentation
The reference for autograd, dtypes, memory behaviour and determinism. The place where questions about reproducibility actually get settled.
The place where questions about determinism and memory behaviour actually get settled.
PyTorch tutorials
The official guides, from a first tensor operation to distributed training. Short, runnable and continuously updated.
For getting into the library most research is written in.
Retrieval-Augmented Generation
The paper that joined retrieval and generation. The origin of the pattern that makes your own documents usable with citations.
For anyone making their own documents usable; the origin of the pattern.
Robust Speech Recognition
Speech recognition that copes with noise, accents and language switching. The benchmark dictation solutions are measured against.
For anyone introducing dictation or transcription who needs a benchmark.
scikit-learn user guide
Not a manual but a textbook with code. Every method comes with a note on when it does not fit, which textbooks rarely state so plainly.
For anyone using classical methods; each one comes with a note on when it does not fit.
Segment Anything
Segmentation without task-specific training, steered by points and boxes. Changes the preparatory work in image analysis considerably.
For image analysis with little data of your own; it changes the preparatory work noticeably.
Short courses on AI tooling
Units of around an hour on prompting, retrieval over your own documents, agents and evaluation. Free, and close to what is actually being deployed.
For practitioners with a specific question; a unit takes about an hour.
spaCy
A library for classical language processing. For recognising names, parts of speech and structure it is often faster, cheaper and more checkable than a language model.
When names, parts of speech or structure are needed: often faster, cheaper and more checkable than a language model.
Speech and Language Processing
Jurafsky and Martin, the standard work on language processing, free chapter by chapter. Covers classical methods and language models in one arc.
For anyone learning language processing systematically, classical and modern in one arc.
Switch Transformers
Only a fraction of the parameters compute per token. The paper that brought mixture of experts into wide use.
For understanding why large models do not compute every parameter per token.
Teachable Machine
Train an image classifier in the browser, without code, in ten minutes. Demonstrates overfitting faster than any explanation.
For a first look without code; it demonstrates overfitting in ten minutes.
TensorFlow Playground
A neural network in the browser with sliders for layers, activation and learning rate. You see in seconds what each knob does.
For anyone who wants to see what learning rate, layers and activation actually do.
The Commission's regulatory framework
The official overview of the AI Act with timeline, guidelines and pointers to implementing acts. The starting point for any question about deadlines.
The starting point for any deadline question, because the official timeline sits here.
The Elements of Statistical Learning
The statistical view of machine learning, the reference on bias, variance and model selection for twenty years. Demanding, free as a PDF.
For readers with a statistics background; without one the entry is hard going.
Training Compute-Optimal Large Language Models
The calculation showing that most large models were trained on too little data. Data volume has not been a side issue since.
For anyone comparing model sizes who needs to know why data volume counts.
Training Language Models to Follow Instructions
How a text continuation engine becomes an assistant. The paper behind alignment from human feedback.
For anyone asking how a text continuation engine becomes an assistant.
U-Net
Segmentation from few examples, developed in medical imaging. Still the first choice for segmentation.
For segmentation from few examples; still the first choice.
Understanding Deep Learning
A modern textbook with unusually clear figures that already covers transformers and diffusion models in full. Free as a PDF.
For a modern entry point; it already covers transformers and diffusion in full.
Very Deep Convolutional Networks
The paper showing that depth with small filters wins. The architecture convolutional networks are usually explained with.
For getting into convolutional networks; the architecture they are usually explained with.
vLLM
A high-throughput server with continuous batching and paged attention cache. The reason self-hosting becomes economical at volume.
For self-hosting at volume; the reason it can pay off at all.
You Only Look Once
Detection in a single pass instead of proposals and checks. The reason real-time object detection became possible.
For real-time object detection; explains why it became possible at all.