LEARNING
// For members: the Collective's own presentations and guides, and a picked library of videos, courses and references. Never include patient information in anything you add.
> LOADING...
// Videos, courses and references
// Picked by the Collective from outside sources. Most of it is free; some courses and tools charge. Videos load from YouTube only when you press play.
// Start here: how AI works
No background needed. Pick one and watch it before anything else.
A short visual explanation of what a large language model is and how it produces text. The best first watch.
A hand-drawn overview of generative AI: what it is, what it can do, and how to work with it.
An introductory lesson on generative AI and how it differs from earlier machine learning.
A general-audience talk on how large language models work, where they are heading, and their security problems.
// Go deeper: what's under the hood
For when you want to understand the machinery. More mathematical, still visual.
The first chapter of a visual series on deep learning: what a neural network is, built up from scratch.
How the transformer, the design behind today's language models, processes text.
A step-by-step look at attention, the mechanism at the centre of a transformer.
A long, thorough walk through how models like ChatGPT are built and how to think about what they do.
// Using AI well: prompting and everyday practice
Practical skill. Never paste patient information into an AI tool that isn't approved for it.
A practical tour of using AI assistants day to day, with worked examples.
A hands-on, code-first guide to what language models can and can't do. Best if you like to tinker.
-
Prompt engineering overview
Anthropic
A structured guide to writing clear, effective prompts.
-
Interactive prompt engineering tutorial
Anthropic, on GitHub
A step-by-step course with exercises you work through yourself.
-
Prompt engineering guide
OpenAI
OpenAI's own guide to getting better results from its models.
// Video series: AI and the environment
Eight videos, about an hour and a half in all, in an order that builds: electricity first, then water, then the wider picture. They do not all agree, on purpose: the size of AI's footprint is still argued over, so watch more than one and check the figures against their sources. The Collective has not checked the claims made in them.
Start here: a short explainer on where the electricity goes.
The numbers, from the agency that tracks the world's energy use.
A short introduction to the magazine's own calculation of AI's energy use.
The water side of the story.
A longer video that challenges how AI's water use is usually reported. Watch it after the BBC piece and compare the two.
A longer investigation into AI data centers.
A physicist's view on the energy numbers and what she thinks matters more.
A talk by an AI researcher who studies the technology's footprint, on its present-day costs, including to the environment.
// Courses
Longer, structured learning. Check each provider's page for current cost and time.
-
Artificial Intelligence in Health Care Series
American Medical Association (ChangeMedEd) with the University of Michigan
A seven-part module series introducing AI and machine learning concepts in health care. Written for medical students and suitable for residents and faculty. Offered at no cost on the AMA Ed Hub.
-
AI in Healthcare Specialization
Stanford University, on Coursera
A multi-course series for clinicians and engineers on how AI is built on health data and how to judge whether it is safe to use.
-
AI for Medicine Specialization
DeepLearning.AI, on Coursera
A more technical, hands-on series on machine learning for diagnosis, prognosis and treatment. Assumes some programming.
-
AI For Everyone
DeepLearning.AI, on Coursera
A non-technical course on what AI can and cannot do. A good choice for people who lead or teach.
-
Elements of AI
University of Helsinki and MinnaLearn
A free online course on the basics of AI for people with no technical background.
// Guidance, policy and evidence
What the major bodies say, and where the research is published.
-
Principles for the Responsible Use of Artificial Intelligence in and for Medical Education
AAMC
The Association of American Medical Colleges' principles for schools and programs bringing AI into teaching.
-
Ethics and governance of artificial intelligence for health: guidance on large multi-modal models
World Health Organization
WHO guidance on the risks and governance of large multi-modal models in health care.
-
List of Artificial Intelligence-Enabled Medical Devices
U.S. Food and Drug Administration
The FDA's running list of AI-enabled devices it has authorized. Useful for checking whether a product is cleared.
-
NEJM AI
NEJM Group
A journal of AI in clinical medicine, from the publisher of the New England Journal of Medicine.
// Build your own
The Collective's own route from idea to working tool.
-
ClinicalToolsDEV
AI Medicine Collective (custom GPT; needs a paid ChatGPT subscription)
A guided assistant that walks you step by step through building a clinical tool app, with no prior coding experience.
See what members have already built in Tools. Members can add their own, and find presentations and guides in Learning Materials.