Learning paths

Understand how modern AI works, use it well at work, and know enough to talk with technical teams. No coding.

1. Elements of AI: Introduction to AI

   University of Helsinki and MinnaLearn · ~30 hours, self-paced
   Free
2. AI For Everyone

   DeepLearning.AI · ~7 hours (4 modules)
   Free to audit · ~$49
3. Anthropic's Prompt Engineering Interactive Tutorial

   Anthropic · 9 chapters, self-paced
   Free
4. Google AI Essentials Specialization

   Google · Under 10 hours, self-paced
   $49/mo
5. Career Essentials in Generative AI by Microsoft and LinkedIn

   Microsoft and LinkedIn · ~4-6 hours (5 courses)
   Free

The classic route: Python and math foundations, then the Stanford/DeepLearning.AI curriculum, then deep learning and shipping models to production.

1. Python for Everybody Specialization

   University of Michigan · 2 months at 10 hrs/week (5 courses, ~79 hours)
   Free to audit · $49/mo
2. Mathematics for Machine Learning and Data Science Specialization

   DeepLearning.AI · 12 weeks at 5 hrs/week (3 courses, ~94 hours)
   Free to audit · $49/mo
3. Machine Learning Specialization

   Stanford Online / DeepLearning.AI · 2 months at 10 hrs/week (~95 hours, 3 courses)
   Free to audit · $49/mo
4. Deep Learning Specialization

   DeepLearning.AI · 3 months at 10 hrs/week (~129 hours, 5 courses)
   Free to audit · $49/mo
5. Practical Deep Learning for Coders

   fast.ai · 9 lessons, ~90 minutes each plus notebooks
   Free
6. Machine Learning in Production

   DeepLearning.AI · ~11 hours (3 modules)
   Free to audit · $49/mo
7. Made With ML

   Anyscale (Goku Mohandas) · Self-paced (foundations + MLOps lessons)
   Free

Go from calling a model API to building agents with tool use and MCP, using official training from Anthropic, Hugging Face and Microsoft.

1. Claude Platform 101

   Anthropic · Self-paced (~12 lessons)
   Free
2. Anthropic's Prompt Engineering Interactive Tutorial

   Anthropic · 9 chapters, self-paced
   Free
3. Building with the Claude API

   Anthropic · Self-paced (multi-module with final assessment)
   Free
4. Introduction to Model Context Protocol

   Anthropic · Self-paced
   Free
5. AI Agents Course

   Hugging Face · ~15-20 hours (5 units plus bonus units)
   Free
6. Claude Code in Action

   Anthropic · Self-paced (4 modules)
   Free
7. Hugging Face LLM Course

   Hugging Face · 12 chapters, self-paced
   Free

Build intuition with the best visual explanations, then code neural networks from scratch and follow the MIT and Stanford university courses.

1. Neural networks

   3Blue1Brown (Grant Sanderson) · ~3 hours total (8 videos)
   Free
2. Neural Networks: Zero to Hero

   Andrej Karpathy · 8 lectures (~15+ hours), ongoing
   Free
3. MIT 6.S191: Introduction to Deep Learning

   MIT · 9 lectures plus 3 software labs (~2 weeks intensive)
   Free
4. Dive into Deep Learning

   d2l.ai (Zhang, Lipton, Li & Smola) · Self-paced (full textbook)
   Free
5. CS231n: Deep Learning for Computer Vision

   Stanford University · 10 weeks (one quarter)
   Free
6. CS224N: Natural Language Processing with Deep Learning

   Stanford University · 10 weeks (one quarter)
   Free

Build strategic fluency fast with free material, then decide whether a branded executive program from MIT or Oxford is worth the investment.

1. AI For Everyone

   DeepLearning.AI · ~7 hours (4 modules)
   Free to audit · ~$49
2. Elements of AI: Introduction to AI

   University of Helsinki and MinnaLearn · ~30 hours, self-paced
   Free
3. Artificial Intelligence: Implications for Business Strategy

   MIT Sloan School of Management and MIT CSAIL · 6 weeks at 6-8 hrs/week
   $3,850
4. Oxford Artificial Intelligence Programme

   University of Oxford, Saïd Business School · 6 weeks at 7-10 hrs/week
   $3,280

Stack recognized credentials: start with the entry-level AWS and Azure AI exams, then move to the professional-grade Google Cloud ML Engineer.

1. AWS Certified AI Practitioner

   Amazon Web Services · Exam: 90 minutes, 65 questions
   $100 exam
2. Microsoft Certified: Azure AI Fundamentals (Exam AI-901)

   Microsoft · Exam: approx. 45-65 minutes
   $99 exam
3. Generative AI Leader

   Google Cloud · Exam: 90 minutes, 50-60 multiple choice questions
   $99 exam
4. IBM AI Engineering Professional Certificate

   IBM · 13 courses, approx. 4 months at 10 hrs/week
   Free to audit · $49/mo
5. Professional Machine Learning Engineer

   Google Cloud · Exam: 2 hours, 50-60 multiple choice questions
   $200 exam

From the accessible Helsinki ethics course to the technical safety curriculum used by alignment researchers.

1. Ethics of AI

   University of Helsinki · 7 chapters, self-paced (approx. 20 hours)
   Free
2. Introduction to AI Safety

   Center for AI Safety · Approx. 4 hours, 3 modules
   Free
3. AI Alignment

   BlueDot Impact · 12 weeks, approx. 5 hrs/week (8 weeks learning plus 4-week project)
   Free
4. Intro to ML Safety

   Center for AI Safety · Approx. 8 weeks of lectures and assignments, self-paced
   Free

The sequence serious RL learners follow: the DeepMind lectures, a hands-on course where you train agents, then Berkeley's deep RL and the OpenAI reference material.

1. Reinforcement Learning (UCL Advanced Topics 2015, COMPM050/COMPGI13)

   UCL / Google DeepMind (David Silver) · 10 lectures, ~1.5 hours each
   Free
2. Deep Reinforcement Learning Course

   Hugging Face · 8 units plus bonus units, self-paced
   Free
3. Spinning Up in Deep RL

   OpenAI · Self-paced (docs, exercises, and code)
   Free
4. Deep Reinforcement Learning (CS 285)

   UC Berkeley · One semester (~25 lectures)
   Free