Uploaded July 2026 | Updated November 2026, 2 days ago
Book: Fundamentals of Active Inference
Principles, Algorithms, and Applications of the Free Energy Principle for Engineers, by Sanjeev V. Namjoshi https://mitpress.mit.edu/9780262050951/fundamentals-of-active-inference/
Register: http://textbook-group.activeinference.institute/
Code: github.com/ActiveInferenceInstitute/fundamentals
Playlist of Textbook Group: youtube.com/playlist?list=PLNm0u2n1IwdrpoTg-9l37YIBiDrrOzTw-
Direct link to group document: coda.io/d/_dZHp_UGhVZE/Meetings_su3ToqMp
------
CHAPTERS
0:00 Intro: Part One review and a live coding session
0:42 Cloning the fundamentals repo and setting up Cursor
1:03 Installing the Hermes agent harness
1:22 Finding free models on OpenRouter
1:48 Two ways in: the web interface and the terminal
2:07 Chapter-by-chapter animations and visualisations
2:44 Running the codebase with run.sh
3:22 Asking where variational free energy is calculated
4:18 You do not need any of this to read the repo
5:15 Requesting more focused, browser-only demos
6:18 Watching the model use tools rather than just complete text
7:19 Comparing the two harnesses' answers
8:04 Where to take an investigation question next
9:34 Building something new: the demo folder
10:51 Picking example situations from the chat
11:36 src, tests and thin orchestrators: why the separation matters
13:03 Collecting demo suggestions: bicycles, drones, saccades
14:04 Switching to plan mode, and why plans help long-horizon work
15:23 Example 2.3 and where the outputs are written
16:49 Local LLMs and free model options
17:32 Plan mode asks a clarifying question about integration
18:20 Modular, signposted software as a precondition for this speed
19:08 Reviewing the plan and hitting build
20:00 Asking for a learning path given only frequentist background
21:24 IDEs, plugins, and different ways to work
22:00 How integrated should a demo be with the core package?
23:03 Why reusable, tested software beats a blank-slate generation
24:02 Attention allocation across parallel agent conversations
25:50 Where you modify the repo vs where you only use it for context
27:04 Q&A: it does not have to be this complicated
28:42 Co-creating interfaces that put learners in flow
28:49 A browser-only Chapter 2 demo with sliders
30:43 Q&A: what basic skills do you need to start?
31:39 Step zero: IDE, Git, and the agent harness
32:23 How much the setup burden has changed
33:16 Porting the chapters to Jupyter notebooks in one prompt
33:43 Knowing what is trivial and what needs a work plan
34:20 Different interface surfaces, from Perplexity to local execution
36:01 Q&A: learning resources for Python and Git
37:48 W3Schools, DataCamp, Coursera, and project-driven learning
38:51 What a notebook is and why it helps
39:27 Vibe coding and missing the forest for the trees
39:53 The notebooks land, generated from the tested package
41:16 The deeper question: I want to do it by hand
42:41 What does by hand actually mean?
43:32 Epistemic value vs pragmatic value in how you learn
44:20 Turning learn-by-hand into a contributable document
45:05 A note on adding a licence file to your repo
45:52 Recap of the modes, from browsing to cloning to running
47:26 Reviewing the generated demos
48:52 Bicycle balance and drone flight demos
49:46 Metacognition: how much should you trust a generated method?
50:56 The saccade demo and its link to the ant foraging model
51:49 Computer science as one entry point among many
52:57 Closing notes for learners at different levels
54:12 The short setup checklist
55:10 Thanks and wrap-up/_dZHp_UGhVZE/Meetings_su3ToqMp
------
Active Inference Institute information:
Website: https://activeinference.institute/
X: https://x.com/InferenceActive
Discord: https://discord.activeinference.institute/
Donate: https://donate.activeinference.institute/
YouTube: youtube.com/c/ActiveInference
Active Inference Livestreams: https://video.activeinference.institute/
Book: Fundamentals of Active Inference
Principles, Algorithms, and Applications of the Free Energy Principle for Engineers, by Sanjeev V. Namjoshi https://mitpress.mit.edu/9780262050951/fundamentals-of-active-inference/
Register: http://textbook-group.activeinference.institute/
Code: github.com/ActiveInferenceInstitute/fundamentals
Playlist of Textbook Group: youtube.com/playlist?list=PLNm0u2n1IwdrpoTg-9l37YIBiDrrOzTw-
Direct link to group document: coda.io/d/_dZHp_UGhVZE/Meetings_su3ToqMp
------
CHAPTERS
0:00 Intro: Part One review and a live coding session
0:42 Cloning the fundamentals repo and setting up Cursor
1:03 Installing the Hermes agent harness
1:22 Finding free models on OpenRouter
1:48 Two ways in: the web interface and the terminal
2:07 Chapter-by-chapter animations and visualisations
2:44 Running the codebase with run.sh
3:22 Asking where variational free energy is calculated
4:18 You do not need any of this to read the repo
5:15 Requesting more focused, browser-only demos
6:18 Watching the model use tools rather than just complete text
7:19 Comparing the two harnesses' answers
8:04 Where to take an investigation question next
9:34 Building something new: the demo folder
10:51 Picking example situations from the chat
11:36 src, tests and thin orchestrators: why the separation matters
13:03 Collecting demo suggestions: bicycles, drones, saccades
14:04 Switching to plan mode, and why plans help long-horizon work
15:23 Example 2.3 and where the outputs are written
16:49 Local LLMs and free model options
17:32 Plan mode asks a clarifying question about integration
18:20 Modular, signposted software as a precondition for this speed
19:08 Reviewing the plan and hitting build
20:00 Asking for a learning path given only frequentist background
21:24 IDEs, plugins, and different ways to work
22:00 How integrated should a demo be with the core package?
23:03 Why reusable, tested software beats a blank-slate generation
24:02 Attention allocation across parallel agent conversations
25:50 Where you modify the repo vs where you only use it for context
27:04 Q&A: it does not have to be this complicated
28:42 Co-creating interfaces that put learners in flow
28:49 A browser-only Chapter 2 demo with sliders
30:43 Q&A: what basic skills do you need to start?
31:39 Step zero: IDE, Git, and the agent harness
32:23 How much the setup burden has changed
33:16 Porting the chapters to Jupyter notebooks in one prompt
33:43 Knowing what is trivial and what needs a work plan
34:20 Different interface surfaces, from Perplexity to local execution
36:01 Q&A: learning resources for Python and Git
37:48 W3Schools, DataCamp, Coursera, and project-driven learning
38:51 What a notebook is and why it helps
39:27 Vibe coding and missing the forest for the trees
39:53 The notebooks land, generated from the tested package
41:16 The deeper question: I want to do it by hand
42:41 What does by hand actually mean?
43:32 Epistemic value vs pragmatic value in how you learn
44:20 Turning learn-by-hand into a contributable document
45:05 A note on adding a licence file to your repo
45:52 Recap of the modes, from browsing to cloning to running
47:26 Reviewing the generated demos
48:52 Bicycle balance and drone flight demos
49:46 Metacognition: how much should you trust a generated method?
50:56 The saccade demo and its link to the ant foraging model
51:49 Computer science as one entry point among many
52:57 Closing notes for learners at different levels
54:12 The short setup checklist
55:10 Thanks and wrap-up/_dZHp_UGhVZE/Meetings_su3ToqMp
------
Active Inference Institute information:
Website: https://activeinference.institute/
X: https://x.com/InferenceActive
Discord: https://discord.activeinference.institute/
Donate: https://donate.activeinference.institute/
YouTube: youtube.com/c/ActiveInference
Active Inference Livestreams: https://video.activeinference.institute/










