Uploaded June 2025 | Updated November 2026, 4 days ago
The Value of Open Source Software
Manuel Hoffmann, Frank Nagle, Yanuo Zhou
papers.ssrn.com/sol3/papers.cfm?abstract_id=4693148
Abstract: The value of a non-pecuniary (free) product is inherently difficult to assess. A pervasive example is open source software (OSS), a global public good that plays a vital role in the economy and is foundational for most technology we use today. However, it is difficult to measure the value of OSS due to its non-pecuniary nature and lack of centralized usage tracking. Therefore, OSS remains largely unaccounted for in economic measures. Although prior studies have estimated the supply side costs to recreate this software, a lack of data has hampered estimating the much larger demand-side (usage) value created by OSS. Therefore, to understand the complete economic and social value of widely-used OSS, we leverage unique global data from two complementary sources capturing OSS usage by millions of global firms. We first estimate the supply-side value by calculating the cost to recreate the most widely used OSS once. We then calculate the demand side value based on a replacement value for each firm that uses the software and would need to build it internally if OSS did not exist. We estimate the supply-side value of widely-used OSS is $4.15 billion, but that the demand-side value is much larger at $8.8 trillion. We find that firms would need to spend 3.5 times more on software than they currently do if OSS did not exist. The top six programming languages in our sample comprise 84% of the demand-side value of OSS. Further, 96% of the demand-side value is created by only 5% of OSS developers.
Active Inference Institute information:
Website: https://www.activeinference.institute/
Activities: https://activities.activeinference.institute/
Discord: https://discord.activeinference.institute/
Donate: http://donate.activeinference.institute/
YouTube: youtube.com/c/ActiveInference
X: https://x.com/InferenceActive
Active Inference Livestreams: https://video.activeinference.institute/
The Value of Open Source Software
Manuel Hoffmann, Frank Nagle, Yanuo Zhou
papers.ssrn.com/sol3/papers.cfm?abstract_id=4693148
Abstract: The value of a non-pecuniary (free) product is inherently difficult to assess. A pervasive example is open source software (OSS), a global public good that plays a vital role in the economy and is foundational for most technology we use today. However, it is difficult to measure the value of OSS due to its non-pecuniary nature and lack of centralized usage tracking. Therefore, OSS remains largely unaccounted for in economic measures. Although prior studies have estimated the supply side costs to recreate this software, a lack of data has hampered estimating the much larger demand-side (usage) value created by OSS. Therefore, to understand the complete economic and social value of widely-used OSS, we leverage unique global data from two complementary sources capturing OSS usage by millions of global firms. We first estimate the supply-side value by calculating the cost to recreate the most widely used OSS once. We then calculate the demand side value based on a replacement value for each firm that uses the software and would need to build it internally if OSS did not exist. We estimate the supply-side value of widely-used OSS is $4.15 billion, but that the demand-side value is much larger at $8.8 trillion. We find that firms would need to spend 3.5 times more on software than they currently do if OSS did not exist. The top six programming languages in our sample comprise 84% of the demand-side value of OSS. Further, 96% of the demand-side value is created by only 5% of OSS developers.
Active Inference Institute information:
Website: https://www.activeinference.institute/
Activities: https://activities.activeinference.institute/
Discord: https://discord.activeinference.institute/
Donate: http://donate.activeinference.institute/
YouTube: youtube.com/c/ActiveInference
X: https://x.com/InferenceActive
Active Inference Livestreams: https://video.activeinference.institute/
![ActInf ModelStream 019.1: Active Inference in Discrete State Spaces from First Principles (P. Kenny)
Active Inference in Discrete State Spaces from First Principles
Patrick Kenny
https://arxiv.org/abs/2511.20321
We seek to clarify the concept of active inference by disentangling it from the Free Energy Principle. We show how the optimizations that need to be carried out in order to implement active inference in discrete state spaces can be formulated as constrained divergence minimization problems which can be solved by standard mean field methods that do not appeal to the idea of expected free energy. When it is used to model perception, the perception/action divergence criterion that we propose coincides with variational free energy. When it is used to model action, it differs from an expected free energy functional by an entropy regularizer.
[Submitted on 25 Nov 2025 (v1), last revised 10 Dec 2025 (this version, v2)]
Cite as: arXiv:2511.20321 [cs.AI]
(or arXiv:2511.20321v2 [cs.AI] for this version)
https://doi.org/10.48550/arXiv.2511.20321
Active Inference Institute information:
Website: https://www.activeinference.institute/
Activities: https://activities.activeinference.institute/
Discord: https://discord.activeinference.institute/
Donate: http://donate.activeinference.institute/
YouTube: https://www.youtube.com/c/ActiveInference/
X: https://x.com/InferenceActive
Active Inference Livestreams: https://video.activeinference.institute/
CHAPTERS
0:00 Welcome and introductions
1:30 Active inference from first principles
3:00 Prediction and a KL divergence criterion
4:30 Intractable target distributions
6:00 The mean field approximation
7:30 Clamping observed variables
9:00 Hidden and observable variables
10:30 Variational free energy
12:00 Prediction horizon and hidden sequences
13:30 Approximating the posterior over the future
15:00 Constraints from the hidden Markov model
16:30 Standard update formulas
18:00 Retrodiction: updating past states
19:30 Predictive distributions for future observations
21:00 Connection to the free energy principle
22:30 From joint to marginal divergence
24:00 Comparison with expected free energy
25:30 Entropy regularization
27:00 Dropping the independence assumption
28:30 A subtle difference from the FEP
30:00 Learning HMM parameters
31:30 Posterior inference from past observations
33:00 Summary of the three problems
34:30 Mean field updates and convergence
36:00 Conclusion
37:30 Discussion: writing the free energy
40:30 Discussion: forms of the functionals
42:00 Discussion: conditional effects of actions
43:30 Folding actions and policies into HMM states
46:30 Dynamic approximating distributions
48:00 Action and perception as divergence minimization
49:30 Factorized approximations explained
51:00 Past clamped, future open
52:30 Fixed points of the objective
55:30 Entropies of emission distributions
57:00 Clarifying the KL divergences
1:00:00 Derivations in the appendix
1:01:30 Mean field and model adequacy
1:04:30 A discrete HMM at the center
1:06:00 Reflections on the paper
1:07:30 Motivations for the work
1:09:00 Debugging variational algorithms
1:10:30 Divergences and non-divergences
1:12:00 Continuous time and step sizes
1:15:00 From in silico to wetware
1:18:00 Continuous-state models
1:21:00 Random fluctuations across coordinates
1:22:30 Expected free energy: conventional vs. proposed
1:25:30 Free energy of the expected future
1:27:00 The physics of the free energy principle
1:28:30 Deriving policy priors
1:31:30 Habits and policy priors
1:33:00 Passive observers vs. agents
1:36:00 Convergence across approaches
1:37:30 Policies without explicit encoding
1:40:30 Deliberate action
1:43:30 Agent and niche
1:45:00 Predicting actions via sensations
1:48:00 The lowest levels of active inference models
1:49:30 One generative model for everything?
1:51:00 Closing ActInf ModelStream 019.1: Active Inference in Discrete State Spaces from First Principles (P. Kenny)](https://i.ytimg.com/vi/q6Z-4aXO36U/mqdefault.jpg)









