Skip to main content

Blankets All the Way up – the Economics of Active Inference

  • Conference paper
  • First Online:
Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML PKDD 2021)


A direct implication of active inference, by way of minimizing expected free energy, is the ability to reframe optimization problems as they relate to biological systems. Instead of employing objective functions in order to maximizing an agent’s exposure to some exogenous measurable quantity, active inference describes how biological systems optimize by minimizing a divergence (KL) between a posterior probability density and a generative density, by definition endogenous to the system. This particular framework can be shown to underwrite many seemingly disparate disciplines in economics, and may prove to be a source of new insights for the field.

This is a preview of subscription content, log in via an institution to check access.

Access this chapter

USD 29.95
Price excludes VAT (USA)
  • Available as PDF
  • Read on any device
  • Instant download
  • Own it forever
USD 129.00
Price excludes VAT (USA)
  • Available as EPUB and PDF
  • Read on any device
  • Instant download
  • Own it forever
Softcover Book
USD 169.99
Price excludes VAT (USA)
  • Compact, lightweight edition
  • Dispatched in 3 to 5 business days
  • Free shipping worldwide - see info

Tax calculation will be finalised at checkout

Purchases are for personal use only

Institutional subscriptions


  1. 1.

    The observation that goods or services are preferred sooner rather than later, all else being equal. In economics, the prefixes “high” or “low” is sometimes used in order to differentiate between various levels of “impatience”. High time preference agents will value time at a high rate and display high levels of impatience, while low time preference agents will display low levels of impatience. High or low time preferences are therefor often used as explanations for various levels of propensities to consume or save in an economy.

  2. 2.

    Revealed preferences is a way to infer an agents utility function by observing past behaviour. As such, agents cannot change their preferences once they have been revealed, since this would change the utility function and hereby greatly complicate economic modelling practices. The concept of revealed preferences is tightly linked to the transitivity axiom [18, 19].

  3. 3.

    The idea, that in situations where optimal solutions do not exist, an agent will search until a solution is deemed to be good enough.

  4. 4.

    There is at present no corollary in the economic literature to a rate of remuneration as used in this presentation. Normally a remuneration rate simply refers to a salary or stream of payments due for work or services rendered. Here, a rate of remuneration refers the expected energy input for a system given energy output, where the minimum requirement is long run homeostasis. For this reason the term may be ill conceived. Conversely, the term as used herein perfectly captures the observation, that agents have a preference for increasing over declining sequences not strictly “permissible” under a rational expectations framework [25, 26].

  5. 5.

    Normally present value refers to the value at present of a discounted future cash flow where \(PV= \frac{CF}{{\left(1+r\right)}^{n}}\). Here, the term refers to subjective value given a time component. As such, it is the expected value of something that by necessity must lie in the future, and therefore must be discounted to some degree, considering a generative model that takes surprise or uncertainty into account. We can therefore also treat present value and utility (discounted) as interchangeable.


  1. Ramstead, M.J., Kirchhoff, M.D., Friston, K.J.: A tale of two densities: active inference is enactive inference. Adapt. Behav. 28(4), 225–239 (2020).

    Article  Google Scholar 

  2. Friston, K.J., Schwartenbeck, P., FitzGerald, T., Moutoussis, M., Behrens, T., Dolan, R.J.: The anatomy of choice: active inference and agency. Front. Hum. Neurosci. 25, 598 (2013).

    Article  Google Scholar 

  3. Henriksen, M.: Variational free energy and economics optimizing with biases and bounded rationality. Front. Psychol. 11, 549187 (2020).

    Article  Google Scholar 

  4. Friston, K.J., FitzGerald, T., Rigoli, F., Schwartenbeck, P., O’Doherty, J., Pezzulo, G.: Active inference and learning. Neurosci. Biobehav. Rev. 68, 862–879 (2016). ISSN 0149-7634.

  5. Arthur, W.B.: Foundations of complexity economics. Nat. Rev. Phys. 3, 136–145 (2021).

    Article  Google Scholar 

  6. Arthur, W.B.: Complexity and the Economy. Oxford University Press, Oxford (2015). ISBN 978-0-19-933429-2

    Google Scholar 

  7. Farmer, J., Foley, D.: The economy needs agent-based modelling. Nature 460, 685–686 (2009).

    Article  Google Scholar 

  8. Parr, T., Friston, K.J.: Generalised free energy and active inference. Biol. Cybern. 113(5–6), 495–513 (2019).

    Article  MathSciNet  MATH  Google Scholar 

  9. Milling, B., Tschantz, A., Buckley, C.L.: Whence the Expected Free Energy? Neural Comput. 33(2), 447–482 (2021).

    Article  MathSciNet  MATH  Google Scholar 

  10. Friston, K.J., Rigoli, F., Ognibene, D., Mathys, C., Fitzgerald, T., Pezzulo, G.: Active inference and epistemic value. Cogn. Neurosci. 6(4), 187–214 (2015).

    Article  Google Scholar 

  11. Kim, C.S.: Bayesian mechanics of perceptual inference and motor control in the brain. Biol. Cybern. 115(1), 87–102 (2021).

    Article  MATH  Google Scholar 

  12. Botvinick, M., Toussaint, M.: Planning as inference. Trends Cogn. Sci. 16(10), 485–488 (2012).

    Article  Google Scholar 

  13. Roseboom, W., Fountas, Z., Nikiforou, K.: Activity in perceptual classification networks as a basis for human subjective time perception. Nat. Commun. 10, 267 (2019).

    Article  Google Scholar 

  14. Zakharov, A., Crosby, M., Fountas, Z.: Episodic Memory for Learning Subjective-Timescale Models (2020). arXiv:2010.01430

  15. Lopez-Parsem, A., Domenech, P., Pessiglione, M.: How prior preferences determine decision-making frames and biases in the human brain. eLife 5, e20317 (2016).

  16. Friston, K., Samothrakis, S., Montague, R.: Active inference and agency: optimal control without cost functions. Biol. Cybern. 106, 523–541 (2012).

    Article  MathSciNet  MATH  Google Scholar 

  17. Simon, H.A.: Rational choice and the structure of the environment. Psychol. Rev. 63(2), 129–138 (1956).

    Article  Google Scholar 

  18. Samuelson, P.: A note on the pure theory of consumers’ behaviour. Econometrica 5, 61–71 (1938).

    Article  Google Scholar 

  19. Von Neumann, J., Morgenstern, O.: Theory of Games and Economic Behaviour. Princeton University Press, Princeton, NJ (1947)

    MATH  Google Scholar 

  20. Kahneman, D., Knetsch, J.L., Thaler, R.: Anomalies: the endowment effect, loss aversion, and status Quo bias. J. Econ. Perspect. 5, 193–206 (1991).

    Article  Google Scholar 

  21. Ariely, D., Loewenstein, G., Prelec, D.: Coherent arbitrariness: stable demand curves without stable preferences. Q. J. Econ. 118(1), 73–106 (2003).

    Article  MATH  Google Scholar 

  22. Coase, R.H.: The problem of social cost. J. Law Econ. 3, 63–94 (1960)

    Article  Google Scholar 

  23. Williamson, O.: The economics of organization – the transaction cost approach. Am. J. Sociol. 87(3), 548−577 (1981).

  24. Williamson, O.: Transaction Cost Economics - An Introduction; Discussion Paper No. 2007-3 (2007).

  25. Loewenstein, G.F., Prelec, D.: Preferences for sequences of outcomes. Psychol. Rev. 100(1), 91–108 (1993).

  26. Scholten, M., Read, D.: Better is worse, worse is better: Violations of dominance in intertemporal choice. Decisions 1(3), 215–222 (2014).

    Article  Google Scholar 

Download references

Author information

Authors and Affiliations


Corresponding author

Correspondence to Morten Henriksen .

Editor information

Editors and Affiliations

Rights and permissions

Reprints and permissions

Copyright information

© 2021 Springer Nature Switzerland AG

About this paper

Check for updates. Verify currency and authenticity via CrossMark

Cite this paper

Henriksen, M. (2021). Blankets All the Way up – the Economics of Active Inference. In: Kamp, M., et al. Machine Learning and Principles and Practice of Knowledge Discovery in Databases. ECML PKDD 2021. Communications in Computer and Information Science, vol 1524. Springer, Cham.

Download citation

  • DOI:

  • Published:

  • Publisher Name: Springer, Cham

  • Print ISBN: 978-3-030-93735-5

  • Online ISBN: 978-3-030-93736-2

  • eBook Packages: Computer ScienceComputer Science (R0)

Publish with us

Policies and ethics