Poster session slides for ADDER paper. Presented in DRUID 2012 conference in June 2012 in Copenhagen.

1. ADDER
a new model for simulating the evolution of
technology,
with observations on why perfectly
knowledgeable agents cannot launch
technological revolutions
Janne M. Korhonen, Aalto University School of Economics
Julia Kasmire, Delft University of Technology

2. Background
• From the viewpoint of social scientist,
existing simulation models of
technology evolution have several
issues
• Environments are exogenous and static
• Internal “structure” of technologies is
missing

3. The requirements:
1. Early technologies form using existing primitive technologies as
components.
2. These new technologies in time become possible components, or
building blocks, for the construction of further new technologies.
3. This implies that technologies have an internal structure, a hierarchy
of subsystems and sub-subsystems.
4. The complex technologies form using simpler ones as components.
5. The overall collection of technologies bootstraps itself upward from
the few to the many and from the simple to the complex.
6. The evolution of technology happens both as a result of essentially
random events and boundedly rational search for and evaluation of
new solutions.

4. Solution: the ADDER
• Inspired by Arthur & Polak’s logic circuit
model of technological evolution (2006,
2009)
• Models endogenous change
• Models internal structures
• Is easy to implement (even for a social
scientist)

5. Advantages
• Models (or allows to model) easily several
stylized facts of technological evolution:
• Obsolescence, including “waves of
destruction”
• Keystone technologies
• Bounded rationality
• Variable difficulty of search

6. Nelson (2005): two key variables seem to
control the rate and direction of technological
advance:
1. the strength of technological understanding
2. the knowledge of user needs
The hypothesis was tested by building a model
incorporating
a. uncertainty about technologies
b. uncertainty about user needs
c. a learning curve for both

7. Example:
Nelson (2005): two key variables seem to
control the rate and direction of technological
advance:
1. the strength of technological understanding
2. the knowledge of user needs
The hypothesis was tested by building a model
incorporating
a. uncertainty about technologies
b. uncertainty about user needs
c. a learning curve for both

8. Example:
rational agents and (un)certain needs
Nelson (2005): two key variables seem to
control the rate and direction of technological
advance:
1. the strength of technological understanding
2. the knowledge of user needs
The hypothesis was tested by building a model
incorporating
a. uncertainty about technologies
b. uncertainty about user needs
c. a learning curve for both

9. no uncertainty lots of uncertainty
Technology
Tech level
time time

10. no uncertainty lots of uncertainty
Technology
Tech level
time time
Uncertainty about outcomes
seems to be necessary for
technological development!

11. Details
• Very simple idea:
• Simulation starts with primitive technology (“1”)
• Goal is to satisfy needs, expressed as real
numbers (e.g. 5, 8, 14, 45, 106)
• New technologies are arithmetical expressions
(e.g. 1 + 1 + 1 - 1 = 3)
• Developed technologies* can serve as
components (e.g. 1 + 1 + 3 = 5)
* Provided the technology is among the list of
“possible” technologies

12. Details (2)
• Criteria for successful technologies; e.g.
• Cost: counting the number of primitives in
each technology (e.g. tech 6, composed of 1
+ 2 + 3, or 1 + (1 + 1 + 1 - 1) + (1 + 1 + 1) cost
= 8)
• Number of components
• “Anti-targets:” specific unfeasible technologies

13. Contact
• For further info, contact Janne M.
Korhonen
• janne.m.korhonen@aalto.fi,
jmkorhonen.net
• Or Julia Kasmire,
• j.kasmire@tudelft.nl

14. Main references
• Arthur, B.W., 2009. The Nature of Technology: What it is
and how it evolves. Free Press, New York.
• Arthur, W.B., Polak, W., 2006. The evolution of technology
within a simple computer model. Complexity 11, 23-31.
• Nelson, R. R. (2005). Perspectives on technological
evolution. In K. Dopfer (Ed.), The Evolutionary Foundation
of Economics (pp. 461-471). Cambridge: Cambridge
University Press.
• See paper for more!