Artificial Agent Coordination
AntNet as an example of environment-mediated coordination.
Mechanism
This entry is limited to stigmergic coordination. It makes no claim of success or direct transfer to modern LLM agents.
Claims & sources
EvidenceIn AntNet, artificial agents share routing information through indirect, asynchronous communication mediated by the network.
Confidence is an editorial assessment of support, not a mathematical probability.
AntNet: Distributed Stigmergetic Control for Communications Networks (opens in a new tab)
- Authors
- Gianni Di Caro, Marco Dorigo
- Organization
- AI Access Foundation
- Source type
- Peer-reviewed paper
- Publication date
- Publication / identifier
- Journal of Artificial Intelligence Research 9, 317–365
- DOI
- 10.1613/jair.530
- Accessed
- Source availability
- Available
Supports — The publisher abstract explicitly identifies network-mediated communication; the scope is AntNet.
Publisher abstract, https://jair.org/index.php/jair/article/view/10217
SynthesisReading Ant System and AntNet together, SWI infers that shared environmental state can serve as a design pattern for artificial-agent coordination.
Confidence is an editorial assessment of support, not a mathematical probability.
Ant system: optimization by a colony of cooperating agents (opens in a new tab)
- Authors
- Marco Dorigo, Vittorio Maniezzo, Alberto Colorni
- Organization
- IEEE
- Source type
- Peer-reviewed paper
- Publication date
- Exact date not established; see publication details
- Publication / identifier
- IEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernetics) 26(1), 29–41; February 1996 (day not established)
- DOI
- 10.1109/3477.484436
- Accessed
- Source availability
- Available
Contextualizes — Artificial trail variables provide the optimization example for the synthesis.
Section II; reviewed manuscript PDF pp. 5–6: https://jmvidal.cse.sc.edu/library/dorigo96a.pdf#page=5
AntNet: Distributed Stigmergetic Control for Communications Networks (opens in a new tab)
- Authors
- Gianni Di Caro, Marco Dorigo
- Organization
- AI Access Foundation
- Source type
- Peer-reviewed paper
- Publication date
- Publication / identifier
- Journal of Artificial Intelligence Research 9, 317–365
- DOI
- 10.1613/jair.530
- Accessed
- Source availability
- Available
Supports — AntNet supplies the routing example; transfer to modern LLM agents is untested.
Publisher abstract, https://jair.org/index.php/jair/article/view/10217
Freshness
- Reviewed
- Updated
- Freshness evaluated