Download Understanding Agenticity and Multi-Agent Systems: Communication, Autonomy, and Learning and more Slides Artificial Intelligence in PDF only on Docsity!
Introduction
1
- As our programs become more “intelligent” we begin to provide them with certain autonomy and certain purpose.
- These new programs in the parlor of AI are known as “Agent” that is to say program with a purpose.
- Yet what differentiates a program from an “Agent? Well that is a good semantic discussion in AI.
- Let us for the time being just say that agent have a new quality called “agenticy” and this new quality requires certain characteristics in a program
- Docsity.com
Agenticy in a program: Communication, autonomy, capability
2
- Autonomous performance (e.g. threads)
- Information handling (e.g. data base access)
- Knowledge handling (e.g. beliefs, desires, intentions)
- Goal seekers (i.e. purposeful)
- Goal setters (i.e. purposive, very very very few)
- Dynamic behaviors (i.e. not single I/O mapping)
- May be embedded in another program
- It is user enabled yet “empowered” to act on its own
- Ability to work with other agents (i.e. multiple agents)
- Ability to improve its behavior over time (i.e. learning)
- Awareness and understanding of an environment (e.g. ontology driven)
MASs definition
Definition : A multi-agent system (MAS) consists of a network of loosely-coupled computational autonomous agents who can perform actions, they have resources at their disposal and they possess knowledge, capabilities or skills. They are situated in a common environment and they can interact through a set of rules, namely an interaction protocol.
It is rare for an agent to act in isolation and it is even more rare for an agent to be useful on its own
The need for a common framework:
Ontologies
5
Once we get involved with common knowledge then we have a problem of common understanding among programs ( aka agents ).
How do we standardized our concepts, this is in AI the field of ontologies.
What is the meaning of meaning?
Rather than using my own ppt I will use some links to explain this ever growing area of study
A semantic continuum
[Mike Uschold, Boeing Corp]
Shared human consensus
Text descriptions
Semantics hardwired; used at runtime
Semantics processed and used at runtime
Pump: “a device for moving a gas or liquid from one place or container to another”
(pump has (superclasses (…))
Implicit Informal (explicit)
Formal (for humans)
Formal (for machines)
Further to the right means:
- Less ambiguity
- More likely to have correct functionality
- Better inter-operation
- Less hardwiring
- More robust to change
- More difficult
Tightness of Coupling & Semantic Explicitness
Implicit, TIGHT
Explicit, Loose
Local^ 1 System: Small Set of Developers
Far
Same Process Space
Same Address Space
Same CPU
Same OS Same Programming Language
Same DBMS
Same Local Area Network Systems of Systems
Enterprise
Community
Internet
Same Wide Area Network Client-Server
Same Intranet
Federated DBs
Data Warehouses
Data Marts
Workflow Ontologies
Compiling
Linking
Agent Programming Web Services: SOAP
Distributed Systems OOP
Applets
Semantic Mappings (^) Semantic Brokers
Looseness of Coupling
Semantics Explicitness
XML, XML Schema
Conceptual Models
RDF/S, OWL Web Services: UDDI, WSDL
OWL-S
Modal Policies
Middleware Web
Peer-to-peer
N-Tier Architecture EAI
From Synchronous Interaction to Asynchronous Communication
Performance = k / Integration_Flexibility
