©IBM Corporation
©IBM Corporation is probably best known for developing
Deep Blue IIı, the computing sytem that defeated Kasporov in a chess
match sponsored by the Association of Computing Machinery from
February 10 -17, 1997. The shear computing power put to task to
accomplish this feat was impressive indeed. (see
http://www.ibm.com/Stretch/EOS/dbo.html).
Its architecture proves clearly that machines can be programmed to
effectively select from a large array of options and outperform
humans in strictly mathematical calculations. Much has yet to be done
to see how these capabilities can be applied to the field of
artifical intelligence.But one thing is certain. As we progress
quickly into the next millenium new scales of computing architectures
will be developed and existing massive parallel processing systems
will be scaled down for use in personal desktop computing
environments which will be essential in the development of more
advanced MC models. Linking this power to a rich and diverse world of
resources such as thoseavailable on the internet are opening up
exciting possiblities that need to be researched much more
extensively. ©IBM is also doing some amazing work in
the area of Intelligent Agentsı. The people at IBM understand the
massive commercial potential for developing systems capable of
learningı about the needs of their clients and responding
accordingly. In developing their system they talk about ³Agency,
Intelligence and Mobility² and provide a structure upon which to base
their designs. On their web site they talk about agents in these
terms:
Delegation of function (1) is an important response to these
trends, because through delegation the computer and the networks
behind it can assume more of the work. For example, software can
take on responsibility for information finding, retrieval and
filtering, can personalize human-computer interaction, and can
enable tasks to be carried out on behalf of users whether they are
present or absent, and with guidance rather than direct control by
the user. Intelligent Agents are software which implements this
delegation, thus managing complexity, supporting user mobility,
and lowering the entry skill level for new users. Intelligent
agents can best be thought of as a design model (2) , much like
client / server computing, rather than a technology or a product
offering. Though integration of intelligent agents into
applications or services requires a mindset change, this new model
is expected to be one of the key computing paradigms over the next
ten years (3).
http://www.networking.ibm.com/iag/iagptc2
As agentsı expand into the networks their functions extend beyond
the desktop computer out into the world. They become capable of
dynamically addressing the needs of the learner even when the learner
is absent or disconnected from the network. Upon reconnection the
agentı springs into action, bringing to the learner information or
carrying out more complex job tasks.
IBM© has several tools in development that deserve
further investigation. Click on any of the following for further
details. [Alter Ego]
[Activist]
[Globenet, RAISE]
[Knowledge Utility (KnU),
Aquí]
[IBM
Neural Network Utility]
[Lotus
Notes] [OBJECT
REXX] [Flow
Mark]
[Conversational
Agents]. The most significant aspect of this research and
development is the desire to establish an open standards architecture
or a ³common object model² upon which intelligent agents can be
developed. This will allow the utiliztion of agentsı across computer
platforms, networks and between applications all working together to
solve the learnerıs problems. Although IBM has directed their R and D
at commercial ventures there is considerable promise in adapting
these technologies to a myriad of more specific educational learning
environments.
Pick a different
paper