COGNITIVE SYSTEMS: Perfect Companions

Cognitive systems are those whose

behavior changes in response to and in anticipation of changes in the

environment, user characteristics and goals.

Computers have no knowledge of

the environment or the user–they don’t ‘know’ the existence of either.

We human beings tend to have an anthropomorphic view of machines. Hence terms

such as ‘stupid’ and ‘dumb’ are often applied to machines. A normal PC

can be made to acquire environment data such as temperature and light intensity,

with a little modification. Such data can be used to increase the effectiveness

of computers.

Rapid standardization is

occurring in the technologies of client-server computing and multimedia. Due to

the proliferation and ease of use of PCs, there is little that technology can do

to differentiate one product from another. "Company X has better multimedia

than company Y because they have better technology" is not a credible

sentence any more.

Conception and design will thus

be the key differentiators between future products. And both are cognitive

processes. Thus, the future of products would seem to depend on cognitive

engineering.

Building cognitive systems

Anyone who is in the business of

building systems that are easier to use than existing ones, should learn to

build cognitive systems. In particular, educational software producers, game

developers, kiosks and public-use terminal developers... and, of course, generic

application developers too. 

Currently, the crux of a

cognitive system is to try and make applications practical and feasible, with as

little work as possible in terms of programming. In other words, doing no

complex programming and yet coming out with a product that appears to exhibit

complex behavior.

Everything in nature is built

like this. Small cells that perform specialized functions interact with each

other and give rise to a complicated organism. While developing a cognitive

system, there is no need to design complexity but to let it evolve by itself.

Designing a cognitive bot

Bots are software robots capable

of learning, making decisions and responding to humans. In some jobs, bots could

replace human intervention. New-generation bots will be able to read stories to

children, take orders, provide interesting game interfaces and change the design

and look of future computer apps.

What does it take to make a bot?

Not a lot of programming. What you need is to create small ‘designlets’–small,

self-contained applets. These plug together, so that the result is emergent,

evolving and unpredictable.

Human beings use only one kind of

interface–the human interface. We speak and gesture. We listen, look and

emote. If we could understand this human interface and reproduce something akin

to it in a machine, it would be the friendliest and most natural interface you

can imagine. It is time programmers realized this and bring the man-machine

interface closer to the human user.

Can your computer do all these

things? The answer is yes. Let’s take a closer look at Cognitive systems.

Two types of entities fall under

the scope of Cognitive systems–humans and non-humans. Under ‘human’ we

have the natural and the artificial.

The former includes people who

mainly deal with ‘other people’, such as HR staff. When they interact

together, they display ‘Connectionism’. An example is the evolution of

Hotmail and Geocities. Those who created Hotmail could not imagine many people

all over the world would be using it. Hotmail has evolved through the enormous

usage by people. This is ‘Connectionism’.

The artificial human includes

pacemakers and robots. Modern pacemakers react cognitively. For example, when a

person is climbing the steps, the pacemaker pumps more blood to the heart

enabling the person to climb without difficulty. It is proactive and adaptive.

The non-human segment too has

natural and artificial members. The natural non-human includes animals. Take

animal behavior, such as in an ant colony or a beehive. A single bee or an ant

cannot produce anything concrete. Collections of these create beehives or

anthills. And there is a discontinuity in numbers. Three bees cannot build a

beehive, a thousand can. These phenomena display ‘Collectivism’. The

emergence of the Linux OS is a good example of a collective phenomenon that is

artificial, non-human and cognitive.

Finally, there are artificial

non-human systems. These include computer-based training packages and other

softwares.

Paradigm shift

In the near future, our PCs will

act as watchdogs, personal secretaries or companions. They will cajole us, scold

us and even pamper us. They will take the role of just about anybody and

everybody. So, do not be surprised if suddenly, while you are working, a message

pops up saying, "Hey Ruchi, you have a meeting at two," and a few

minutes later, "Ruchi, I think you really should get moving if you want to

make it on time". Or if you are angry and are pressing the keys too hard,

it might respond, "Look, don’t take out your anger on me! I have feelings

too!"

The next generation PCs will no

longer be ‘dumb’ or ‘stupid’. They will be responsive, adaptive and more

‘human-like’ in their interactions. They will respond differently to

different users, even throw tantrums on being misused and refuse to work.

Ritu

Dangwal