CodeX
Published in

CodeX

The Difference Between Symbolic AI and Connectionist AI

AI Development

Whenever there are two categories of something, people do not wait to take sides and then compare the two. The same is the situation with Artificial Intelligence techniques such as Symbolic AI and Connectionist AI. The latter has found success and media’s attention, however, it is our duty to understand the significance of both Symbolic AI and Connectionist AI.

Truth being said, these two sets of techniques have their own place as the pedestal. There’s no such thing as AI (Artificial Intelligence) that can be used any and everywhere. There are various AI development services available for various uses and for multiple audiences. Each of the AI techniques has its own strengths and weaknesses, however, choosing the right thing is a bit of a task.

What is Symbolic AI?

Symbolic AI is more commonly known as rule-based AI, good old-fashioned AI (GOFA), and classic AI. Earlier AI development research was based on Symbolic AI which relied on inserting human behavior and knowledge in the form of computer codes.

We humans have used symbols to drive meaning from things and events in the environment around us. For example, imagine you told your friend to buy you a bottle of Coke. Your friend would first have an image of a bottle of coke in his mind. This is the very idea behind the symbolic AI development, that these symbols become the building block for cognition.

Any application made with Symbolic AI has a combination of characters signifying real-world concepts or entities through a series of symbols. These symbols can easily be arranged through networks and lists or arranged hierarchically. Such arrangements tell the AI algorithms how each symbol is related to each other in totality.

Information in Symbolic AI is processed through something that is called an expert system. It is where the if/then pairing directs the algorithm to the parameters on which it can behave. These expert systems are man-made knowledge bases. The inference engine is a term given to a component that refers to the knowledge base and selects rules to apply to given symbols.

Pros and Cons of Symbolic AI

Symbolic AI is well suited for applications that are based on crystal clear rules and goals. If you want this AI to beat a human in the game of chess then we need to teach the algorithm the specifics of chess. This framework acts like a boundary that helps it operate properly.

Symbolic AI falls short when it is required to encounter variations. Taking an example of machine vision, which might look at a product from all the possible angles. It would be tedious and time-consuming to create rules for all the possible combinations. The real world contains huge amounts of data and numerous variations. It is difficult to anticipate all the possible alterations in a given environment.

What is a Connectionist AI?

Earlier experts focused on the symbolic type AI for many decades however, the Connectionist AI is more popular now. This AI is based on how a human mind functions and its neural interconnections. This technique of AI software development is also sometimes called a perceptron to signify a single neuron.

An application built with Connectionist AI tends to get more intelligent as we keep on feeding data and learning patterns and relations associated with the environment and with itself. On the other hand, symbolic AI gets hand-coded. To understand connectionist AI let’s take the example of an artificial neural network. Each one is made up of hundreds of single units processing elements and artificial neurons. They are a layered format with weights forming connections with this structure where weights are adjustable parameters.

In Connectionist AI all the processing elements have weighted units, output, and a transfer function. However, it is to keep in mind that the transfer function assesses multiple inputs and then it combines them into a single output value. Each weight in the algorithm efficiently evaluates directionality and importance and eventually the weighted sum is the component that activates the neuron. When all is done then the activated signal passes through the transfer function and produces one output.

Pros and cons of Connectionist AI

When you have high-quality training data Connectionist AI is a good option to be fed with that data. Even though this AI model gets smarter as data is fed into it, it still needs the support of accurate information to start the whole learning process. Connectionist AI is quite famous in the healthcare industry. It is most commonly used when there is a heap of medical images that are required to be verified by humans for correctness and assign annotations for contexts.

With all the pros, this AI often cannot explain how it reached a solution. Thus it is advised not to select this AI as the primary or the sole choice as the conclusions were drawn by it cannot be explained and would require the help of a third party. For example, using connectionist AI to decide if a person is a murderer or not. People are bound to consider it unjust and cruel to rely on AI, which does not explain how it reached the conclusion that it did.

Conclusion

AI software development is now becoming more mainstream than before as it is widely used and accepted. However, this article shows why it is important to understand how these AI operate and choose the right one for them.

--

--

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
Nora Winkens

Nora Winkens

Daffodil Software is a partner in software technology for more than 100 organizations around the world.