Join online with today’s leading executives at the Data Summit on March 9th. Register here.
Although many enterprises are starting to dip their toes into the AI pool with Rudimentary Machine Learning (ML) and Deep Learning (DL) models, a new form of technology called Symbolic AI is emerging from the lab on how AI works. And the ability to improve both how it relates to its human observers.
Followers of Symbolic AI say that it follows the logic of biological intelligence more closely as it analyzes symbols, not just data, to arrive at more intuitive, knowledge-based conclusions. It is commonly used in linguistic models such as Natural Language Processing (NLP) and Natural Language Understanding (NLU), but it is rapidly finding its way into ML and other types of AI where it can bring much needed visibility into algorithmic processes. .
What is old is new
According to expert Luca Scagliarini, the technology dates back to the 1950s, but was considered old-fashioned in the 1990s, when the demand for sensory and motor processes increased. Now that AI has been entrusted with the task of high-end systems and data management, the ability to engage in the presentation of logical thinking and knowledge is again great.
One of the keys to symbolic AI success is how it works in a rules-based environment. Typical AI models deviate from their original purpose because the new data influences changes in the algorithm. Scagliarini says that the rules of symbolic AI resist drift, so the model can be started very quickly and with very little data, and then requires less training once it enters the production environment.
Because they are bound by rules, however, symbolic algorithms cannot improve themselves over time, which, after all, is one of the key value proposals that AI brings to the table, says Jans Asman, CEO of knowledge graph solutions provider Franz Inc. That’s why Symbolic AI is being integrated into ML, DL, and other forms of rule-free AI to create a hybrid environment that provides the best for both worlds: complete machine intelligence with a logic-based brain that improves with every application.
This, in turn, enables AI to be trained using a variety of techniques, including semantic inference and supervised and unsupervised learning, which will eventually create AI systems that can reason, learn and engage in natural language question-and-answer interactions with humans. Already, the technology is finding its way into complex tasks such as fraud analysis, supply chain optimization, and sociological research.
Jelly Harper of Analytics Week says this is a turning point for the enterprise. Data fabric developers like Stardog are working to combine both logical and numerical AI to analyze classified data; That is, the data that is classified in order of importance for the enterprise. Symbolic AI plays a crucial role in interpreting the rules governing this data and in rationalizing its accuracy. This will ultimately allow these organizations to apply multiple forms of AI to cope with virtually any and all situations in the digital realm – essentially using one AI to overcome the shortcomings of another.
Tech journalist Surya Madula says that while organizations are looking forward to the day when individuals can interact with AI, the symbolic AI is how it will happen. After all, we humans have developed reason by first learning the rules of how things relate to each other, then applying those rules to other situations – the way symbolic AI is trained. Integrating this form of cognitive reasoning into deep neural networks allows researchers to create what they call neuro-symbolic AI, learning and maturing using the same basic rules-oriented framework that we do.
While this may be frustrating for some, it should be remembered that symbolic AI still only works with numbers, only in a different way. By building a more humane thinking machine, organizations will be able to democratize technology across the workforce so that it can be applied to the real world situations we face every day.
It will certainly not be able to solve all our problems, but it will relieve us of the most annoying problems.
Venturebeat’s mission Transformative Enterprise is about to become a digital town square for technology decision makers to gain knowledge about technology and transactions. Learn more