The range of AI solutions available today
So, you’ve been mesmerized by the news about the potentials of Artificial Intelligence or you’ve somewhere already known about how it will transform your business to deliver higher productivity. We cannot deny the fact that everyone has an opinion when it comes to AI, it has quickly become the crowned poster child of Buzzwords in the last 5 years. Interestingly, it has come to my notice that there has been a surprising amount of companies that have come up in the recent 2 years that claim to provide AI solutions. Even if the solutions are Robotic Process Automation (RPA) driven or a proposed extension of the definition.
The reality is that almost 95% of the solution providers today are not exactly selling the AI that most businesses need. It is at the best driven by machine learning applications and the closest it can get to is with Automated Reasoning. But the fact remains: businesses that fail to adopt even the available technology today might face various challenges when AI advances in the future, which it will.
So, to help you familiarize with the situation, let me give you picture of the AI technology levels.
The first level begins with Robotic Process Automation (RPA), this is a term that has been widely seen in the news these days. There have been quite some established players in this market. They have set a firm foundation when it comes to enterprises aiming for digital transformations. This technology, to put it in a simple way helps automate manual processes. Especially, the ones that follow a rule-based workflow approach. In other words, any repeatable and rule-based task is suitable for RPA.
The next level sees machine learning driven applications. They make use of standard algorithms that has been modeled to produce highly probable and accurate outputs. So, the solution providers who have better platforms are perhaps utilizing this method to structure the rule-based approach of workflow engines to generate a considerably better outcome. Apparently, the combination of RPA & machine learning turns out to be an interesting approach in certain cases, wherein the platform is as good as an intelligent system. This is what has been mostly positioned as the AI solution these days. But then again, it is not the true artificial intelligence, it can be viewed more of an assistance to get the job done in a better way. It still requires human intervention to completely finish the task and validate it.
Then comes, Automated Reasoning where computer programs are designed to reason completely or nearly completely, automatically. It requires logical rules, predictive models and is able to compute at a level which can beat even an expert system at times.
Then we reach the level of AI that is available today, which is a smart integration of all the technologies mentioned above — RPA, machine learning and automated reasoning. It helps organizations in their quest to automate basic low-value tasks but not the ones which are too strategic in nature.
Finally, the level which everyone actually wants is the true representation of AI, that is cognitive. This is the level that assists organizations to learn from insights and arrive at key decisions to drive the organization growth in the real sense. It has a human level understanding of going through the tasks. The Cognition element at this technological level has a better understanding of unstructured and dark data. It has the knowledge to classify all the information and arrive at a structured output. At this level, there are a handful of companies operating, who are in the process of making it accessible for various business cases.
So, these are the AI technology levels we are dealing with and it helps in establishing the picture as to which level you are in currently and where do you intend to be ultimately. Things are definitely evolving, keep an eye for the players who are working in the domain of cognition for a better understanding.
*Some content from Cere Labs