We are seeing AI bots in more places than ever, right now it's very common to get a chat bubble in a website and understand you are talking with a bot, this means there is a process behind every message you send and the bot will interact differently according to what you say. The standards behind that process are still being defined but we are getting to a point when we can see those standards in practice just by interacting with different bots around the internet.
Standards also help business as a whole in the way that end-users start expecting specific things for a bot to do, when people already expect certain behavior then we can start working on making that behavior better and enhancing it with better algorithms and new ideas, it's kind of crazy to think but as we as a society define those standards and enable other people to work on making those better, we will generate better and better bots, thus creating a healthy competition on seeking who has the better standards for their bots.
Before we define standards it’s important to understand the definition of a bot, which is basically any software that makes itself seem human, good bots utilize machine learning to make themselves better with time and answers. Bots can come in different forms and they can be divided according to the source material they are working with. If the input we are giving the bot is text then it's a text based bot, if it's images then it's an image based bot, if we talk with the bot using our voice like Siri or Alexa then it's a voice based bot. What makes an AI a bot it’s the philosophy behind how it was built. If the AI wants to interact as if it’s an actual human we can classify it as a bot, thus not every AI can be called a bot as it depends if their makers care about it being able to pass as a human or just want a working AI machine.
Profitability is one of the major reasons for building a bot. Increasing response times or being able to attend multiple people at the same time is a big reason behind creating them. Many companies have begun experimenting with their own custom bots and some of them even open source it, so that the community can learn from the mistakes they made and also feed the bots with more data. Standards allow for certain behavior to be seen and understood by the business owners themselves, with standards we understand what we need to keep an eye on and which processes we should focus on to make them better. Also, it’s great for end-users because they can start expecting certain behaviors or actions to happen when they interact with the bot, making it easier for tech leaders to know where they should focus improvement on.
Let’s take driving as an example, the letters P,R,N,D are a global standard for automatic drivers and when someone gets to a car they know that universally the P means parking, R is reverse, N is neutral, etc. This has helped car manufacturers to work behind this standard in making better ways of interacting with the car, as we can now use screens or even our voice to make a driving mode change. Pedals are also a good example of standards in driving, as we always expect the right one to accelerate and the left one to brake.
Defining standards helps to create a healthier relationship with the end-user and make their interaction as simple as possible. By having standards defined we also create competition which at the end of the day helps the industry as a whole because we end up having better and better bots with time. Standards also gives companies new to the business a guide in where to start and what to track.