While some bots are nothing more than glorified command languages, others are driven by natural language processing, and the most intelligent bots are in the arena of self-driving cars. The technology behind the types of bots varies greatly in its complexity, application, and maturity level. For the purpose of this landscape, we’ll use the bot classifications provided by botnerds.com: Script Bots, Smart Bots, and Intelligent Agents. Script Bots The simplest form of bots are the Script Bots, which are platform based bots that run off of a pre-determined script that dictates what the bot can and cannot do. Currently, there are six main platforms: Slack, Twitter, FB messenger, KIK, Telegram, SMS. Developers need to make the bots work based on the platform’s capabilities. Script Bots are helpful in transactional and informational requests. Take for example @statsbot in Slack. If your company uses Slack for internal/external communications, what a great way to quickly check website stats, speed, traffic sources and more right inside of the platform. Some Script Bots use Natural Language Processing (NLP) to help identify words that might match the script. Due to technological barriers, this approach often causes a poor and confusing user experience. It is important that companies developing bots understand and respect the limitations of Script Bots + NLP. One brand doing this well is Taco Bell’s food ordering bot integrating within Slack, named TacoBot. This is where most of the Bot Landscape exists today, and will exist in the near future. Much of the media buzz is about Smart Bots, which we will discuss next, but in reality the bot landscape growth will be confined to Script Bots + NLP for several years. For one, these bots are confined to Messaging Platforms. Even with all the buzz, Facebook’s Bot Platform relies on an older NLP technology. When companies respect the limitations, Script Bots prove to be incredibly helpful agents. Smart Bots The next level of bot intelligence is the Smart Bot. A Smart Bot relies on AI with human intervention as a backup layer for complex commands. Think of when you call into your bank. There is a set of automated questions that you have to go through. 10 years ago- those commands were much more limited than they are today, but now we can handle most tasks before talking to a customer service. Intelligent Agents Siri, Chatbots, IBM Watson, Amazon’s Alexa, and Google’s Self-Driving Car are all examples of Intelligent Agents. Powered by Machine Learning (ML) and/or Artificial Intelligence (AI), these bots are designed to run without any human intervention. One Intelligent Agent that is particularly interesting is x.ai, a bot named Amy that schedules meetings with zero oversight (including managing all of the back-and-forth messages). While it is still in its beta stage of development- we have yet to access and try out this seemingly highly useful bot.