More than a chatbot, this AI is an artificial member of staff for your business!
Adam Milton-Barker | Jul 29, 2016 | Artificial Intelligence | 14344
The history of Artificial Intelligence, or AI, goes back to around the 1940's when philosophers began to try to compare the human brain with algorithms. In the 1950's Alan Turing wrote a paper which put the question on whether machines can think, and proposed the Turing Test, which tests the intelligence of a machine by seeing if a human can detect whether or not it is human by its responses. For many years work in artificial intelligence was not really of public knowledge, for most, their only understanding of Artificial Intelligence was what they saw in the movies. Films like The Terminator, Space Odyssey, iRobot, the Matrix and others began to give AI a very bad public image, it is only recently that people are becoming aware of the existence of AI in the world around them, and many are still skeptical about how safe it is. What most people did not know is that Artificial Intelligence was being integrated into the core of many of the online platforms they use daily. Facebook, Google, Microsoft, among others have been key players in integrating the power of AI into their online services for many years.
In 2013, I began taking my first steps into the world of Artificial Intelligence, guided by a close friend and mentor Shawn E.Carter. My first AI programs were developed under the oIsCore platform and included personal assistants developed in C and C++, as well as projects such as JMNN, an Artificial Intelligence Medical Neural Network built in Java. Not long after that TOA, the TechBubble Online Assistant, was created using a markup language known as Artificial Intelligence Markup Language, check out the TechBubble Online Assistant Video Archive for examples of the early day TOA. The role of TOA was to assist both visitors to our clients websites and staff members whilst managing their business.
The way AIML works is you train the interpreter by feeding it questions and statements, and by providing a set of responses that the bot could reply with. The interpreter would then analyse user input and use advanced pattern matching techniques to match the user input with entries in the AIML database. This method meant that you needed a very large database of possible questions and answers, to make the bot seem more human required many many hours of training, and developers would have to think of pretty much every possible combination of ways that a sentence could be said in order for the bot to provide an accurate answer. This issue meant that AIML was very hard to scale and response times would be slowed down as the database size increased.
I started to approach these issues and began to develop my own implementation of AIML known as TAIML, the TechBubble Artificial Intelligence Markup Language. With this new implementation I restructured the AIML files and completely modified the AIML interpreter. I integrated entities that would allow users to identify key features of the sentence and break them down to help with faster and more accurate matching with data in the database. I also implemented intents which allowed for multiple combinations of user input to be connected to the primary query table which saved on storage space and made queries faster. The last main feature I integrated into TAIML was the ability to assign functions to the responses, with the exception of using Javascript in the bot responses, there was no real functionality for integrating custom functions into the responses that could interact with say PHP functions etc, with TAIML I solved this issue.
As time went on it seemed that AIML (Artificial Intelligence Markup Language) was beginning to lose credibility in the AI world. I decided to put the work on TOA to the sideline for a few years as I focused on completing the TechBubble development, IoT and AI platforms. During this time there was a boom of interest in AI, companies like Google, Facebook and Microsoft began open sourcing their software for developing neural networks, cloud platforms such as IBM and API.AI began to offer AI cloud solutions, and machine learning began to become a very popular subject in the development world.
This inspired me to revive the project and over the last few months I decided to focus my attention on our AI platform and began to rebuild TOA using the latest capabilities of Artificial Intelligence. The project has advanced quickly and she is now accessible via the GUI and able to assist businesses to automate their day to day.
Quick Overview Of The TechBubble Technologies EcoSystem
The TechBubble Technologies EcoSystem combines three main groups of technologies, web/mobile/desktop and Facebook development, the Internet of Things and Artificial Intelligence. Businesses can purchase many types of online CMS systems and business administration systems to build a modular online platform, tailor made for their business, and control it all via a central control panel, the TechBubble GUI. Business owners can create teams and assign permissions which allow their staff to carry out their work on the various systems provided by TechBubble.
The Integration Of Artificial Intelligence
Since rebuilding TOA I have been teaching her how to assist users on the TechBubble GUI using machine learning, natural linguistics and facial recognition. Currently businesses can interact with her like they would with any member of staff. Using voice recognition users can speak to TOA and provide her commands to execute functions on the TechBubble GUI, navigate around the systems and even manage their schedule, emails and job tasks. Facial recognition has now been added which allows TOA to identify members of staff and adds an extra layer of security. The GUI has also been integrated with voice synthesis that allows staff members to hear TOA speaking back to them and read out instructions that can be found around the GUI.
How Does This Benefit Businesses?
With her new capabilities TOA becomes an integral part of the work force, able to carry out tasks that free up staff to focus on other areas. For small businesses and self employed she can save on their staffing costs and help them become more efficient. Soon she will have the capability to autonomously carry out tasks without the need to be prompted, she will learn how your business works and with a bit of training, be able to manage jobs off her own back.
What Are The Future Plans For TOA?
Soon TOA will of been taught how to manage every function available on the TechBubble GUI including managing systems such as E-Commerce, sales management and blogs. In addition to being able to help staff members, she is also being integrated into our front end systems such as E-Commerce systems, blogs, photo galleries etc and will provide a new kind of hybrid web application to be known as AI Hybrid Apps. These revolutionary applications will be accessible on the web, mobile devices and inside the Facebook platform as Facebook Applications and Tab Applications. In addition to this she is also being integrated into the TechBubble IoT JumpWay to bring Artificially Intelligent IoT devices and networks.
We are in talks with some of the most well known machine learning providers in the fields of natural linguistics and computer vision and hope to have some very exciting news soon, and the most exciting news, we are now working with John Gotts and Family Office Venture preparing to go through our first round of funding.
Overall, Artificial Intelligence will play a major role in the future of TechBubble Technologies and will be integrated into all of our products and systems. We will be working on machine learning to assist with our research into brain computer interfaces and our brain controlled IoT and robotics as well as many other projects so make sure you keep up to date with what is going on via our social channels.
Please check out the videos attached to this article to see how TOA can assist your business, more can be found on my YouTube. Anyone that is interested in knowing more please feel free to get in touch.