Creating a Chat Bot – The Future of Training

Here @chatsecrets we just published a training course called ‘Marketing & Selling Using Chat‘. In the first post we talked about the course itself with the modules and lectures. The second post was a deep dive into the technology and architecture of the bot and this final post a discussion on the possibilities teaching bots open up for a future of accelerated information exchange and real time learning experiences

As far as I know this is the first time a chat bot has been used in this way and it opens some interesting ideas for the future. Before diving in to that, lets look first at the challenges.

Lets go?

Architecture

The future does not scale if we need to code all the lectures and modules by hand. We need some way of allowing teachers to create lesson plans using tools they are familiar with. We can then automatically have this transformed into a conversation. There has been some work recently on using a spreadsheet to formally specify these interactions. Our next course will likely dive deeper into this.

Proof that a student enrolled in our course and was successful

We need to allow sign up and payments for courses directly in the chat bot. We also would like to be able to automatically detect when a student has completed all the modules and also all of the assignments. The goal is to have our students send solutions to more complex assignments as links to a git repository. The bot then clones this, mounts some specific storage in a docker container and runs a series of tests. Live output is streamed back to the bot.  JavaScript.ninja has a prototype of this working already (the UI is in Russian for now but it shows the main idea)

After this automated grading of the assignments we can then create a certificate of completion. If everything is OK we can add that to the block chain. This then creates a permanent record of completion. The training should not end there though as we can set further goals and follow up with users

Future Possibilities

We can better track goals.

For our in-company training workshop we can end with a promise from the participants for further learning. Being held accountable by another person (in this case, a chat bot) makes participants more likely to achieve their goals. We can send regular reminders, ask about progress, and provide resources for our participants using broadcast messages.

Introducing new concepts.

The limited time we are given in such workshops means we are limited with the amount of valuable content we can teach. With a chat bot extending the conversation we can add new concepts to the learning experience and deliver them a long time after the classroom experience has finished.

The Spacing Effect Hack.

Repeating old lessons right when you are about to forget them is an optimal learning hack. It’s called The Spacing Effect

Using machine learning, our chat bot can learn when students are most likely to forget information and remind them to repeat it

Personality Based Learning

Many clever students struggle with learning in a formal environment. The promise here is that we can provide a massively automated course which at the same time puts the student’s own personality and interests at the core of what they are actually taught. It would mean adapting content to the pace of the learner and means collecting feedback regularly however we are already doing this so we just need to build a usable interface onto this process and the branching logic into the lesson plan.

 

Here at Chat Secrets HQ we hope you have enjoyed this series of posts and hope you will check out one of our courses soon!

Leave a Reply