Movie Mashup Chatbot

Have you ever wanted to bring your favourite movie character to life and have a conversation? Well, now you can with the help of AI! In this workshop we are going to create a mashup of your favourite movie character into a personal chatbot that you can have a conversation with.

Movie Mashup Chatbot

Program Details

In this workshop, we are going to look into natural language processing or NLP. The way we understand each other and have conversations, AI can also be taught to do the same. AI analyzes the intent of our sentences to understand what is being said. The way AI can be trained to interpret our words and our natural language then make a decision is natural language processing. 

Chatbot conversation between a person and Steamlabs bot

In this activity, you will create a mashup of your favorite movie characters to create a new character. By working with your friends and their mashup, you will create brand new movie scripts. We will use an AI model called ConveRT – conversational Representations from Transformers to choose the best response from your character based on the prompt you provide.

Did you know that ConveRT was trained on 727 million conversations on Reddit to learn how to choose the response to any given prompt? The responses aren’t generated itself, it just chooses the best one from a list of responses. With your movie character mashups, you will choose from over 300,000 lines of dialogue from 9662 movie characters to create your ultimate mashup. ConveRT will then choose the lines to use.

AI can also learn from human texts and responses and mimic a response back to us. This enhances computer to human communication which is what makes this so important currently in our society. Some examples of Natural language Processing that are currently being used are automated voice systems being used by large corporations such as Costco, Amazon, RBC, etc.

 

Chatbot conversation between person and Steamlabs bot

Discord chatbot conversation between a person and Steamlabs bot

Our chat engine API has many robust features, our free tier which can be accessed by everyone offers the response selection mentioned above and will allow for the whole activity to be fully completed. In addition, a Google Colab notebook will also be provided for teachers who want to challenge their students by integrating this Chatbot into Discord.

 

 

 

 

Taking it further!

AI Text Generation and French Translation Access

You can apply to get access to a paid API key with AI text generation (powered by GPT-2) & french translation (powered by Amazon Translate) capabilities, which can be used to enhance students’ learning experiences and give them a more holistic understanding of AI natural language processing. 

AI text generation allows students to explore natural language processing at a deeper level. Opposed to the free tier where the ConveRT AI only selects existing responses from the response list you have chosen. The AI text generation capabilities use GPT-2 AI to work hand in hand with ConveRT. This allows the AI to take the students’ prompt/question into consideration and generate text from scratch to enhance responses. Discussions such as why the AI wrote what it did, and what influenced it can be fostered from such responses. Allowing the student to think about ethical and real-life implications of AI generating text. 

This paid tier can be accessed for FREE for educators who teach a student body where at least 40% of students belong to underserved or underrepresented communities (this includes, but is not limited to Indigenous, black, recent immigrant, low income, racialized, 2SLGBTQ+, and other marginalized communities).   

Our Chat Engine paid tier API key can be applied here: https://docs.google.com/forms/d/1cWJwjehq58Ql7CpzZvKQm5QHolm3OMXscyvfdd-bVXw/edit

 

What's included

Technical Concepts

  • Natural language processing
  • How AI detects intent from speech
  • How AI can be trained to respond back

Ethical Concepts

  • What are some societal impacts of using NLP?
  • Can there be fairness among different diversity groups when using NLP?
  • Could there be any discrimination interpreting different natural languages?

Upcoming dates

Program Name Ages Location Date Time Price All taxes included  
No Program Date Found