Using AI in Rise or Storyline

In my opinion, the best option for using AI in such cases is to have your own server with open-source AI, such as Falcon GPT or Llama (regarding Llama, it’s important to clarify its legal use... but Falcon can definitely be used legally in these cases). The company will control all its expenses. You pay once for the server and then have a very low maintenance cost. It’s easy to plan expenses. Plus, as a bonus, you can set up the server to work securely with courses created on the frontend (of course, this can also be done with a server that stores the API key for OpenAI).

The idea of having your own server will definitely be appealing to the information security department since all data and processes will be contained within the company’s network and fully controlled by the security staff.

Hello! Integrating AI into educational content is definitely the future of eLearning. I had a similar case where I developed a dialogue simulator for selling mobile phones. The salesperson interacted with a customer, whose role was played by AI. It works really great, and such courses are much more productive than the traditional approaches we’re used to.

Integrating AI into any product is not that difficult. OpenAI provides access to its language model through an API. The code takes literally a dozen lines, and even someone without programming skills can understand it (essentially, you can simply ask AI to write this code for you, and it will all work).

However, this approach has a number of limitations that need to be considered when developing educational content:

1. The OpenAI API is paid. Businesses are used to paying once for the development of e-learning content and then using it as much as they want. Here, the approach is different. Each time we use AI, money will be deducted from the company's account. In most companies, especially large ones, there are no business processes that operate under this model. Additionally, budget forecasting is quite a challenging task (especially when launching such training for the first time). We don't know how many attempts it will take students to complete our simulator, how many mistakes they will make... or the length of their dialogue (the longer the message we send, the more expensive it becomes).
2. To use the OpenAI API, it's essential to work with a server (i.e., part of the code needs to be done on the backend). Currently, 99% of courses are developed exclusively on the frontend (Articulate is no exception). Why is backend needed when working with AI? When we query the AI, we use an API Key (you can think of it as your password for the service. You approach the service, indicate who you are, and confirm it with a password. After that, charges will be incurred for the service). If you add your API Key directly into the course created in Storyline, anyone can see it simply by opening the code of your course. Malicious individuals will certainly see your password and steal all your money, no doubt about it.

In modern business processes, it's very challenging to use the backend in companies because, for decades, there has been no need for server infrastructure in course development. Businesses are still comfortable with things working as they currently do.

Conclusion: It's not that difficult to develop such a product, but rather challenging to implement such a project in companies due to bureaucratic processes and the high inertia of large organizations.

P.S. To use AI in Storyline, you need to use JS, and to use AI in Rise, you need to embed a Storyline block into it. Currently, there is no such option in Rise itself.