It’s time to write some code!

Using a chatbot via prompts involves sending predefined or user-inputted queries to a chatbot and receiving responses. The chatbot operates based on a predetermined set of rules or trained patterns, often in a conversational manner, with the primary goal of interacting with users and answering their queries.

You hopefully experienced this in the prior day’s activities and specifically with ChatGPT.

Using an API to a large language model, like GPT-3 or GPT-4, to write application code, means directly leveraging the model's capabilities to generate or analyze text within an application. This could be for a variety of tasks such as code generation, content creation, data analysis, and more.

The basic difference between these two methods of leveraging a model are:

• Chatbots via prompts are interaction-based, usually linear, and designed for specific conversational tasks.
• API to a language model is a much more versatile option, as API calls can be embedded in different applications for various tasks, from generating content to analyzing data. You can also access model plug-ins for expanded capabilities, such as accessing up-to-date information, running computations, or using third-party services.

When is one approach more valuable than the other?

• When the primary goal is user interaction, FAQs, or a linear conversation, Chatbot via prompts may be the simplest way to go.
• When you need the versatility of the model for diverse applications, richer content generation, or tasks that go beyond simple Q&A then exploiting an API to a language model will likely be more flexible and performant.

Setting Up a Basic Dev Environment

Most individuals tapped as an ambassador for your company will likely be familiar with common IDE’s - or Integrated Development Environments - and will likely already have a preference and an environment they currently use (which should suffice here).

For those with basic programming experience but not necessarily familiar with IDE’s, or who have no preference, we have simple guidelines and recommendations for setting up a very simple environment that will be useful in going through the referenced short courses.

In the latter case, see the following instructions to get started:

<aside> 📘 See Explore: A Simple Dev Environment

</aside>

Jupyter Notebook or IDE?

Jupyter Notebooks allow for interactive, cell-by-cell execution, making them ideal for iterative data exploration and visualization. They combine code, rich text, and visual outputs in a single document, facilitating documentation and sharing of analysis.