Content

Image models can be used to produce a range of creative assets and can be prompted via text instructions along with context added by the user. Some examples are below

In this exercise, we'll experiment with a range of image generation models and image generation tasks in order to get a sense of which services might be best suited for particular tasks. To begin with, prompt or search for the top models available for the task that support a free tier for use. Then, create accounts on at least 2 services. Options might include:

Select two different categories from above and produce a prompt for generating an image in each. Then, send each prompt to the two services. For example, one can use the prompt below for a flyer about this course:

Generate a flyer for CS 112: Generative AI Studio.  The course covers a range of topics 
such as prompting, context engineering, conversational chatbots, agents, code generation, 
multimedia generation.

and the prompt below for a drawing derived from a photo.

Generate a simple pencil drawing of the attached image.

Image models can be used to produce data and information visualizations given a prompt and the desired information. Some examples are below

Select two different categories from above and produce a prompt for generating a visualization in each. Then, send each prompt to two different services. For example, one can use the prompt below for a histogram:

Generate a histogram plot of the word occurrences in the Declaration of Independence

and the prompt below for plotting data from a CSV file temps.csv in this week's Google Drive folder.

Generate a line graph of the temperatures in the attached file.

Voice generation models can realistically generate human voices for a range of tasks. A range of creative assets can be generated via text prompts that can also include context added by the user. Some examples are below

In order to get a sense of which services might be best suited for voice generation, prompt or search for the top voice generation services available that support a free tier for use. Then, create accounts on at least 2 services. Options might include:

Select two different categories from above and produce a prompt for generating speech in each. Then, send each prompt to two different services. For example, one can attach the syllabus for the course Syllabus.pdf in the Google Drive folder for this week and prompt a service for a summary of the course:

Summarize what CS 112: Generative AI Studio is about in 30 seconds.  I've attached 
the syllabus.

or one can utilize the prompt below for a voice clone.

Utilizing the voice in the audio attached, summarize what CS 112: Generative AI Studio 
is about in 30 seconds

Music and sound models can be used to produce a range of creative assets and can be prompted via text instructions. Some examples are below

In order to get a sense of which services might be best suited for particular tasks. To begin with, prompt or search for the top image generation models available that support a free tier for use. Then, create accounts on at least 2 services. Options might include:

Select two different categories from above and produce a prompt for generating music in each. Then, send each prompt to two different services. For example, one could ask for a jingle to advertise the course:

Create a 30-second jingle about CS 112: Generative AI Studio based on its attached
syllabus

or one can utilize the prompt below for a cello instrumental.

Create mood music using a cello

Video models can also handle a range of useful tasks. Some examples might include:

In order to get a sense of which services might be best suited for particular tasks. To begin with, prompt or search for the top video editing sites available that support a free tier for use. Then, create an account on a service of your choice. Options might include:

Create a short video clip from an existing source video using the video service you've chosen. For example, one can use the prompt below for a video clip of a basketball blooper that is suitable for a YouTube short that is taken from the video file SportsBloopers2016.mp4 in this week's Google Drive folder:

Find a basketball blooper in the attached video, clip it to a video of less than 
5 
seconds, and then overlay a funny caption on top of it.

There are many different models for producing video content as well. Some examples include:

In order to get a sense of which services might be best suited for particular tasks. To begin with, prompt or search for the top video generating sites available that support a free tier for use. Then, create an account on a service of your choice. Options might include:

Create a video clip using the video service you've chosen. Example prompts might include

Turn this article into a 30-second explanatory video: <Article URL>

or

Create a 30 second rap about a class on Generative AI that includes labs for students 
to practice prompts, context engineering, chatbots, agents, coding agents, image generation, 
audio generation, and video generation.

Note that you may exceed the free tier limits if you ask for anything beyond 30 seconds. If so, ask the service to return what has been completed. Then,

Build-your-own product marketing campaign

For your homework, you will create a marketing campaign for a fictional company and product utilizing any or all of the techniques in this week's labs. For example, you can prompt Generative AI for:

Example product ideas might include

Rubric

We will be using the following rubric to evaluate your homework.

Instructions followed properly including length of screencast and video camera initially turned on

Walkthrough of products produced via Generative AI

Walkthrough of prompts utilized

Submission

Upload your completed screencast on MediaSpace. Ensure that it is published as "Unlisted". Then, in Canvas, submit the URL that your unlisted screencast on MediaSpace is located.