Attachment (1) Content Skeleton for Gen AI Course (Draft).pdf
PDF 91 KB Posted
- Attached to
- Request for Information (RFI) Gen AI Training Federal contract opportunity
- Solicitation number
- A045924
About this file
This is a content skeleton document outlining a 10-session training curriculum for Generative AI courses. The training program, requested by the Department of the Air Force Materiel Command Lifecycle Management Center at Maxwell AFB, consists of 50-minute sessions followed by Q&A periods. Each session builds progressively from basic AI concepts to practical applications.
The curriculum begins with a general AI overview and moves through hands-on experience with tools like NIPRGPT, Gemini, and ChatGPT. Later sessions focus on work-related applications, including data analysis, code creation, and testing. Key features include virtual handouts, practical demonstrations, and emphasis on security considerations for paid accounts. The program incorporates hands-on exercises using tools like Notebook LM, ChatPDF, RAG implementation, and culminates in a three-session coding project. Sessions will be recorded for future use, with initial sessions mandatory PEO-wide, and subsequent sessions optional. The document notes that users will likely need to sign a User Commitment Agreement with supervisor approval for DAF system access.
View the file
Other files for this federal contract opportunity
| File | Type | Posted |
|---|---|---|
| Request for Information (RFI) Gen AI v2.pdf |
On GovTribe
Work with this file on GovTribe
- Download the original file
- Contacts named in this file
- Similar government files
- Ask GovTribe AI about this file
Text version
Content Skeleton for GenAI Course
Assumptions:
- Content simple enough that our mothers, fathers, and kids could understand the content and on their own perform the examples
- Target is 50 minutes of content for each class and remaining time is for answering questions chatted in—this is the standard for every class
- Focus is on engendering fun and curiosity and creativity
- The entire PEO will participate in the first session
- Sessions after the first one will be done PEO-wide, though not mandatory
- The recordings will be available for individual or small group viewing
- Virtual handouts will accompany each class so students can follow along and repeat activities on their own
1. General overview of what is the umbrella called AI with examples in pictures or videos, and they can consist of content that’s free over the internet
a. The start or perhaps first 5-10 minutes should be obvious illustrations in pictures or videos to get the audience’s attention.
b. We should follow with a myriad of different examples before drilling what is Generative AI—and then follow that with still more examples
c. Examples for how people are using it today
d. How they can get an account—just walk through the process
e. Demo using NIPRGPT and Gemini and ChatGPT, paid accounts for the latter two
f. Emphasize only paid accounts can ever get used for things you don’t want the whole world to know
g. Each session will have a handout that is an outline of the class with links and simple directions to accomplish all the exercises in the class
h. Do a few “cool” things that will be in the handout so they can also try it on there own
i. This will be a “group-focused” session
j. Provide the handout before the class so the participants can follow along during the close if they so choose
k. Before going to questions, conclude the first session with the following and paraphrasing or rewriting to fit your taste is acceptable: “The Air Force may provide users of GenAI on DAF systems and data with guardrails. One of the guardrails the DAF will likely use is the signing of User Commitment Agreement, including a supervisor’s signature, where you will commit to follow some simple rules, in much the same way our Robotic Processing Automation programmers commit to always have Functional or Domain leadership permission before they read or write data from any of our Mission systems. We’re still working on the Agreement. However, that shouldn’t stop anyone from experimenting with GenAI or continuing to use it in those Mission areas where it is already approved. For example, like in our on-going use of GenAI in the understanding and conversion of the Stock Control System.”
2. This will be an on-their-own or with co-workers focused session
a. Spend the first 20 minutes covering an overview of previous material
b. Walk them through creating an account for a service over their phone or personal computer—optional whether they actually follow along or just watch—Gemini would be preferred since we’re going to give some of them an Impact Level 4 environment for their program
c. Drop handout into the beginning of the session or provide ahead of time, so people can decide whether they want to do it during or after class, their call—no pressure
d. Walk them through 5 simple exercises they can do with the service offerings—easy and fun examples
3. Ditto 2
a. Spend 10 minutes on the original material
b. Spend 10 minutes on the Session 2 examples
c. Walk them through 3-5 new example uses using five different service offerings:
easy and fun examples
d. The goal of the first three one hour sessions is to engender a desire for more practice exercises and general use of the technology
e. You can also provide them more examples, time-permitting
4. Continuation of three but the examples will transition towards work-type or knowledge work activities but not their actual data
a. Spend 5 minutes reviewing early examples
b. Spend 5 minutes reviewing the Session 3 examples
c. Walk them through 3-5 examples where people are using the tech for work-related activities
d. Use Notebook LM, ChatPDF, and a widely used Service comparable to Notebook
LM
e. Include how they can have documents ingested or sites linked to a Prompt
f. Introduce RAG and begin an evolution of its use across future classes
5. Continuation of four but the examples will transition towards different work-type activities but not their actual data
a. Spend 10 minutes refreshing the Session 4 examples
b. Walk them through 3-5 new uses where people are using the tech for increasingly difficult or complex knowledge worker activities…stick to the previous models used in 4
6. Continuation of Session 5 with new examples
a. Spend 10 minutes reviewing the Session 5 examples
b. Walk them through 3-5 new knowledge worker examples using other models
(continue increasing complexity/difficulty)
7. Continuation of Session 6 with new example tasks
a. Spend 10 minutes reviewing the Session 6 examples
b. Using a spreadsheet of data and a simple Mongo or Postgres database, show them how GPT, Gemini, and NIPRGPT can analyze and manipulate the data in a variety of manners
c. Show how one of the models can create some simple code to analyze the data
d. Using the previous code, use CoPilot and / or Gemini to enhance the code and complete an updated analysis or series of pipelined sub analyses
e. Show how Chat Bots can be created to do simple tests of different aspects of the code
8. Begin a three-session code creation project—simple game(s) is fine
a. Settle on a simple game as your focus and show them how you break it into pieces and you can create the code step by step to complete the game
b. Include the creation of additional code or Chat Bots to Test the code created
c. This can span both Sessions 8 and 9
9. Review previous session for 20 minutes and add features during remaining 30 minutes
a. Finish your targeted capability with the class
b. Show them how you can create chat bots as Test Agents and automatically test the capability—again, we want to keep it simple
c. Tell them to send in examples of their work and we’ll give them two weeks
10. Review the two previous sessions for 20 minutes
a. Cover any new or emerging capabilities with your completed game for 20 minutes
b. Give the top three submissions 5-minutes each for presenting their games
File details come from the government source that posted it. Updated .