Lesson 07: Where to go from here

The video explains the main idea. Use the sources, instructions and checks below to complete the practice.

Plan learning and authorized onboarding. Build a judgment you can explain and a result another reader can inspect.

Prerequisites: know the difference between generated text and source evidence. For configuration vocabulary, use the AI workflow reference. This is public self-study, not official government training, accreditation, a clearance or a professional credential.

Observable outcomes: Produce a conditional 30/60/90-day onboarding plan with approved-environment gates, recovery evidence and a fallback.

Part 1 Plan learning and authorized onboarding

Choose a path from demonstrable tasks, mission interest, location, work conditions and actual opportunities. Government, contractor, research and small-business paths each vary; a clearance does not guarantee compensation or job security. Keep public portfolio work wholly synthetic or already authorized for release. Record learning evidence instead of predicting an inevitable market trajectory.

In a new role, first establish approved hardware, software, repositories, data, role permissions, incident routes and deployment procedures. Never install a local AI stack on government or employer hardware without authorization. A harmless personal experiment can introduce unapproved code, weights, telemetry or dependencies. Ask for an approved sandbox and use synthetic tests until the work boundary is established.

A 90-day plan should include one skill gap, one reproducible artifact, one feedback loop and a stop or revision trigger. Career choices remain yours; practice requires no accounts, applications, networking messages, paid courses or real work records.

Part 2 Synthetic worked example

Fictional onboarding request: “Download the model to your work laptop and show me something Friday.” Riley requests the approved environment, artifact provenance and software authorization. While waiting, Riley prepares an offline architecture and ten synthetic test cases. This advances the task without bypassing the installation boundary.

Guided practice: write the tempting conclusion first, then underline the fact that does not support it. Rewrite the conclusion with that limitation visible. Identify one additional observation that would change the decision.

Practice exercise · 30 MIN

Independent synthetic practice

Build a 30/60/90-day plan for a fictional applied-engineering role. Include access confirmation, approved sandbox, a failure-recovery test, review criteria and a fallback if the sandbox is unavailable.

Use paper or a local notes file. All names, records and outcomes are invented. No model account, API, installation, paid service or private file is needed. If you optionally use an assistant on these synthetic cases, keep its output as an unverified draft and apply the same review.

Answer feedback and acceptance: Pass when installation and data permissions precede execution, milestones produce inspectable evidence, and the plan can revise after a failed test. Cumulative check: explain why local inference, a cert and a clearance each fail to establish complete authorization.

Retry: explain the failed criterion, repair the artifact and test a new invented case with a different boundary or consequence.

Next step: carry your artifact into the AI workflow reference; retain assumptions and unanswered questions.

Where to go from here

Advance learning while waiting for required tool approval.

A made-up teaching example

Ask yourself

Start with these checks

Where to go from here

0:00 · What you will learn

0:06 · The idea

0:22 · One useful term

0:31 · A worked example

0:48 · Your turn to think

1:05 · Why this answer works

1:13 · Check 1

1:19 · Check 2

1:24 · Check 3

1:32 · Now try a different case

Part 1 Plan learning and authorized onboarding

Part 2 Synthetic worked example