Agency before automation.
Use AI to extend a decision you can understand—not to make invisible choices on your behalf.
Six original video tutorials · Critical AI literacy
Keep the intelligencehuman.
00
The central question is not whether AI can produce an answer. It is whether the person using it can still explain, evaluate, and take responsibility for the result.
Begin the seriesA position on integration
Ethical AI is not a permission slip.
It is a practice of judgment.
Responsible use begins before the prompt and continues after the output. It asks what should be automated, what must remain human, whose knowledge is represented, what evidence can be trusted, and who remains accountable when something goes wrong.
Evidence before fluency.
Persuasive language is not proof. Trace claims to sources and examine the source itself.
Disclosure before ambiguity.
Know when and how to communicate AI's contribution to the work, especially when trust is at stake.
Privacy before convenience.
Do not trade sensitive information for a faster answer. The prompt is also a data decision.
The practical series
Six tutorials. One standard: remain accountable for the work.
These demonstrations move from authorship and pre-writing to research, source-grounded analysis, professional communication, and presentation design.
01AuthorshipJuly 2024
Use AI after the first act of thought—not instead of it.
Using ChatGPT as a Reviewer Over a Producer
Begin with your own response, then use ChatGPT to test its clarity, structure, and blind spots. The system can become a demanding second reader, but the argument, decisions, and final language should remain yours.
02IdeationJuly 2024
A useful outline should reveal your thinking, not conceal its absence.
Using ChatGPT for the Pre-Writing Stages
Use ChatGPT to surface questions, possible structures, subpoints, and counterarguments before drafting. Treat every outline as provisional: reorganize it around your purpose, evidence, and emerging point of view.
03ResearchJuly 2024
A confident citation is not evidence.
Finding High-Quality Sources Using ChatGPT
AI can accelerate discovery by suggesting search language, authors, and candidate studies. It cannot perform the final act of verification. Trace every source to the original publication and judge its relevance, quality, and context yourself.
04Source-grounded AIOctober 2024
Constrain the system before you trust the summary.
NotebookLM Demonstration
NotebookLM can organize documents, surface themes, answer questions across a selected corpus, and create audio overviews. Its value comes from grounding the interaction in sources you deliberately choose—and continuing to inspect those sources directly.
05CommunicationJuly 2024
Professional polish should not erase personal responsibility.
How to Use ChatGPT to Help You Write Formal Emails
Use ChatGPT to review tone, structure, completeness, and clarity in a formal email. Before sending, restore the context and specificity only you possess, verify every name and detail, and ensure the message still represents your intent.
06PresentationOctober 2024
Automation can accelerate design; it cannot decide what deserves emphasis.
Gamma.AI Demonstration
Gamma can turn documents, PDFs, and web material into a fast presentation draft. Use that speed to create room for deeper editorial work: strengthen the argument, correct simplifications, replace generic language, and cite generated or borrowed material.
Before you submit, send, or present
Five questions that keep responsibility visible.
- 01
Can I explain what I asked the system to do—and why?
- 02
Have I verified every important claim against the original evidence?
- 03
Does the work express my judgment, language, and purpose?
- 04
Have I protected private, personal, and institutionally sensitive information?
- 05
Would I be comfortable disclosing exactly how AI contributed?