https://workyoucanexplain.com/tags/ai-verification/ Recent content in AI Verification on Work You Can Explain Hugo en-US Copyright © Work You Can Explain Tue, 08 Sep 2026 00:00:00 -0500 https://workyoucanexplain.com/posts/verify-a-citation-before-you-use-it/ Tue, 08 Sep 2026 00:00:00 -0500 https://workyoucanexplain.com/posts/verify-a-citation-before-you-use-it/ <p>A citation is usable only when you can identify the work it names. Match the reference against an authoritative publication or library record, open the matching work, and check which version you have. Resolve discrepancies before copying the citation into an assignment.</p> <p>This establishes identity. It does not establish that the work is reliable or supports your sentence. Keep those judgments separate.</p> <h2 id="compare-the-reference-with-the-actual-work">Compare the reference with the actual work</h2> <p>Begin with the reference as you received it, whether it came from a class reading, an AI answer, a reference manager, or your own notes. Search the exact title through a publisher, library catalog, or scholarly index. A DOI, when supplied, gives another route to a record.</p> https://workyoucanexplain.com/posts/check-whether-ai-is-replacing-your-learning/ Thu, 03 Sep 2026 00:00:00 -0500 https://workyoucanexplain.com/posts/check-whether-ai-is-replacing-your-learning/ <p>Finishing a problem with AI does not by itself show that you learned the method. After the session, close the chat and test four things: can you explain the method, reproduce the key step, transfer it to a different example, and verify the result?</p> <p>The check is diagnostic. A failed step is not a verdict about your ability or about AI in general. It tells you which part of the learning process still needs work.</p> https://workyoucanexplain.com/posts/make-templates-easier-with-guided-tools/ Tue, 28 Jul 2026 00:00:00 -0500 https://workyoucanexplain.com/posts/make-templates-easier-with-guided-tools/ <p>A complicated template does not have to be completed as one long writing task. Open a dedicated chatbot thread, paste the template, provide only safe relevant context, and ask the chatbot to collect the information one field at a time. After you review the answers, have it produce the complete document in the required format.</p> <p>The chatbot may extract, clarify, organize, and format. It must not invent an answer or make the final decision for you.</p> https://workyoucanexplain.com/posts/test-ai-generated-claims-with-the-scientific-method/ Sun, 26 Jul 2026 00:00:00 -0500 https://workyoucanexplain.com/posts/test-ai-generated-claims-with-the-scientific-method/ <p>An AI-generated claim becomes useful knowledge only after the evidence supports it. Turn the claim into a testable statement, make its assumptions explicit, decide what observation could contradict it, and choose an independent check appropriate to the decision.</p> <p>The goal is not to prove that an AI system is generally reliable or unreliable. The goal is to reach a bounded conclusion about one material claim.</p> <h2 id="begin-with-the-decision">Begin with the decision</h2> <p>Verification effort should match the consequence of being wrong.</p> https://workyoucanexplain.com/posts/ai-can-be-confident-and-wrong/ Fri, 24 Jul 2026 00:00:00 -0500 https://workyoucanexplain.com/posts/ai-can-be-confident-and-wrong/ <p>AI confidence is a feature of presentation, not evidence of correctness. Generated output can contain false citations, nonexistent APIs, code that works only for the example, hidden assumptions, or unnecessary complexity.</p> <p>Break the response into claims. Verify factual claims in primary sources and technical behavior with official documentation and reproducible tests. Mark unsupported claims as uncertain or reject them.</p> <h2 id="confabulation-can-sound-specific">Confabulation can sound specific</h2> <p>NIST&rsquo;s <a href="https://www.nist.gov/publications/artificial-intelligence-risk-management-framework-generative-artificial-intelligence">Generative AI Profile</a> uses <strong>confabulation</strong> for erroneous or false content that a generative AI system presents confidently. The report notes that output can also contradict the prompt or earlier output.</p> https://workyoucanexplain.com/posts/how-to-review-ai-generated-code/ Fri, 24 Jul 2026 00:00:00 -0500 https://workyoucanexplain.com/posts/how-to-review-ai-generated-code/ <p>Treat AI-generated code as an untrusted proposal. Read every line before execution, compare it with the specification, identify inputs, outputs, assumptions, dependencies, and effects, then test expected, boundary, and invalid cases.</p> <p>Do not accept code you cannot explain. Passing one sample is evidence about one sample, not proof of general correctness or safety.</p> <h2 id="confirm-that-generated-code-is-allowed">Confirm that generated code is allowed</h2> <p>Course policies can distinguish among explanation, completion, code generation, review, and disclosure. Before generating or using code, determine:</p> https://workyoucanexplain.com/posts/never-blindly-paste-a-command-from-the-internet-or-an-ai/ Fri, 24 Jul 2026 00:00:00 -0500 https://workyoucanexplain.com/posts/never-blindly-paste-a-command-from-the-internet-or-an-ai/ <p>Never run an unfamiliar command merely because a webpage, classmate, or AI assistant presents it confidently. Before execution, identify every segment, input, target, privilege, network action, and possible secret exposure. Verify the behavior in official documentation and define how you will inspect the result.</p> <p>If you cannot explain a segment, do not run the command yet. Ask for an explanation, not reassurance.</p> <h2 id="treat-a-command-line-as-a-small-program">Treat a command line as a small program</h2> <p>A line can contain more than one operation. Shell operators may:</p>