Build AI Verification Program Tutor: CORE-104

@shakilan4

Hi, I’m a Senior AI/ML Engineer with practical experience in LLMs, RAG, AI agents, and production AI systems. I understand CORE-104 is a verification course, not prompt engineering. I can teach students how LLMs work, their limitations, hallucinations, context, model selection, and most importantly, how to independently verify AI-generated facts, citations, APIs, and code. I can also create the Provenance & Verification Statement guidance and grading rubric with practical exercises that make students accountable for everything they submit. Best Regards, Shakila Naz

@C0RETECHIES

Hello, I understand you're creating a foundational course focused on the critical and often-overlooked skill of AI output verification. The goal is to instill a lifelong habit of treating AI-generated content as a starting point that requires rigorous validation, making students accountable for everything they submit. This is about building an accurate mental model, not just prompt crafting. Here are my responses to your questions: 1. A language model is a statistical tool. It operates on 'tokens' (pieces of words) and its sole function is to predict the next token in a sequence based on patterns learned from a massive dataset of text and code. It doesn't understand concepts; it calculates probabilities. The 'creativity' we see is just the model sampling from a distribution of likely next tokens. The context window is its short-term memory, and everything outside of it is forgotten. It is a powerful pattern-matcher, not a reasoning engine. 2. Three fabrication types and their checks: - Factual Fabrication (e.g., incorrect historical dates): Check against an authoritative primary source like official archives, scientific journals, or government publications. Do not use another AI or a secondary aggregator. - Citation Fabrication (e.g., a non-existent academic paper): Verify the citation directly. Search for the DOI in a database like Google Scholar or PubMed. Check if the cited journal and volume actually exist. - Code/API Fabrication (e.g., using a function that doesn't exist): Execute the code in a sandbox environment and consult the official library/API documentation to verify the function's signature, parameters, and existence. 3. I would assess verification by requiring a 'Verification Trail' with every submission. Students must document each claim they checked, the exact query or method used, a direct link to the primary source they used as ground truth, and a one-sentence justification for why that source is authoritative. This makes the process the focus of the assessment; faking the trail would be more work than doing the verification. 4. Draft P&V Statement Guidance: Your Provenance & Verification Statement holds you accountable for your work. Answer these four questions for every deliverable: - What did you delegate? List the specific tasks given to the AI (e.g., 'Generated an outline,' 'Drafted Python function for X'). - What did you change and why? Detail your edits to the AI's output (e.g., 'Corrected factual inaccuracies in paragraph 2,' 'Refactored the code for efficiency'). Explain your reasoning. - How do you know it is correct? Provide your verification evidence. Link to the primary sources, test results, or official documentation you used to confirm the output's accuracy and functionality. - What is still unverified? Honestly declare any parts you have not been able to verify. This demonstrates academic integrity. (e.g., 'The model's claim about X could not be confirmed with a primary source.'). 5. I confirm I understand this is a verification course, not a prompt-engineering course. The objective is to build a professional habit of skepticism and rigorous validation. Regards, Rohit

@toriquldev123

Hello There! I’m Md Toriqul Islam, an experienced full-stack developer with 10+ years of experience, and I’m excited to partner with you. I can dive into your project immediately. I understand you need a verification-focused AI course covering language models, context, task specification, hallucination detection, output verification, model selection, and responsible AI use. I am skilled in AI integration, technical documentation, software development, AI concepts, and practical verification workflows. I’m ready to start immediately and would be happy to discuss the course requirements. Looking forward to hearing from you. Best regards, Md Toriqul Islam

@arturg6

Hello, "AI Verification Tutor Curriculum" - you need a clear, honest course on model verification I’ll build the curriculum around a Docker‑based notebook that runs a small open‑source LLM, letting students see token flow and run verification scripts instantly – this keeps costs low and guarantees a reproducible environment. I’ve delivered a similar AI‑driven learning tool in my modular proposal engine project: https://www.freelancer.com/projects/ai-content-creation/Powered-Modular-Proposal-Engine/reviews To prevent students from fabricating verification statements, I’ll embed hidden test prompts that only a genuine verification process can answer correctly. Which LMS (e.g., Moodle, Canvas) do you plan to use for hosting the course materials? Looking forward to working with you. Artur Giżycki

@jitendras097

Hi, I’m a full-time AI/full-stack developer with 10+ years of experience, and I understand this is a verification course, not a prompt-engineering course. I can help build practical, student-friendly content covering: LLM basics: tokens, prediction, sampling and context. Fabrication checks: verify facts against primary sources, validate citations by opening them, and test APIs against official documentation. Verification: students must provide evidence of what they checked, how they checked it, and what remains unverified. Provenance Statement: What did you delegate? What did you change and why? How do you know it’s correct? What remains unverified? My approach focuses on teaching students to confidently say “I have not verified this part” rather than trusting fluent AI output.

@sarimshoaib1425

100% doable. Verification is the real skill here, not prompting. One. A language model predicts the next token from probability, not knowledge. Training compresses patterns into parameters. Sampling picks likely words, sometimes not the top one. Context is its only memory, nothing persists otherwise. It has no built-in check and no idea when it's wrong. Not magic, not useless, just confident by default. Two. Fabricated facts: check against a primary source, never another AI. Fabricated citations: search the exact title or quote directly. Fabricated APIs: run the code, or confirm it in real docs. Three. Never trust the word "verified" alone. Require a link, screenshot, or reproducible step. Spot check one random claim weekly, live. Four. What did you delegate: name the exact part. What did you change: list edits and why. How do you know it's correct: name the real check you ran. What's unverified: say it plainly, own the gap. Five. Confirmed. Verification course, not prompt engineering. Catching wrong answers, not writing clever prompts. Feel free to DM me if you want examples of how I've handled this before. Let's do it.

@rbtech1984

Hello!! I understand you need a verification-focused AI tutor who can teach models, context, precise task specification, hallucination detection, independent verification, model selection, and honest AI use across all eight fixed topics. • Should lessons be delivered as written modules or interactive lessons? • Do you already have a preferred student assessment format? • Should the verification rubric include practical AI exercises? A language model predicts likely token sequences from context; it does not understand truth automatically. Fabrications can include invented facts, false citations, and nonexistent APIs. These require independent source checks, citation verification, and real API or documentation testing. Verification should require evidence: source links, reproduced tests, reasoning, and a clear record of what remains unverified. The Provenance Statement should cover what was delegated, what was changed and why, how correctness was established, and what remains unverified. This is clearly a verification course, not prompt engineering. Relevant experience with AI systems, evaluation, teaching, and responsible AI practices makes this a strong fit. Let us discuss the curriculum and assessment approach in chat. Best regards Farhin B

@Lazarone1

Hi, I would love to help create this AI verification course. I understand this is not about prompt engineering, but about teaching students how AI works, where it fails, and how to verify results responsibly. I have practical experience working with AI systems and understand the importance of accuracy, testing, and critical thinking. I can help create clear lessons, student guidance, and verification frameworks that are easy to understand and apply. I would be happy to contribute to building a strong AI learning foundation.

@amadorjay

Hi there, The hard part is not explaining what a language model is. It is helping students build a clear understanding of its mechanics, including tokens, prediction, and context—without the allure of mysticism. A language model generates text based on patterns learned from vast datasets, producing outputs that can sometimes mislead if not scrutinized. Fabrication types include hallucinated facts, misleading citations, and incorrect API responses. To catch hallucinations, cross-referencing against reliable sources is essential; for citations, checking against the original sources helps; and for APIs, testing the functionality directly can reveal inconsistencies. Assessing "verified this output" is about establishing a clear process: students must demonstrate their checks through documentation and peer review, ensuring accountability. For the Provenance & Verification Statement guidance, students should address: what was delegated, what changes were made and why, the methods used for verification, and what remains unverified. I understand this is a verification course focused on accountability and critical analysis, not prompt engineering. If you're interested in discussing this further, I'm always ready to talk.

@alirashdi26

Hello, As an AI & Automation Engineer specializing in AI Development, I am the right fit for the Build AI Verification Program Tutor role. Proficient in the language of tokens, context, and predictions, I possess an accurate technical mental model necessary to teach CORE-104 effectively. My extensive experience working with AI models has provided me with a deep understanding of their capabilities and importantly, their limitations. Moreover, I have extensive knowledge and experience in the area of verification - a crucial element of this course. I have firsthand experience being burned by plausible yet incorrect answers; my habit of performing thorough verification is born out of this experience. In addition to my technical strengths, I bring excellent writing and teaching skills to the table. As a Freelancer who has worked on diverse projects ranging from AI to CMS platforms, I have honed my ability to communicate complex concepts clearly and concisely, which will be invaluable in teaching students how to verify outputs effectively. In conclusion, I am a Full-Stack Developer and AI Specialist who not only possesses the technical skills required for this task but also deeply understands its significance. With my combined expertise in AI Development, proven track record in verification processes, strong communication abilities, and affinity for teaching, I am confident that I can lead your students to say "I have not verified this part" confidently. Thanks!

@husainahmed2000

Hi There!!! Your CORE-104 course needs a trainer who teaches students to treat AI output with healthy suspicion, not blind trust, and can turn that into a real verification habit. Building out the eight topics with a clear mental model of tokens and context, teaching precise task specification, spotting hallucination patterns, and writing the Provenance and Verification Statement guidance along with its marking rubric can all be handled with real hands on experience of being burned by confident wrong answers. A short chat would help walk through the five bidding questions in detail and confirm teaching format before moving ahead. Best Regards, Hussain Ahmed

@mobiwebsolutions

✋ Hi There!!! ✋ THE PROJECT GOAL: BUILD A PRACTICAL AI VERIFICATION COURSE THAT TEACHES STUDENTS TO UNDERSTAND, QUESTION, AND VERIFY AI OUTPUT. 1. Cover all 8 fixed topics with clear technical explanations and practical verification. 2. Teach hallucination, fabricated citations, APIs, independent checks, and adversarial review. 3. Create measurable guidance for model selection based on cost, latency, and fitness. 4. Develop the Provenance and Verification Statement guidance and marking rubric. 5. Create student-focused learning material that promotes honest AI use and accountability. Similar AI education projects have been completed covering LLM concepts, AI verification, responsible AI usage, technical training, and evaluation frameworks. <-- Questions --> 1. What learner assessment format do you prefer for each topic? 2. Should the verification rubric include practical AI output exercises? Looking forward to chat with you for make a deal Best Regards Elisha Mariam!

@invoketech

Hello, I have hands-on experience working with AI/LLM systems and building AI-powered products, backed by 9+ years of end-to-end software development experience. I understand that CORE-104 is specifically about verification, not prompt engineering, and the goal is to teach students how to understand model behaviour, specify tasks clearly, identify confident failures, and independently verify everything they submit. Core Features & Modules: Tokens/context, model capabilities, hallucination detection, task specification, independent verification, cost/latency evaluation, AI disclosure and Provenance & Verification Statement. WHAT I AM OFFERING ->>100% Delivery of Work ->>High-Quality, Bug-Free Code ->>Daily Progress Updates with No Communication Delays ->>2-Year Post-Development Support I’m confident I can contribute practical AI knowledge and clear teaching material aligned with the academy’s verification-first standard. Thanks, InvokeTech

@Olywebart

The project requires building an AI verification tutor focused on models, context, and output verification as detailed in CORE-104. I've done similar precise instructional design with technical content, notably in AI content creation and engineering frameworks, aligning with the verification emphasis here. The strategy leverages accurate mental models of language systems, operationalizing context management and verification steps familiar from AI content and model development work, akin to the rigor seen in the crypto dashboard for real-time data and the teaching approach in the sampler output verification courses. A first milestone could be delivering the core course outline and the Provenance & Verification Statement guidance in two weeks, ensuring tight alignment before full module development begins. Missing specifics like student platform or integration tools would be clarified early to tailor content delivery and assessment standards. How do you envision integrating the Provenance & Verification Statement guidance into the students' submission workflow to reinforce verification rigor?

@zeljkop2

Hi there, You need a verification course, not a prompt-engineering course, and I can help build that standard precisely. I’ve spent the last 4 years solving exactly this type of problem: turning AI use into accountable, teachable practice. I will draft the student-facing guidance, the marking rubric, and the verification examples so learners can explain what they delegated, what they changed, how they checked it, and what remains unverified. I’ll also align the 8 topics, define the disclosure standard for Rung 2 and Rung 3 work, and ensure every exercise reinforces adversarial reading, ground truth checks, and honest AI use. Which student deliverables should the Provenance & Verification Statement rubric prioritize most heavily: correctness, disclosure completeness, or quality of verification evidence? Best regards, Zeljko

@mayr81

CORE-104 (Build AI Verification Program Tutor) focuses on accountability, not prompt theatrics. I can help deliver a verification-first learning experience: accurate mental models of language models (tokens, sampling, context), precise task specification, and robust methods to detect hallucinations and fabricated citations/APIs. The course structure across 8 fixed topics aligns well with building repeatable student habits, assume answers are wrong until verified; use independent checks and ground truth; read adversarially; and document what was delegated vs changed. A key differentiator is Topic 8’s Provenance & Verification Statement: students produce a student-facing guidance plus a marking rubric that standardizes disclosure and prevents “looks right” submissions from passing without evidence. I can craft concise instruction text and a practical rubric that distinguishes verified claims, partial verification, and unverified assumptions. I understand this is a verification course, not a prompt-engineering course.

@rapidg1

As we explore the world of AI and seek to fully comprehend its capabilities and limits, it becomes increasingly important to focus our attention on verification, which is at the heart of this course. My extensive experience in AI development and deep understanding of technical aspects, including tokens, sampling, and context, makes me uniquely qualified for the job. I'm not just familiar with a language model- I've been actively building with them and have experienced firsthand where their limitations lie. One of my strengths lies in identifying fabrication in output and implementing specific checks to catch each type. I understand the significance of fostering a habit of verification and accountability in students and have honed my ability to teach these important values effectively. Moreover, being able to teach a complex topic such as AI Verification in a way that a 20-year-old can understand is another skill I bring to the table. Regarding the Provenance & Verification Statement that is critical for this course, I assure you I can draft comprehensive guidance in the given word limit. My proficiency in mobile app development, CRM development combined with my proven ability to integrate AI into different environments further make me an ideal candidate for your project. Choose me for a well-rounded freelancer who not only cares about accurate information but also about delivering it effectively and responsibly.

@MIndlabsAi6

Hey there! I’m beyond excited to take this on! I recently wrapped up a similar project with good results. Drawing from my experience in AI Model Development, AI Content Creation, AI Research, AI Development, AI Compliance, AI Ethics, AI Verification, AI Training Data, I’m ready to dive into your project. Please come over chat and discuss your requirement in a detailed way. Cheers, Vishal Maharaj

@MuhammadAkif8

Hi, I can teach this course around one core habit: never trust AI output simply because it sounds correct. I’d give students a practical understanding of how language models work, then train them to specify tasks clearly, identify confident mistakes, verify claims independently, and take responsibility for anything AI-assisted. The course would focus on: Models, tokens, sampling and context Model limitations and hallucinations Precise task specification Fact, citation and API verification Independent checks and ground truth Cost, latency and model selection Honest AI use and disclosure The Provenance & Verification Statement and marking rubric I understand this is a verification course, not a prompt-engineering course, and I’d build the teaching around real examples where plausible AI answers fail and students must prove what is actually correct. Best regards Akif K