HNK Institute / AI literacy / human knowledge
Human knowledge for new technology.
HNK Institute is a place where students learn to explain AI, data, and technical systems in human language before they are asked to operate them. Responsible AI education here is a weekly practice, not a warning slide.
Curriculum
Six tracks for STEM knowledge practice.
Choose a track by the knowledge gap, not by a product name. Each track on hnk.org is unique to HNK's explanation-first method.
Method
A literacy loop, not a tool tour.
HNK students move through explanation, mapping, limited use, and teach-back so generated fluency cannot impersonate understanding.
Explain
Human account first
Map
Show the holes
Limit
Write the refusals
Use
Tool as object of study
Test
Closed-interface check
Teach
Someone else can mark it
Institute
What HNK is for.
HNK Institute exists because students are meeting generated answers before they have language for evidence, limits, and responsibility. The public promise on hnk.org is narrow on purpose: human knowledge for new technology. That means AI literacy for students is taught as explanation, mapping, and teach-back, not as speed at prompting.
The institute serves secondary students, families who need a way to hear understanding, schools that need markable modules, and mentors who must protect student thinking from fluent machines. It does not serve product theater, certificate shopping, or unverifiable staff mythology. Faculty standards are published as duties: do not ghostwrite, do not complete the map, do not reward invasive personal data, do require limitation notes.
Evidence at HNK is boring in a way families can use. A later reader should find drafts, knowledge maps, source notes, and a teach-back. If those are missing, a beautiful export is not a result. Progress is a clearer sentence the student can say when the laptop is closed.
HNK Institute publishes this page on hnk.org because human-centered AI literacy is easy to fake with fluent tools and hard to own as human knowledge. HNK Institute teaches students to keep human knowledge in front of new technology, so explanation comes before tools. The institute's tagline — human knowledge for new technology — is a working rule here, not a decoration beside the pink mark. Students, families, and schools should read the page as a description of practice: what the student will explain, what artifact will hold that explanation, and what the work will refuse.
This page on human-centered AI literacy is written for secondary students, families, and schools who want responsible AI education rather than product training. It is not written for tool-first workshops that confuse fluency with understanding. If a visitor came to hnk.org looking for a shortcut around thinking, human-centered AI literacy will feel slow on purpose. AI literacy for students is treated as STEM knowledge practice: repeated, reviewable, and independent of whichever interface is fashionable this term.
What human-centered AI literacy demands of a student
What human-centered AI literacy demands of a student sits at the center of this HNK page because speed is not the scarce resource. The student must be able to speak the idea without borrowing a fluent machine voice. Students practicing human-centered AI literacy should write a human-language account of human-centered AI literacy with the interface closed. They should also circle every term in human-centered AI literacy the student cannot define without a slogan. Those two moves keep responsible AI education attached to a real student mind rather than to a transcript.
A reviewer looking at an explanation card for human-centered AI literacy can ask the closed-interface question: can the student still talk? Common failure looks like opening a tool first and writing the explanation of human-centered AI literacy afterward. HNK treats that failure as a literacy gap. The next assignment is not a more impressive tool. It is a clearer human sentence.
- Write a human-language account of human-centered AI literacy with the interface closed.
- Circle every term in human-centered AI literacy the student cannot define without a slogan.
- Name the human decision hidden inside human-centered AI literacy and who is responsible for it.
- List one thing human-centered AI literacy cannot do, even when the output looks fluent.
- Can the student explain the idea without looking at the interface?
Explanation before operation
Under the heading Explanation before operation, HNK asks adults to listen differently. If the interface is required for the student to sound knowledgeable, the knowledge is still on loan. Families do not need to become engineers to review human-centered AI literacy. They need to hear a limitation note and a teach-back. If the student can only perform while a model is completing sentences, the knowledge is still on loan.
Keep a source note that separates a claim from a guess about human-centered AI literacy in the folder. Refuse treating a fluent answer as proof that human-centered AI literacy is understood. Responsible AI education is rehearsed in small refusals: not pasting other people's work, not sending private records, not claiming certainty the map does not support.
- Can the student explain the idea without looking at the interface?
- Which human decision is the student still responsible for?
- What would make this explanation false?
- Which words are decorative rather than known?
- What should not be sent to a model or a public form?
Human artifacts worth keeping
Human artifacts worth keeping is where STEM knowledge practice becomes visible on paper. Reviewers want maps, limitation notes, and teach-backs more than they want exports. The student should produce a source note that separates a claim from a guess about human-centered AI literacy and a revision log showing how the explanation of human-centered AI literacy changed before anyone talks about publishing, showcasing, or applying the work to a larger story.
Mentors are held to a negative duty as much as a positive one. They may not complete the explanation of human-centered AI literacy. They may not replace a weak student sentence with a better generated one. They may ask: What should not be sent to a model or a public form? That question protects the student from fluent machines and from helpful adults.
- Opening a tool first and writing the explanation of human-centered AI literacy afterward.
- Pasting generated text about human-centered AI literacy as if it were student knowledge.
- Treating a fluent answer as proof that human-centered AI literacy is understood.
- Hiding uncertainty so human-centered AI literacy looks finished.
- Collecting screenshots of human-centered AI literacy without a human account of what they show.
Responsible boundaries
Responsible boundaries sits at the center of this HNK page because speed is not the scarce resource. Literacy includes practiced refusal: privacy, honesty about sources, and not using other people's data. Students practicing human-centered AI literacy should list one thing human-centered AI literacy cannot do, even when the output looks fluent. They should also keep a limitation note beside any example of human-centered AI literacy. Those two moves keep responsible AI education attached to a real student mind rather than to a transcript.
A reviewer looking at a source note that separates a claim from a guess about human-centered AI literacy can ask the closed-interface question: can the student still talk? Common failure looks like hiding uncertainty so human-centered AI literacy looks finished. HNK treats that failure as a literacy gap. The next assignment is not a more impressive tool. It is a clearer human sentence.
- List one thing human-centered AI literacy cannot do, even when the output looks fluent.
- Keep a limitation note beside any example of human-centered AI literacy.
- Compare two explanations of human-centered AI literacy and keep the one a classmate could mark.
- Record a question about human-centered AI literacy that still has no honest answer.
- Which words are decorative rather than known?
How reviewers will push back
Under the heading How reviewers will push back, HNK asks adults to listen differently. A reviewer will ask which sentence would be false and which words are still decorative. Families do not need to become engineers to review human-centered AI literacy. They need to hear a limitation note and a teach-back. If the student can only perform while a model is completing sentences, the knowledge is still on loan.
Keep an explanation card for human-centered AI literacy in the folder. Refuse opening a tool first and writing the explanation of human-centered AI literacy afterward. Responsible AI education is rehearsed in small refusals: not pasting other people's work, not sending private records, not claiming certainty the map does not support.
- Can the student explain the idea without looking at the interface?
- Which human decision is the student still responsible for?
- What would make this explanation false?
- Which words are decorative rather than known?
- What should not be sent to a model or a public form?
What this page refuses
What this page refuses is where STEM knowledge practice becomes visible on paper. HNK will not treat a generated paragraph as a student's mind. The student should produce an explanation card for human-centered AI literacy and a responsible-use boundary list for human-centered AI literacy before anyone talks about publishing, showcasing, or applying the work to a larger story.
Mentors are held to a negative duty as much as a positive one. They may not complete the explanation of human-centered AI literacy. They may not replace a weak student sentence with a better generated one. They may ask: What would make this explanation false? That question protects the student from fluent machines and from helpful adults.
- Opening a tool first and writing the explanation of human-centered AI literacy afterward.
- Pasting generated text about human-centered AI literacy as if it were student knowledge.
- Treating a fluent answer as proof that human-centered AI literacy is understood.
- Hiding uncertainty so human-centered AI literacy looks finished.
- Collecting screenshots of human-centered AI literacy without a human account of what they show.
A practical next step
A practical next step sits at the center of this HNK page because speed is not the scarce resource. Write the human account first, then decide whether any tool is even needed. Students practicing human-centered AI literacy should record a question about human-centered AI literacy that still has no honest answer. They should also translate a technical claim about human-centered AI literacy into ordinary speech, then back again. Those two moves keep responsible AI education attached to a real student mind rather than to a transcript.
A reviewer looking at an explanation card for human-centered AI literacy can ask the closed-interface question: can the student still talk? Common failure looks like pasting generated text about human-centered AI literacy as if it were student knowledge. HNK treats that failure as a literacy gap. The next assignment is not a more impressive tool. It is a clearer human sentence.
- Record a question about human-centered AI literacy that still has no honest answer.
- Translate a technical claim about human-centered AI literacy into ordinary speech, then back again.
- Write a human-language account of human-centered AI literacy with the interface closed.
- Circle every term in human-centered AI literacy the student cannot define without a slogan.
- Which human decision is the student still responsible for?
Audiences
Different people need different tests of understanding.
HNK routes visitors by the decision they actually have, not by a single promotional story.
studentsOwn the explanation.If you cannot say it without the interface, you do not own it yet.
familiesListen with the screen dark.A family review is a conversation, not a dashboard.
schoolsMarkable literacy modules.One concept, one artifact, one responsible-use rule.
mentorsCoach the sentence.Do not donate a better generated paragraph.
Questions this institute actually answers
Does HNK teach a specific AI product?
No. HNK Institute teaches AI literacy for students as a human knowledge practice. Product names change. The ability to explain a system, name a limit, and refuse an unsafe use is what hnk.org organizes. A course may use a tool as an example, but the assessed work is the student's explanation, map, and teach-back.
What does AI literacy mean at HNK?
AI literacy means a student can say, in ordinary language, what a system is doing, what data or patterns it depends on, what it cannot know, and what the student remains responsible for. It is not speed at prompting. It is not a certificate title. It is STEM knowledge practice applied to new technology.
Do students need coding experience?
No. Many HNK pathways are explanation-first. Coding can appear later if the knowledge target requires it. Students who already code are still asked to close the editor and teach the idea back. Technical depth should match the student's current language, not the family's ambition.
How does HNK treat generated text in student work?
Generated text is a claim to examine, not a student's voice by default. Responsible AI education at HNK requires students to mark borrowed lines, test them, and keep a human account. Quietly submitting generated work as personal knowledge is treated as an integrity problem.
What does a student actually produce?
Typical artifacts are explanation cards, knowledge maps, limitation notes, source notes, teach-back outlines, and revision logs. Some labs include a small made object, but the object is not accepted without the human account. Families should expect to hear the student speak without the screen.
How can a family review progress without a dashboard?
Ask the student to explain this week's idea with the laptop closed. Ask what the system cannot do. Ask which sentence they revised after feedback. If those answers are thinner than the exported files, the literacy is not yet owned.
Can schools use HNK as classroom modules?
Yes, when the school wants a shared explanation target, a responsible-use boundary, and a markable artifact. HNK is a poor fit for a one-day gadget assembly with no follow-up language. Teachers should be able to mark the work without being software specialists.
What is a knowledge map in this institute?
A knowledge map is a student-drawn picture of what an idea depends on, what is known, what is borrowed, and what is still empty. It is made before heavy tool use so a model cannot invisibly fill the holes with unearned confidence.
Start