EduCon 2027What Does It Mean to Know Something?
What happens to students when some of the work of learning can be handed to a machine?
For educators, some of the most meaningful evidence of learning isn’t a test score or completed assignment. It’s the moment a student makes a connection they couldn’t make before. The sudden “aha.” The first time a student realizes they have something to say. The growing confidence that comes from knowing something well enough to explain it, question it, expand on it, or use it in a new way. Not because the facts were impressive, but because the student built them, piece by piece.
Those moments are often the result of knowledge accumulated, many times slowly and imperfectly. ReadWorks has spent two decades operating on the premise that background knowledge is essential to reading comprehension, and we still believe it matters. But AI is creating a live tension in that work as teachers and students increasingly encounter AI-generated texts and answers.
We’ll explore what happens to those moments when AI can provide information, explanations, connections, and interpretations almost immediately. How is faster access to knowledge changing the experience of reading and learning? What happens to the growth in confidence that we’ve traditionally seen? And as we remove work that may be busywork, what else might we accidentally lose, and what might we do as educators to preserve students’ joy in discovering something new for themselves?
Conversational Practice
How We'll Make This a Conversation No slides beyond a single opening provocation. The room does the thinking.
Open (8 min): We'll share the live version of this tension — including a real, unresolved moment from our own work where we weren't sure what to tell a teacher who asked this question point-blank. Then one prompt to the room: What's something you know that you'd be genuinely sad if a kid never bothered to know, even though a machine could tell them? No discussion yet — just everyone jots one thing down.
Small groups (20 min): Groups of 4–5, organized loosely by role (so a classroom teacher, a librarian, a counselor, and a parent or student end up in the same group where possible) to compare what people wrote and push on each other's answers. Guiding sub-questions on cards: Is that thing valuable because of what it is, or because of what knowing it does to how your brain works? Would your answer change for a six-year-old versus a sixteen-year-old?
Cross-pollination (10 min): Groups post their sharpest disagreement — not their conclusion, their disagreement — on chart paper. Everyone adds a sticky-note reaction.
Reframe (10 min): We'll introduce the counter-pressure we keep running into: that "just look it up" assumes equal, reliable access to the looking-up tools, and that knowledge is also about belonging, culture, and shared reference points, not just retrieval speed. Does that change anyone's answer?
Trust protocol (15 min): In the same small groups, we'll hand out a real AI-generated answer to a genuine, slightly loaded question (something with a contested history or an underrepresented perspective at stake). Groups apply three questions we originally built for evaluating whether a printed nonfiction text was trustworthy enough for a classroom: What signals trustworthiness here? What raises a flag? Whose voice is centered — and whose is missing? The exercise is the point, not the specific answer — we want the room to feel, hands-on, whether those same questions still work when there's no author or publisher to interrogate, only a model.
Whole-group conversation (20 min): Open floor, fishbowl-style if the room is large. We'll ask people to argue from where they actually land, not where they think they're supposed to land — including on whether the trust protocol changed anything for them.
Close (7 min): Everyone writes one sentence: the thing I'm no longer sure about or the thing I'm more sure about than I was 90 minutes ago — and we read a few aloud before we go.
No comments have been posted yet.
Log in to post a comment.