My 14-year-old saw a news segment about CRISPR curing a kid's sickle cell disease and turned to me with the four most dangerous words a parent can hear at 8pm: 'wait, how does that actually work?' I have a biology degree. It is also twenty years old and was, if I'm honest, mostly about frogs and the Krebs cycle I've since forgotten. I know CRISPR is 'gene scissors' the way I know a car has an engine. Explaining the molecular mechanism — accurately, to a teenager, on a Tuesday, after dinner, with her own homework deadline glaring at us both from the kitchen table — was a different sport entirely, and I was losing badly. My first attempt involved the phrase 'so there's this bacterial immune system' and I watched her eyes begin their slow journey toward her phone.
What I did instead of bluffing
Rather than fake my way through and teach her something subtly wrong, I ran 'CRISPR-Cas9 gene editing' through the Explain Any Scientific Concept With a Concrete Analogy and Its Limits prompt. Audience: 'a curious 14-year-old who knows basic cell biology from school'. Depth: 'explain the mechanism'. Claude Opus 4.8 reached for a word processor, which was almost suspiciously perfect, because this child lives inside Google Docs and has more comments resolved than I do. The genome is the document. The guide RNA is the Find function — you type in the sequence you want to land on, and it scans the whole document for that exact string. Cas9 is the pair of scissors that cuts the document precisely where Find stopped. Then the cell's own repair machinery is the autocorrect that swoops in afterward and either deletes the snipped section or pastes in whatever new text you handed it.
She got it immediately, which was lovely. Then, unprompted, she got the safety problem on her own, which is the exact moment I knew the analogy was doing real teaching and not just sounding nice. She said, 'but what if Find lands on the wrong word?' Reader, that is the entire field of off-target effects, articulated by a ninth grader because the right metaphor handed it to her for free.
The mapping did the teaching, not me
We worked through the mapping table line by line, and each row turned a piece of jargon into something she already understood from a tool she uses every day. I genuinely contributed very little beyond reading aloud and occasionally saying 'right?' in an encouraging voice.
- She figured out why 'off-target' edits are the big worry entirely on her own — it's Find matching the wrong word elsewhere in a long document and the scissors cutting there too, in a place you never meant to touch.
- The 'where it breaks down' section flagged that, unlike autocorrect, the cell's repair is genuinely sloppy and not always faithful to what you intended — which led straight into why editing a blood cell (somatic, affects only that person) is different from editing an embryo (germline, passed to every descendant).
- The Surprise note explained that the edit isn't 'undo-able' — once you cut and the cell repairs, there is no version history, no Ctrl+Z, no document recovery. That landed with real weight for a kid whose entire relationship with text assumes infinite undo.
- It also clarified that CRISPR doesn't 'add good genes' like a magic upgrade; it mostly makes precise cuts and lets the cell's repair do the consequential, error-prone part — which reframed the whole thing for her as 'controlled risk', not 'science fiction button'.
We spent a solid twenty minutes reading the explanation and the mapping table out loud, trading off paragraphs like a bedtime story for nerds. At the end she said, 'so they really did cure actual people of sickle cell with this?' and then — without me asking, without me even suggesting it — she pulled out her own phone and looked up the clinical trials herself. That curiosity cascade, from a headline, to a mechanism she could hold in her head, to a teenager independently fact-checking the news, is the entire reason I keep this prompt around. I didn't make her curious. I just got out of the way of a good explanation.
The one setting that does most of the work
The honest caveat: I picked the audience field carefully, and that's where most of the magic actually lives. When I'd earlier tried a rushed version set to a generic, unspecified audience, the analogy drifted toward concepts a teenager simply doesn't carry around, and the mapping got abstract and bloodless. Tell it exactly who's asking — 'a 14-year-old who lives in Google Docs' produced a completely different and far better explanation than 'a curious person'. When she later wanted to understand the disease itself rather than the editing tool, I just switched the same prompt's inputs to the topic 'why sickle cell mutations deform red blood cells' and we kept going down the rabbit hole together.
If you've got a kid, a student, or a curious partner who keeps lobbing 'how does that work?' at you over dinner, grab the Explain Any Scientific Concept With a Concrete Analogy and Its Limits prompt on Prompt Dock and let it build the bridge for you. Set the audience to exactly who's asking — the right analogy changes completely between a ten-year-old and a colleague from another field, and aiming it precisely is the whole point.