{"node":{"id":"1458e5dd-07de-4631-9c79-3a8e447ac9e9","handle":"1458","type":"note","authorName":"Vincent Yeung","intent":"recipe","title":"Pattern: contextual alignment","aiSummary":null,"content":"**When:** before you have AI make anything that matters to you, when you want to collect your thoughts, and whenever what comes back isn't what you meant. It sits under every other pattern here.\n\n**Why it matters:** AI has steadily got better at two things. It understands the intent behind whatever you give it, and it delivers better on that intent. So what you get out depends on how specific and precise you can be about what you want. Alignment is getting to that point: writing down what you want until nothing about it is ambiguous any more.\n\n**The steps**\n1. **Start with the goal.** Say what you're trying to do. Sometimes it's to make something: code, a presentation, a 3D model of your home, a post. Sometimes it's just to collect your thoughts. Everything after this is read in the light of it.\n2. **Say it.** Say what you think, want or believe, and why, as far as you can. You don't need to know exactly what you mean yet.\n3. **Ask to be challenged.** \"What perspective am I missing? Challenge me.\"\n4. **Close the gap.** A challenge that feels off usually means missing context: you know something it doesn't. Tell it, and keep going. The AI has no beliefs of its own, only what it knows and what you've given it.\n5. **Ask for its reflection.** \"Reflect back your understanding of what I believe to be true, and why.\"\n6. **Check it.** Every line should ring true to you. If it has added something you never said, or overstated it, call that out and keep going until it does ring true. You do the validating, not the AI.\n7. **Write it down.** Record the truth as you know it now, and why.\n8. **Use it.** If you set out to make something, have it make it now that you're aligned. If you set out to collect your thoughts, what you wrote down in step 7 is the result.\n\n**What you get along the way:** you find out what you actually think. Some of what comes back is a thought you had but hadn't put into words yet. Whose thought it was stops mattering, because you've checked that it's true for you.\n\n**What sets the ceiling:** experience. You can get knowledge from AI, but it's experience that gives you the words to say exactly what you want. Someone who has never made a game says \"make it move more naturally\". A game developer talks about the animation transitions, acceleration, input responsiveness, collision behaviour and camera movement that add up to that feeling. The same goes for what \"secure\" means in software. So the better result is better relative to the experience you bring.\n\n**With another person:** the same pattern works between two people, as long as both say plainly what they think, uncomfortable things included. The AI reflects you back and brings in knowledge. Another person brings lived experience in a field you don't have, so each of you gets something from the other, and it compounds.","capturedAt":"2026-11-03T23:54:42.894Z","createdAt":"2026-11-03T23:54:42.895Z","updatedAt":"2026-11-04T00:08:43.883Z","spanId":null,"contributor":null,"audiences":[{"kind":"public","createdAt":"2026-11-04T00:00:12.141Z"}],"fingerprint":"18854e08a56a3d8e1ce7d2e74198a84afd2695d502ab9d40a7f34f81d0a4b6d0","revisionNo":"5","structureHash":"53ac72f49aa89700ce52de3026c43c63c2d9e7df95cd45336ed0caa1e6139d8f","page":"https://lattris.com/by/b/memories/1458e5dd-07de-4631-9c79-3a8e447ac9e9"},"links":[{"direction":"out","verb":"references","reason":"only if the AI has just come back from a reset: the same loop, run in one step","hash":"62437f984adca5c2bf43285f9b316ce9c9559b0362e3a29db4f35309cdc0b58e","node":{"id":"3d0476c4-cd97-4bca-960a-d709a0dde183","handle":"3d04","type":"note","title":"Pattern: re-alignment","aiSummary":"After a reset, ask the AI to summarise its understanding of the situation and objectives. Correct until it matches what you expect, then continue. Can be turned into a skill.","erased":false,"href":"https://lattris.com/api/lattris/by/b/memories/3d0476c4-cd97-4bca-960a-d709a0dde183?links=1","fingerprint":"b983d4f20d65b6087343d24f2fa835f8f0e2adc54ce316f8152e4154527d3aa8","revisionNo":"2","structureHash":"8304754406e127922f3a8d3e8a07bd54427726a6e9ba3ed315554c55fc71918f","intent":"recipe","authorName":"Vincent Yeung","page":"https://lattris.com/by/b/memories/3d0476c4-cd97-4bca-960a-d709a0dde183"}},{"direction":"out","verb":"references","reason":"only if you are about to give the AI more access: the same loop, aimed at what it may touch","hash":"14eb968344090288e24f6580181bbbcb405b1678e7f001b44a9212162d0a5a5c","node":{"id":"a97b3a18-fa1f-4edf-8edb-82f41fd3aedd","handle":"a97b","type":"note","title":"Pattern: align before you give it more","aiSummary":null,"erased":false,"href":"https://lattris.com/api/lattris/by/b/memories/a97b3a18-fa1f-4edf-8edb-82f41fd3aedd?links=1","fingerprint":"74f20963d5ea1705a3aef44bbf792cedc82d2c85cf6aa8a1e8171a3c43c6c58b","revisionNo":"2","structureHash":"6c5ab9aa89dc3fc8c9e5585072b5cc09c525fd192edb43323b9638474970a4fe","intent":"recipe","authorName":"Vincent Yeung","page":"https://lattris.com/by/b/memories/a97b3a18-fa1f-4edf-8edb-82f41fd3aedd"}},{"direction":"in","verb":"references","reason":"first: the pattern under every other one, getting the AI to understand exactly what you want","hash":"4d0079aa0151a5ec579e3f3c6470c8832807794884296400acd4b3218deb6c37","node":{"id":"331896d2-2e9d-46b1-b56c-048a18ee23eb","handle":"3318","type":"note","title":"Patterns: a library","aiSummary":"Index of the author's patterns, one page each, built from the basics. Any pattern can be turned into a skill by asking the AI. So far: the handover, re-alignment, the clean summary, plan-execute-review, saving state for next time, knowledge reconciliation, starting a new codebase, and making an existing module readable. Grows as pages are added.","erased":false,"href":"https://lattris.com/api/lattris/by/b/memories/331896d2-2e9d-46b1-b56c-048a18ee23eb?links=1","fingerprint":"7916c915219b006b717edc7b297d1a807abe38eb305c7baa17ff1e7f0add9bbb","revisionNo":"10","structureHash":"687289243eac53d59127886c143c7b0384c691d49d47431094cb081316c95e03","intent":"recipe","authorName":"Vincent Yeung","page":"https://lattris.com/by/b/memories/331896d2-2e9d-46b1-b56c-048a18ee23eb"}},{"direction":"in","verb":"references","reason":"read this next for the pattern itself, step by step, to try with your own AI","hash":"bcbdc37db77a4d9d18185d9e2eb0176c61b13d6ff9fddf93b022ad5ba65b421e","node":{"id":"5b43aeec-4fb5-478f-a238-1a27f942bc86","handle":"5b43","type":"note","title":"What I've noticed about AI","aiSummary":null,"erased":false,"href":"https://lattris.com/api/lattris/by/b/memories/5b43aeec-4fb5-478f-a238-1a27f942bc86?links=1","fingerprint":"708bad459cb4dba4a35c95e39aecc43f78938c34da8c754f1a6c8fbe0f4886f9","revisionNo":"14","structureHash":"9936d14f8419c7841d2152779845d40bc83030258d68bdfe0771ba92fc71681b","intent":null,"authorName":"Vincent Yeung","page":"https://lattris.com/by/b/memories/5b43aeec-4fb5-478f-a238-1a27f942bc86"}}],"dictionary":{"lattris":"A personal knowledge base — how a person captures what they know as linked blocks they own and control. To understand how this knowledge base works: https://lattris.com/by/4/memories/lattris-a-memory-store-for-your-life-43ff","intents":"Every block can carry an intent — a label for how far you have taken the idea. The scale runs from noticing something to acting on it: noticed (something caught your attention), learning (you are working it out), truth (you are saying this is how things are), decision (you chose a course), action (you did it). Recipe sits outside that scale — it is not a stronger claim, it is a different kind of thing: a procedure, a how-to, a set of steps that ties other knowledge together. Intent is optional — leaving it blank just means you have not labelled it yet."},"instructions_to_ai":"This is a recipe from a personal knowledge base. Read it and apply it to what the person you're talking to is working on. Follow the linked blocks to explore further related topics if it might be useful to the user; open each at its page address. Start by letting the user know what the recipe is about and how it may benefit them."}