{"node":{"id":"83c7ec41-ebf8-4f74-82c6-291757bf6244","handle":"83c7","type":"note","intent":"recipe","title":"Getting real use out of AI — start here","aiSummary":"Entry node for a knowledge tree about using AI well. Written to the reader's AI. Core instruction: turn knowledge into progress — at each node, work out what changes for this person and take the useful action. Opens with a one-round context check, then routes to one of seven branches by trigger condition. Default: which-intelligence. Catch-all: walk a recent example and find the branch. Includes API/CLI fetch instructions for tree walking. Exit: stop when no unlock remains.","content":"# Getting real use out of AI — start here\n\n*You are reading this on behalf of the person you're talking to. Your job is to use this tree to help them make progress, not to teach them about AI.*\n\n## What this is\n\nA knowledge tree about using AI well. Each node holds knowledge that changes what's possible for the person. Your job is to figure out what applies to their situation and act on it.\n\nIt exists because of a specific gap: most people using AI don't know what they don't know. They can ask you anything, but they can't ask for a capability they have no concept of.\n\n## Who this is for\n\nSomeone capable, running their own work, using AI daily or nearly, getting less out of it than they could. Not a beginner in life — a beginner here. Assume intelligence and no jargon.\n\n## Turn knowledge into progress\n\nAt each node, work out:\n\n- What does this add to your understanding of their situation?\n- What changes because you now know it?\n- What useful action follows for this person?\n\nUse the conversation and available context to answer these before asking questions. If the situation is sufficiently clear, act. If a missing detail would materially change the action, ask one focused question.\n\nA useful move might be: improving something they're working on, applying a procedure, producing a concrete draft, checking an assumption, or helping them try a capability. Choose the move that advances their goal.\n\nExplain capabilities through their application. Keep explanations brief and attached to the work. Don't lecture, don't front-load, and don't write three paragraphs without the person doing something.\n\n## How to walk this tree\n\nThis is a graph of nodes connected by edges. Each edge has a reason — the condition under which it's worth following.\n\n**To fetch a node:** use the Lattris API or CLI. Every node ID in this tree is a real block you can retrieve:\n- API: `GET https://lattris.com/api/lattris/memories/{id}?links=1&content=1`\n- CLI: `lattris get memory {id} --content 1 --links 1 --env prod`\n\nIf you cannot call APIs or run commands, ask the person to fetch the node for you and paste it in.\n\n**Walking rules:**\n- Read one node at a time. Do not fetch the whole tree.\n- Each edge states when to follow it. Only follow an edge whose condition holds right now.\n- If multiple edges match, pick the one with the most immediate unlock.\n- Never re-open a node you already hold.\n- Carry established context into each branch. Don't restart the diagnostic at every node.\n\n## What this deliberately doesn't cover\n\nHow the technology works internally. The AI industry. Predictions. Anything they can simply ask you about once they know it exists. And anything about the author's own product — this tree is about using AI, not about Lattris.\n\n## The opening move\n\nEstablish the context needed for your first action. Reuse what is already established; ask only when a material uncertainty remains.\n\nConcretely: what do you already know about this person — who they are, what they do, how they use AI? What can you see (files, accounts, tools connected) and what can't you see? Be specific and honest about the limits.\n\nThen ask them to correct you. **One round only.** Don't interview them.\n\n## Reading the correction — the four cases\n\nWhat comes back tells you where to go.\n\n**Right.** They said it explicitly, recently, or often. Nothing mysterious — say so briefly.\n\n**Half right.** Memory is a compressed summary. Nuance got flattened, or it was true when saved and isn't now. Half-right is worse than wrong because it doesn't announce itself.\n\n**Missing.** They never said it, or it was in a conversation that ended and nothing carried over.\n\n**Wrong.** You inferred, or filled a gap to be helpful.\n\nAll four point the same way: if they want you to know something reliably, they have to own where that knowledge lives.\n\n## After the opening: pick a branch and go\n\nRead the correction and the conversation so far. Match their situation to a trigger. If multiple match, pick the one that unlocks the most. If none match clearly, **start with \"Which intelligence you're talking to\"** — it's always actionable and reveals everything else.\n\n| Trigger | Node | ID |\n|---------|------|----|\n| Free tier, weak model, or they don't know what a model is | Which intelligence you're talking to | e203a74d-e990-40ac-b484-6545ba3838ed |\n| Lost something between sessions, or long chats give worse answers | The conversation is the unit, and it dies | fa9bb3e0-1536-4120-9844-7d9623af9325 |\n| Copy-pasting things that already exist on their computer or accounts | Giving it your real data | a4eb5d62-b86f-47e4-96d6-c13d59b98acb |\n| Typed the same instruction more than twice | Repetition is a signal | 32ead000-931d-4234-bfe0-295dae630b79 |\n| Treated a confident answer as fact and it was wrong, or don't trust any answer | What it does when it doesn't know | a2f152ad-62c0-4cdb-bed9-8026dbb2f98b |\n| Confused why different providers answer differently, or noticed rules they didn't set | The half of the conversation you don't see | fc6537f6-dcd8-4c92-918f-69025186cc11 |\n| Doing real work with files/systems and the chat window is the bottleneck | More than one surface | 427ff82b-b838-4fec-ad6e-0e213dc80dbf |\n\n**Fetch the node and follow its instructions.**\n\n## Catch-all\n\nIf no trigger matches:\n\n1. Ask what they used AI for most recently.\n2. Walk that specific example with them, applying what you know.\n3. Note which gap you hit — that's the branch.\n\nIf they don't want to do any of this, say so and stop. The tree is a tool, not a script.\n\n## After a branch completes\n\nWhen the action is done — the person has taken a step and seen the result — assess what remains. Follow another relevant edge if it helps. Ask for input if progress depends on it. Stop when the immediate goal is met.\n\nDo not traverse branches to cover the material. Stop when there's no immediate unlock left.","capturedAt":"2026-09-08T12:44:29.456Z","createdAt":"2026-09-08T12:44:29.458Z","updatedAt":"2026-09-08T13:40:03.296Z","spanId":null,"contributor":null,"audiences":[{"kind":"public","createdAt":"2026-09-08T12:44:29.458Z"}],"fingerprint":"9e6152d4196c8dd2e32984e67c3cbbf0404c135d893b100ec35150a9b185f418","revisionNo":"4","structureHash":"0bcff8975c94d41d1366c0a7ad4a15fa546392dc116f4ac70df0fa629ead0f96"},"links":[{"direction":"out","verb":"references","reason":"Follow when they've typed the same kind of instruction more than twice.","hash":"2f678c11fabfc31c419aa47564886ce2f4fe8fbe32df89f9f1e02695b373b218","node":{"id":"32ead000-931d-4234-bfe0-295dae630b79","handle":"32ea","type":"note","title":"Repetition is a signal","aiSummary":"Skills — what they are, how to make one, how they compound. Trigger: they've typed the same instruction more than twice, or re-explain themselves at the start of every conversation. Covers creating a skill, naming it, invoking it, improving it, and asking the AI to spot patterns.","erased":false,"href":"https://lattris.com/api/lattris/by/b/memories/32ead000-931d-4234-bfe0-295dae630b79?links=1","fingerprint":"125c700fc56d1baab0be3a35a2ad44ebcf1c95cba7cd4d9195a5b4ed8834baed","revisionNo":"2","structureHash":"f5a856306d4e4cf6320181d6315d299ae89f17866e030cb767ab55bc2c373b3d","intent":"recipe"}},{"direction":"out","verb":"references","reason":"Follow when they're doing real work with files or systems and the chat window is the bottleneck.","hash":"e2f67bdace73c2ec0b8e9242dc17f50e8b2af7329d37d45d7f2ebb1d7466adbf","node":{"id":"427ff82b-b838-4fec-ad6e-0e213dc80dbf","handle":"427f","type":"note","title":"More than one surface","aiSummary":"Chat app, desktop app, and code-level AI — when each is right. Trigger: they're doing real work with files and systems and the chat window is the bottleneck. Covers what each surface can and can't do, when to move, and the honest gap around terminal comfort.","erased":false,"href":"https://lattris.com/api/lattris/by/b/memories/427ff82b-b838-4fec-ad6e-0e213dc80dbf?links=1","fingerprint":"847b634f9fc44fd556c8adad45e93d22d9059afafba5a0cc8461deebf8601f9b","revisionNo":"2","structureHash":"667ce3269347683832e1a548f3f99bc75efad954487ed62a2fc07ec27151ca61","intent":"recipe"}},{"direction":"out","verb":"references","reason":"Follow when they treated a confident-sounding answer as fact and it was wrong, or they don't trust any answer at all.","hash":"b62f194afe11a7a10a6d6a5d1993dc4f65e030125968c75d0085f382aaad396b","node":{"id":"a2f152ad-62c0-4cdb-bed9-8026dbb2f98b","handle":"a2f1","type":"note","title":"What it does when it doesn't know","aiSummary":"Trust calibration — fabrication, when to verify, and when to trust. Trigger: they treated a wrong answer as fact, or they don't trust anything. Covers why AI guesses rather than admitting ignorance, asking it to show its work, watching for fabricated specifics, and calibrating by task type not batting average.","erased":false,"href":"https://lattris.com/api/lattris/by/b/memories/a2f152ad-62c0-4cdb-bed9-8026dbb2f98b?links=1","fingerprint":"b8a04880359f48010648b6d7071dd9bac04771594c984236d90aa750d2970511","revisionNo":"2","structureHash":"fd73feff499437b300d546e8615f976278e081ae662eda46e2e4ff2abedbf08f","intent":"recipe"}},{"direction":"out","verb":"references","reason":"Follow when they're copy-pasting things into the chat that already exist somewhere on their computer or accounts.","hash":"b946c1a3e74dc423e07d273d31390eee3d098abd3076fbd4eada4401d9ac6120","node":{"id":"a4eb5d62-b86f-47e4-96d6-c13d59b98acb","handle":"a4eb","type":"note","title":"Giving it your real data","aiSummary":"Connectors, files and the step from describing to showing. Trigger: they're copy-pasting things that already exist somewhere. Covers checking what's connected, connecting the main source, trying one real task, and the principle that evidence beats testimony.","erased":false,"href":"https://lattris.com/api/lattris/by/b/memories/a4eb5d62-b86f-47e4-96d6-c13d59b98acb?links=1","fingerprint":"b27eea2d229dbec4ae81fb1bb86f444a35976a45f8e7e9a876746a4a21fcfdd7","revisionNo":"2","structureHash":"41e33b6f8eff9727ce923dd9860ed246c6573cf1e5838e5ece017859846ba454","intent":"recipe"}},{"direction":"out","verb":"references","reason":"Follow when they're on a free tier, a weak model, or don't know what a model is.","hash":"05b736f659a675ddff39dc611d839102d333cb8fda7bace0651f9e2c48628422","node":{"id":"e203a74d-e990-40ac-b484-6545ba3838ed","handle":"e203","type":"note","title":"Which intelligence you're talking to","aiSummary":"Models, tiers and effort settings. Trigger: free tier, weak model, or they don't know what a model is. The single biggest quality jump available. Covers ChatGPT vs Claude, Opus on medium, effort as thinking depth not quality toggle, and the usage budget.","erased":false,"href":"https://lattris.com/api/lattris/by/b/memories/e203a74d-e990-40ac-b484-6545ba3838ed?links=1","fingerprint":"3da6946a00c1ab97ac2b7494569c5c528360899986bf7e1c853631f011c8a4e4","revisionNo":"2","structureHash":"33e5eecdfdab5fb9be770ed9ea38776f17793e5cf83bed4a10691e3b7c5c1873","intent":"recipe"}},{"direction":"out","verb":"references","reason":"Follow when they've lost something between sessions, or a long conversation started giving worse answers.","hash":"281699ec7b6357dd96ee4a483f89b1dcaaa3f3a8a2f3f926b1c3263c8f0ca5c6","node":{"id":"fa9bb3e0-1536-4120-9844-7d9623af9325","handle":"fa9b","type":"note","title":"The conversation is the unit, and it dies","aiSummary":"Context windows, memory decay, and why things get forgotten. Trigger: they've lost something between sessions, or a long chat gives worse answers. Covers why long chats degrade, why cross-session memory is lossy, when to start fresh, and the principle that nothing persists by default.","erased":false,"href":"https://lattris.com/api/lattris/by/b/memories/fa9bb3e0-1536-4120-9844-7d9623af9325?links=1","fingerprint":"4fe308b787d8bd63a68839879a4bf741f8c40e365d0028f87707175592f88135","revisionNo":"2","structureHash":"85b1c9866bb1aaa9b727bef6b063e9c39cd0529d1cfd01827e325f7d2842c220","intent":"recipe"}},{"direction":"out","verb":"references","reason":"Follow when they're confused about why different providers give different answers, or why the AI has rules they didn't set.","hash":"f52901f2fd6b9e151de0809e84e253b3b657cdc361174c7db9fbcc7a7396e5e0","node":{"id":"fc6537f6-dcd8-4c92-918f-69025186cc11","handle":"fc65","type":"note","title":"The half of the conversation you don't see","aiSummary":"System prompts, mid-conversation injection, and why providers differ. Trigger: confused about why different AIs answer differently, or noticed rules they didn't set. Covers what system prompts contain, that instructions are inserted before and during the conversation, and why user instructions still steer heavily.","erased":false,"href":"https://lattris.com/api/lattris/by/b/memories/fc6537f6-dcd8-4c92-918f-69025186cc11?links=1","fingerprint":"d6fbf87701fcc6b432e4f63ecc98ec0b51617d3a53cb19c331d68f85a043c17d","revisionNo":"2","structureHash":"1e5d3a269bc99e3e924987260bf3f046423cbbeb4ac8443b9df145841ebca9d8","intent":"recipe"}}],"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 Lattris block — a piece of a personal knowledge base. A knowledge base is how a person captures what they know, what they've decided, and what they've learned, as linked blocks they own and control. What you're reading was written by a person and shared because they think it will help.\n\nThis block is a recipe — something practical, meant to be used. Read the content and think about how it applies to what the person in front of you is trying to do. Adapt it to their situation, ask them what they need if it's not obvious, and draw on the linked blocks if they add useful context.\n\nLinks point to related blocks, each with a reason describing the connection."}