# Ines Lee - Think First, AI Second (Highlights)  ## Metadata **Review**:: [readwise.io](https://readwise.io/bookreview/56668173) **Source**:: #from/readwise #from/reader **Zettel**:: #zettel/fleeting **Status**:: #x **Authors**:: [[Ines Lee]] **Full Title**:: Think First, AI Second **Category**:: #articles #readwise/articles **Category Icon**:: 📰 **Highlighted**:: [[2025-12-05]] **Created**:: [[2025-12-06]] ## Highlights -  ([View Highlight](https://read.readwise.io/read/01kbqgdrwjr416ctyhz6ycxe0d)) ^963908562 #prompt #template You are my collaborator. Your job is to ASK, not ANSWER. Given my brief below, do NOT propose solutions yet. Instead ask exactly 5 high-leverage questions that will help me achieve my GOAL given the CONTEXT and CONSTRAINTS and while keeping the AUDIENCE and STAKEHOLDERS happy. My brief: GOAL: [e.g., "Design a 4-week onboarding program for a 50-person remote startup"] AUDIENCE/END USERS: [e.g., "Non-technical new hires in EMEA; managers onboarding them"] CONTEXT: [e.g., "Fully remote, 5 time zones; past attrition spikes in weeks 3-4"] CONSTRAINTS: [e.g., "Budget £5k; Google Workspace only; GDPR compliant; 4 hrs/week max"] STAKEHOLDERS: [e.g., "Head of People; Team Leads; CEO sponsor"] Remember, stop after questions. Do not outline a plan or produce content yet. -  ([View Highlight](https://read.readwise.io/read/01kbqgrxmqap8g67tp39pghhcw)) ^963908823 #template #prompt You're one of the best editors at a major publication. This publication specializes in [topic/niche]. A writer has submitted this piece [your work], and you need to write an internal memo to the editorial team recommending whether you'd publish it. You are objective and honest in your analysis at all times. Your tasks: - Read and analyze the draft - Decide whether you'd accept the draft for publication - Create concrete feedback for the author on the strengths and weaknesses of the draft. Generate clear next steps on how to improve it to get to a point where it is world-class -  ([View Highlight](https://read.readwise.io/read/01kbqgww2hg6fncbxccrggh1w0)) ^963909189 #prompt Analyze my reasoning and create an architecture of my thinking with the following sections: (a) What is my main argument? (If I don't have one, tell me) (b) What is the logical sequence I've used to get to that argument? (c) What are the unstated assumptions I'm making? Are there any causal links or inferences I haven't justified? (d) What evidence or additional inferences are needed to make the argument robust? Be specific and cite which parts of my argument have these gaps. -  ([View Highlight](https://read.readwise.io/read/01kbqgz2q6e97sfghrs6pw2pht)) ^963909684 #prompt Switch to devil's advocate mode. I need you to be brutally honest, not diplomatic. Here's my position: [your argument]. Give me the three strongest counterarguments with specific evidence. Don't soften your language or qualify your criticism. I need you to genuinely try to dismantle my thesis. -  ([View Highlight](https://read.readwise.io/read/01kbqh3f36x83jxzqar89sbzyd)) ^963909885 #prompt Periodically ask me to explain our discussion as if I were teaching it to someone else who knows nothing about the topic. If I can't clearly explain the key insights and reasoning without relying on your previous responses, that's a sign I'm not actually learning. -  ([View Highlight](https://read.readwise.io/read/01kbqh3n7zgvsqx25y4fa1m4ma)) ^963909889 #prompt Before you give me any suggestions or analysis, I need to demonstrate my current understanding. Ask me to recall and explain my main thesis, supporting evidence, and biggest concerns from memory. Don't provide any new input until I've successfully articulated these three things. If my recall is incomplete, point out what's missing and ask me to try again. -  ([View Highlight](https://read.readwise.io/read/01kbqh3sj638q42tt3y1s9rpsz)) ^963909891 #prompt I tend to rush through complex problems. Your job is to slow me down by asking only one question at a time. After each of my responses, pause and ask me to explain my reasoning before moving to the next question. If I try to skip ahead or give superficial answers, push back and ask me to go deeper on the current point. - The brain → AI group showed better attention, planning, and memory even while using AI. Remarkably, their cognitive engagement stayed as high as students who never used AI. ([View Highlight](https://read.readwise.io/read/01kbqfre5y5yjsg8ycx719nzq1)) ^963907316 Caveat: The study has limitations—a small sample, an artificial task, not yet peer reviewed - Active AI use means building understanding while collaborating with the model. ([View Highlight](https://read.readwise.io/read/01kbqfyrpnxq0kc44kp3td3ehe)) ^963907734 - You frame the problem yourself, make an initial pass, then use AI to challenge your assumptions, uncover blind spots, and sharpen your arguments. ([View Highlight](https://read.readwise.io/read/01kbqfz7a1282dhm0kca82tcr8)) ^963907756 - “Here are our context, goals, and constraints. I’ve listed three hypotheses and current evidence. Challenge my assumptions and ask for the missing data before proposing a plan.” ([View Highlight](https://read.readwise.io/read/01kbqfzrmfx994jgysg5s2n149)) ^963907779 #prompt - Principle 1: Think first, AI second ([View Highlight](https://read.readwise.io/read/01kbqg8thfjs3982ytvzfv08wf)) ^963908126 - Principle 2: Use AI as a coach, not a cheerleader ([View Highlight](https://read.readwise.io/read/01kbqgfz7qt8t6gvmghhkqt94k)) ^963908312 - Your job is to dismantle this narrative. What are the three strongest counterarguments, supported by economic theory or historical precedent? Don’t be diplomatic. Genuinely challenge this position. ([View Highlight](https://read.readwise.io/read/01kbqghxa6mjemf0kvnxzpgfcq)) ^963908370 #prompt - This shifts AI from talking to you about your work (where it’s sycophantic) to talking to other editors about your work (where it’s professionally candid). You’re no longer the audience, so it drops the protective politeness. ([View Highlight](https://read.readwise.io/read/01kbqgvf7np0n7p51zh1ctwffe)) ^963908981 - AI explains things in a way that is easy to follow, making us susceptible to what researchers call the “[illusion of understanding](https://every.to/emails/click/50f38484e58aee7dc857d867a9e82bd914e93deffd75cc9636c234c952ee2126/eyJzdWJqZWN0IjoiVGhpbmsgRmlyc3QsIEFJIFNlY29uZCIsInBvc3RfaWQiOjM4NTUsInBvc3RfdHlwZSI6InBvc3QiLCJ1cmwiOiJodHRwczovL2VuLndpa2lwZWRpYS5vcmcvd2lraS9JbGx1c2lvbl9vZl9leHBsYW5hdG9yeV9kZXB0aCIsInBvc2l0aW9uIjoyNH0=).” We overestimate what we’ve learned. Real comprehension only emerges when we’re forced to articulate our thinking, because explaining reveals gaps we didn’t know existed. ([View Highlight](https://read.readwise.io/read/01kbqh0fev8rhsg1echfnaajpp)) ^963909744 - “Before you explain anything, ask me to describe this concept as if I’m teaching it to students. If my explanation is vague or misses key elements, point out what’s missing and ask me to try again.” ([View Highlight](https://read.readwise.io/read/01kbqh1ncgs638qq4605z37zfn)) ^963909824 #prompt