# How to Use 48 Hours Before the Hackathon > [!question] Lingering Questions > - Does Lovable's queued-prompt-while-sleeping feature actually work as described, or was that an in-passing claim worth verifying before relying on it? > [!tip] Article Ideas > - Draft a one-page PRD template based on Hong's 5-section structure (problem, user, workflow, features, success) for reuse beyond hackathons. > - Request and document Hong's 6-category prompt library (setup, UI, database, debug, refactor, handover) as a reusable resource. > [!link] Cross Reference > - [[3. Resources/Literature Notes/Prepare for SheBuilds — Before the Hackathon]] — same SheBuilds Season 3 prep series, complementary framework > - [[3. Resources/Permanent Notes/Implementation Is Cheap So Taste Becomes the Bottleneck]] — same bottleneck-relocation argument, independently derived from a beginner's hackathon experience rather than formal product theory > - [[3. Resources/MOCs/Product & Startup Instincts MOC]] > [!quote] Vocabulary > - **PRD (Product Requirements Document)** — here reduced to five sections (problem, user, workflow, features, success) specifically to pre-compute clearer AI prompts, not as bureaucratic overhead ## Author's thesis When AI tools generate code faster than any human can, coding speed stops being the hackathon bottleneck — the scarce resource becomes judgment: defining the right problem, talking to real users, cutting features ruthlessly, and protecting the decision-making capacity (via sleep) that all of that requires. ## Key arguments (≤3) 1. **AI shifted the bottleneck from execution to decision quality.** Hong's own framing, arrived at empirically from a non-technical founder's lived hackathon experience rather than formal product training. 2. **User conversations are asymmetrically cheap relative to what they prevent.** One honest conversation surfaced that migrant workers' real pain was high-stakes life decisions, not budgeting tools — the assumption she'd planned to build around. 3. **Sleep is a competitive decision-quality input, not a tradeable cost.** Since Lovable can queue and execute prompts overnight, the bottleneck under fatigue is exclusively the human founder's judgment, which sleep protects. ## Claim atoms ### AI shifts the hackathon bottleneck from coding to judgment > [!quote] ID:rw_001 > "The biggest challenge was never the code during this buildathon. It was more about making good decisions under time pressure." Where/why: ~4 min in; central thesis of the talk. Reframes preparation strategy for any AI-augmented build context, not just hackathons — applies wherever AI automates execution and elevates human decision-making as the scarce resource. ### Clear PRD thinking directly improves AI-generated output quality > [!quote] ID:rw_002 > "If your thinking is foggy, your prompts will be foggy as well, and your output will be unclear as well." Where/why: ~20 min in, Step 4 (PRD). Relevant to anyone using LLMs for code, content, or product work — the PRD is prompt pre-computation, not bureaucratic overhead. ### One real user conversation prevents 100 hours of wrong-direction building > [!quote] ID:rw_003 > "One honest conversation with a real user can save you a hundred hours of building the wrong thing later on." Where/why: ~15 min in, Step 2 (user interviews). Supported by Hong's own case: she assumed budgeting tools were needed; the real pain was high-stakes life decisions. ## Permanent candidates - **Decision quality is the primary competitive variable in AI-augmented builds** — supported directly by H1; promoted below, connected to the vault's existing taste/bottleneck-relocation thread. ## Processing checklist - [x] Ported existing deep-read (Refactor Appendix) into Konik Literature Note format - [ ] Verify Lovable's queued-overnight-prompt feature status - [x] Connect into the relevant MOC — [[3. Resources/MOCs/Product & Startup Instincts MOC]] ## Metadata - Source: [[3. Resources/Readwise/Articles/How to use 48 hours before the hackathon]] ## Links - [[3. Resources/MOCs/Product & Startup Instincts MOC]] — pre-build judgment discipline - [[3. Resources/MOCs/AI & the Future of Knowledge Work MOC]] — the bottleneck-relocation argument (coding to judgment) is a direct instance of this MOC's core theme