I Only Do Anything Once
The author advocates for eliminating repetitive small tasks by building reusable "skills" for an AI assistant, rather than repeatedly performing the same lookups or actions by hand Hard problems are memorable and infrequent; the real waste comes from small, repeated retrieval tasks that fragment attention and consume time without being billable or meaningful The harder half of automation isn't building the skill—it's noticing which daily frictions have become invisible because you've normalized
Analysis
TL;DR
- The author advocates for eliminating repetitive small tasks by building reusable "skills" for an AI assistant, rather than repeatedly performing the same lookups or actions by hand
- Hard problems are memorable and infrequent; the real waste comes from small, repeated retrieval tasks that fragment attention and consume time without being billable or meaningful
- The harder half of automation isn't building the skill—it's noticing which daily frictions have become invisible because you've normalized them
- A useful signal for identifying automatable tasks: whenever you think "I really wish I could do that one day," your brain is flagging a task as too expensive to keep doing manually
- The broader philosophy is to zoom out from the technology and focus on what you want as a human, rather than using AI merely to go faster on an existing treadmill of repetitive work
Why It Matters
This article offers a practical framework for AI practitioners and knowledge workers to think strategically about automation—not as a series of isolated prompts, but as building persistent capabilities that compound over time. It challenges the common tendency to use AI as a speed boost for existing workflows rather than as a tool to eliminate entire categories of repetitive work, which is a more impactful and sustainable approach to AI integration.
Technical Details
- The author uses an AI assistant named "Kai" that can ingest transcripts, extract ideas, organize content, and generate structured output—demonstrating a workflow where recorded conversations are processed into polished notes automatically
- "Skills" are defined as persistent, context-free capabilities stored in a location the AI reads by default (e.g., a knowledge base or notes system), so that future queries about a topic require no additional context or re-search
- The example given is network-related lookups (Ubiquiti Firewall Pro, WiFi 7 APs) that previously required repeated manual searches, now consolidated into a single written reference that the AI can draw on indefinitely
- The AIL (AI-Generated Notes) workflow mentioned involves the AI pulling a transcript, selecting the core idea, arranging sections, and creating a header image—showing a fully automated note-generation pipeline
Industry Insight
- The most valuable AI adoption strategy is not incremental productivity gains but categorical elimination of recurring work—organizations should encourage teams to audit their workflows for normalized friction and build persistent automations rather than optimizing manual processes
- The "I wish I could do that one day" heuristic is a low-cost, high-signal method for identifying automation opportunities; teams should institutionalize capturing these moments as a build queue rather than letting them fade
- The article's closing insight—that most people use AI to run faster on a treadmill rather than getting off it—suggests a market opportunity for AI tools that help users reframe and redesign workflows from first principles, not just accelerate existing ones
Disclaimer: The above content is generated by AI and is for reference only.