Summary This video explains how to leverage AI to improve knowledge work by focusing on the input rather than the process. The proposed workflow involves using AI tools to access and process high-quality academic research to gain insights and create actionable plans, distinguishing this approach from relying on average output from general AI models.
Moving Up the Knowledge Work Value Chain
- AI will increase the volume of information, making it harder to find valuable insights.
- Knowledge workers need to move beyond basic information processing.
- The core of knowledge work involves input, processing, output, and feedback for improvement.
- Most people focus on using AI to process information (e.g., writing), leading to average results due to average input.
- The focus should shift to using AI to improve the quality of input.
- Including academic papers and research in the input can significantly improve output quality.
Accessing High-Quality Input with AI
- Traditional sources like HBR, McKinsey Quarterly, and MIT Sloan Review synthesize academic research for professionals.
- Instead of being a reader or writer of such publications, leverage AI to become the "HBR" for your team/organization.
- This involves understanding complex research and translating it into actionable frameworks for your context.
- AI tools now allow individuals to efficiently process academic papers that were previously only accessible to academics and researchers.
Tools for the AI Workflow
- Elicit: Tool for finding academic papers related to a specific question.
- Allows sorting by citation count to identify well-cited and peer-reviewed papers.
- Provides summaries of top papers.
- Links to sources like Semantic Scholar, which may lead to the full paper.
- Google Search: Can be used to find PDF versions of academic papers.
- Notebook LM by Google: Tool for processing uploaded documents (e.g., PDF papers).
- Uses AI only on the uploaded sources, reducing hallucinations.
- Automatically generates a table of contents for documents.
- Suggests questions about the document's content.
- Allows clicking on AI-generated answers to see the source text and citation.
- Can process multiple documents simultaneously, allowing for questions across all sources.
- Claude: Can be used to refine AI-generated plans and frameworks based on the insights from Notebook LM.
- Can help create detailed plans and matrices based on provided information and sources.
Example Workflow: Improving Team Performance
- Starting with a general question like "How to improve team performance at work" usually yields vague results from general AI.
- Using Elicit to find academic papers on the topic (e.g., team design features).
- Using Google to find a PDF version of a relevant paper.
- Uploading the PDF to Notebook LM.
- Using Notebook LM to understand the paper's table of contents and suggested questions.
- Identifying key areas of focus (e.g., task design, leadership) based on highlighted sections.
- Uploading additional relevant papers found via Elicit to Notebook LM.
- Asking Notebook LM questions that synthesize information from multiple sources.
- Copying the synthesized insights from Notebook LM.
- Pasting the insights into Claude and asking for a detailed, actionable plan.
- Refining the plan in Claude (e.g., focusing on 80/20 changes, creating matrices).
- This process yields specific, research-backed recommendations and frameworks, unlike the general answers from basic AI queries.
Future of Knowledge Work and AI
- Optimistic outlook for knowledge work with AI.
- AI can enable more critical thinking, creativity, and associative thought.
- Generalists can thrive in the AI era.
last updated: