Deep Research
Stop asking AI for answers. Start giving it sources. Use Claude's Deep Research mode to synthesise entire domains in minutes.
- See the difference between a quick AI answer and a cited AI insight
- Learn the five-step anatomy of a deep research workflow
- Practice uploading sources to NotebookLM and generating outputs
- Build a habit of citing your AI-generated content at work
Run a real deep research workflow using NotebookLM. By the end, you will have a cited insight from a credible document that you can use in your actual work.
Choose a research topic
Pick something directly relevant to your current work — a decision you're making, a trend you need to understand, or a problem you're solving.
Find one credible source
An academic paper, industry report, or validated dataset. Not a blog post or news article. Try MIT Sloan, Stanford HAI, McKinsey Global Institute, or Google Scholar.
Upload to NotebookLM
Go to notebooklm.google.com, create a notebook, and upload your source as a link, PDF, or Google Drive document.
Create a mind map or audio overview
Use NotebookLM's built-in tools to understand the material more deeply than reading a summary alone would allow.
Pull your key insight
Identify the single most important finding relevant to your work. Be specific — not "AI is growing" but a data point that could change a real decision.
Submit to your AI Buddy
Head to AIO Labs and submit your work for grading and certification credit.
- Your source — title, author or organisation, and a link
- Your key insight — one specific finding in plain language
- One sentence — what does this change about how you work?
Source Credibility
Named author or institution, directly accessible. Academic or validated industry research — not a blog or AI summary.
Insight Specificity
A number, comparison, or concrete finding — not a general statement that could have come from anywhere.
Relevance to Work
Names a specific decision, behaviour, or challenge the insight affects — not abstract or hypothetical.
