How to use Claude Fable 5.1 low effort for knowledge work

TL;DR

Fable 5.1 on low effort can complete a seven sheet workbook and a 13 slide deck for GoPro deal analysis, while extra effort surfaces additional questions that could change the investment decision. Soul improves verifiability with sources and checks. The film creation is a useful side effect for sharing results, but careful review remains essential.

Transcript

Fable 5.1 built this 37second architectural film entirely from code. I didn't know how to use Blender. It used a tool called Blender. I still don't know how it works. All after I gave it only a single property address in Seattle. Then I put the same model on low, its cheapest and fastest setting, and I gave it a real acquisition assignment for inve... Read More

Key Insights

• Fable 5.1 on low effort completes a seven sheet workbook and a 13 slide deck, providing a solid first draft for a real deal analysis.
• Extra effort surfaces questions that can flip the investment call by adding uncertainty, funding needs, and more sophisticated metrics.
• Soul emphasizes verifiability by including a dedicated sources sheet and a clear check sheet, making it easier for another analyst to audit the work.
• Comparing models shows that low effort is fast but incomplete while extra levels add depth and explicit assumptions.
• The example GoPro and Starman deal demonstrates how a model should handle missing information and still produce actionable insight.
• The base case valuation and bull cases illustrate how small changes in input or approach can shift final recommendations.
• Writing prompts to test models reveal differences in causal reasoning and factual clarity across tools.
• The video argues for a workflow where drafts are iterated between models and human oversight rather than a single final output.
• More videos from this channel:
  • How can AI scale to AGI without full context?
  • How to Run AI Locally on Sensitive Files You Can't Upload
  • How to remember AI model differences with semantic cards
• More from this channel: AI News & Strategy Daily | Nate B Jones

Questions & Answers

Q: How can Fable 5.1 low effort help in knowledge work projects

Fable 5.1 on low effort helps by delivering a seven sheet workbook and a 13 slide deck quickly, enabling a fast initial view of the problem and the final result to be shaped through further review. It provides a structured draft that includes core assumptions, calculations, and a clear recommendation, which can be refined with higher effort levels or additional checks. This approach reduces turnaround time for early-stage analyses while preserving the ability to iterate.

Q: What additional value does extra effort provide

Extra effort surfaces questions that could flip the investment decision by adding uncertainty analysis, explicit funding needs, and refined financial metrics such as a weighted average cost of capital. It expands the workbook to nine sheets and the deck to 15 slides, offering deeper scrutiny and more nuanced scenarios, which helps stakeholders understand potential risks and outcomes before committing to a near final conclusion.

Q: Why is Soul important for knowledge work according to the video

Soul is important because it focuses on making the analysis easier to inspect and verify. It adds a dedicated sources sheet and a check sheet that clearly shows how the logic and formulas were followed, ensuring other analysts can audit the work efficiently. This transparency helps reduce misinterpretation and increases trust in the results.

Q: What is the main takeaway about low versus extra settings

The main takeaway is that low effort gives a strong first draft and a fast path to a usable analysis, while extra effort adds meaningful depth, checks, and questions that can alter decisions. Using both in sequence can improve both speed and reliability of knowledge work, balancing efficiency with rigor.

Q: How does the video describe token efficiency

The video discusses token efficiency by comparing how the low setting uses tokens to produce a basic, coherent workbook and deck, while higher settings invest more tokens to introduce uncertainty, additional questions, and more complete analyses. This illustrates how token use correlates with depth and diligence in the output.

Q: What real-world task is used to test the models in the video

The real-world task used is researching the GoPro acquisition by a company called Starman, building a post-acquisition discounted cash flow model in Excel, and turning the analysis into an executive PowerPoint deck. This scenario tests model capability, writing quality, and the ability to produce actionable financial insights under varying levels of effort.

Q: How do different models compare in producing writing after 100 words

Different models show varying strengths in concise writing. One model may include many details and color, making it harder to follow, while another keeps a tighter causal chain and clearer rationale with better labeling of facts. The comparison demonstrates how model design influences the clarity and usefulness of short-form writing.

Q: What is the recommended workflow for knowledge work according to the video

The recommended workflow is to start with a low effort draft to establish the core structure and initial conclusions, then use extra effort for deeper questions and robustness, and finally employ another tool like Soul to verify sources and checks. This multi-pass approach helps produce a reliable, auditable analysis while maintaining speed and adaptability.

Summary & Key Takeaways

• Fable 5.1 on low effort builds a multi-part workbook and deck that cover a real transaction, showing fast draft capability.

• Higher effort levels add more rigorous structure, questions, and financial checks, revealing why token cost matters for depth and reliability.

• Different models offer complementary strengths: fast draft, structured analysis, and verifiable sourcing, suggesting a multi-pass approach to knowledge work.