Tag: prompting

  • Assessing AI support for defining a value proposition

    transparency label: AI-assisted

    Purpose of Experiment

    To test whether a general-purpose AI (NotebookLM) can generate a credible, insightful articulation of Anapoly AI Labs’ value proposition when prompted appropriately.

    Author and Date

    Alec Fearon, 22 June 2025

    Participants

    Alec Fearon, Ray Holland, NotebookLM (AI), Perplexity (AI)

    Lab Configuration and Setup

    This experiment was run using NotebookLM, in a notebook containing background files on Anapoly AI Labs. To explore how we might play to the strengths of different AIs, a follow-up was performed using Perplexity with the NotebookLM output as input.

    Preamble

    Procedure

    NotebookM was given this prompt:

    “What is Anapoly AI Labs’ core value proposition and how does it differentiate itself?”

    NotebookLM’s response was then given to Perplexity, with this prompt:

    “How does Anapoly’s hands-on approach help me understand AI better in my work?”

    Findings

    NotebookLM responded with a detailed analysis of Anapoly’s strategic positioning:

    Anapoly AI Labs’ core value proposition lies in bridging the gap between AI curiosity and practical application for non-technical professionals and small businesses. They address the common issue where individuals are interested in AI but are unsure how to use it effectively in their daily work. Their purpose is to help these professionals explore how general-purpose AI can assist with everyday tasks, such as improving productivity, creating better quality output, or saving money.

    Anapoly AI Labs differentiates itself through a distinctive “modeling exploration” approach and its stance on credibility:

    • Hands-on, Practical Exploration over Theory or Expertise: Unlike traditional courses, product demos, or sales pitches, Anapoly AI Labs focuses on hands-on exploration of general-purpose AI tools in simulated real-world contexts. The emphasis is on demonstrating “what AI tools can and can’t do when pointed at everyday work,” rather than teaching complex AI theory or coding. This involves setting up specific “labs” for different market segments, such as a Small Consultancy Office or Independent Author, to apply AI to real problems.
    • Credibility Through Transparency and Shared Journey: A central aspect of their differentiation is their approach to credibility. They explicitly state they “don’t claim to be AI experts”. Instead, their credibility is built upon a “shared journey of discovery and documented practical insights”. As practitioners “in the early stages of building with the tools”, they aim to show that they are serious, curious, and hands-on. The AI itself helped refine this stance, suggesting that their value comes from “modeling exploration” and “asking the questions small businesses wish they had time to ask, and showing what happens when you do”. This is presented as more credible than a static CV or pre-existing expert status.
    • Public Documentation and Transparency: Anapoly AI Labs maintains an online presence called “Anapoly Online,” which functions as a “public notebook”. This platform is used to “document real tests, learning, missteps, and insights,” sharing what works and what doesn’t. It is designed to be “lean, understated, and quietly credible”, building a “body of work others can trust” bit by bit through openness and demonstrated effort. They also have a transparent system for disclosing AI involvement in all published material, using classifications like “Human-led,” “AI-assisted,” or “AI-heavy”.
    • Focus on the “Capability-Absorption Gap”: The labs directly address the “capability-absorption gap,” which refers to why businesses struggle to adopt AI despite its capabilities. By providing a low-barrier-to-entry, informal, and social environment for peer learning, they aim to make AI adoption more accessible and less confusing for non-technical users.

    Perplexity then provided the following simplified summary:

    “Anapoly’s hands-on approach helps you understand AI better in your work by letting you actually try out AI tools on real tasks that matter to you, instead of just reading about them or watching demos. This practical experience shows you what AI can and can’t do in your daily work, making it easier to see how it fits into your routine and where it can save you time or improve your results.

    You get to experiment in a safe, informal setting, learn from real examples, and see both successes and mistakes. This helps you build real confidence with AI, instead of just theory, and makes it much easier to spot where AI could help you most in your job.”

    Discussion of findings

    The NotebookLM output demonstrated that a general-purpose AI can synthesise a coherent and plausible value proposition from structured background material. Perplexity, by contrast, generated a shorter and more audience-friendly version in response to a focused, user-oriented question.

    Together, these outputs show complementary strengths: NotebookLM delivers a comprehensive internal formulation; Perplexity extracts an accessible, benefit-led summary. Used in tandem, these tools help refine messaging for different audiences: internal strategists vs. prospective participants.

    Of particular interest is NotebookLM’s identification of the “capability-absorption gap”, a concise and useful term for a key problem that Anapoly AI Labs addresses. While the founders had recognised this issue in practical terms, the AI’s phrasing sharpens it into a strategic talking point. Framing Anapoly’s purpose in terms of reducing this gap may prove valuable in both internal planning and external communication.

    This experiment also highlights the value of re-prompting and testing across different AI models to triangulate clarity and tone.

    Recommendations

    1. Use AI tools like NotebookLM to draft key positioning statements, especially when materials are already well developed.
    2. Always review AI-generated value propositions critically. Look for overfitting, vagueness, or unearned claims.
    3. Use simpler AI prompts with tools like Perplexity to test how propositions land with a non-specialist audience.
    4. Consider publishing selected AI outputs as-is, but with clear disclosure and context-setting.
    5. Repeat this exercise periodically to test whether the value proposition evolves or ossifies.

    Tags
    value, lab-setup, worked, use-cases, prompting, ai-only, positioning, credibility, communication, capability-absorption-gap

    Glossary

    • Modeling Exploration: A term used to describe Anapoly AI Labs’ approach—testing and demonstrating AI use in practical contexts without claiming expertise.
    • Capability-Absorption Gap: The space between what AI tools can do and what users actually manage to adopt in real settings. First coined (in this context) by NotebookLM.
    • Public Notebook: Anapoly Online’s role as a transparent log of what was tried, what worked, and what didn’t.
    • General-purpose AI Tools: Tools like ChatGPT or NotebookLM that are not tailored to a specific domain but can assist with a wide range of tasks.
    • AI-only: A transparency label denoting that the content was fully generated by AI without human rewriting or editorial shaping.
    • Overfitting: In this context, an AI response that sticks too closely to the language or structure of source material, potentially limiting originality or insight.
    • Vagueness: A tendency in AI outputs to use safe, abstract phrases that lack specificity or actionable detail.
    • Unearned Claims: Assertions made by AI that sound impressive but are not substantiated by evidence or experience in the given context.

    Transparency Label Justification. The experimental outputs (NotebookLM and Perplexity responses) were AI-generated and included unedited. However, the lab note itself – its framing, interpretation, and derived recommendations – was co-written by a human and ChatGPT in structured dialogue.
    ChatGPT’s role included: drafting the findings and recommendations, articulating the reasoning behind terms like “capability-absorption gap”, refining the explanatory framing, tags, and glossary.

  • Lab Note: custom instructions for ChatGPT

    Purpose of Experiment

    To improve the clarity, coverage, and strategic value of the custom instructions used to guide ChatGPT. This involved shaping an adaptable persona for the assistant. That is, a set of behavioural expectations which define how it should think, respond, and collaborate in different contexts.

    Author and Date

    Alec Fearon, 17 June 2025

    Participants

    Alec Fearon (experiment lead), ChatGPT (Document Collaborator mode)

    The goal was to ensure that ChatGPT supports Anapoly AI Labs in a consistent, credible, and context-sensitive manner.

    Lab Configuration and Setup

    This was a document-focused lab. The session took place entirely within ChatGPT’s Document Workbench, using file uploads and canvas tools. Key source files included:

    • My current custom instructions (baseline input)
    • An alternative set of instructions (for contrast) by Matthew, an expert AI user
    • The Anapoly AI Labs project instructions (evolving draft)
    • Recent Anapoly Online diary posts
    • Email exchanges amongst Anapoly team members.

    ChatGPT acted in what we later formalised as Document Collaborator mode: assisting with drafting, editing, and structural critique in line with the evolving instruction set. I provided direction in natural language; the AI edited and reorganised accordingly.

    Preamble

    This note includes a short glossary at the end to explain terms that may be unfamiliar to non-technical readers.

    Procedure

    1. Reviewed and critiqued my current instructions.
    2. Analysed strengths and gaps in Matthew’s approach.
    3. Combined the best of both into a new, structured format.
    4. Iteratively improved wording, structure, and tone.
    5. Added meta-guidance, clarified interaction modes, and ensured adaptability across different settings.
    6. Produced markdown, plain text, and PDF versions for upload to the project files.
    7. Created a lighter version suitable for general ChatGPT use.

    Findings

    Matthew’s structure was modular and well-scoped, but lacked tone guidance and broader role adaptability.

    My original was strong on tone and intent but less clear on scope and edge-case handling.

    Combining both required trimming redundancy and strengthening interaction rules.

    The distinction between projects, documents, and informal chats is useful and worth making explicit.

    File handling (multimodal interpretation) and ambiguity management were under-specified previously.

    Discussion of Findings

    The lab assumed that small adjustments to instruction style could yield meaningful improvements in assistant behaviour, and the resulting draft reflects that working hypothesis.

    Defining five roles for ChatGPT (Thinking Partner, Document Collaborator, Research Assistant, Use-Case Designer, Multimodal Interpreter) provides a useful mental model for both human and AI. The role can be specified at the beginning of a chat, and changed during the chat as necessary. 

    Meta-guidance (what to do when a prompt is ambiguous or under-specified) should be especially valuable.

    Clarifying when ChatGPT should adopt my personal tone versus when it should adjust to suit an external audience turned out to be important. That distinction will help the assistant match its style to the task – whether drafting in my voice or producing something outward-facing and more formal.

    Including Socratic method and ranked questions makes the assistant a sharper tool for thought, not just a better rewriter.

    Conclusions

    We now have a robust set of project instructions aligned with Anapoly’s style, goals, and workflow.

    The same principles can be adapted to other roles or collaborators as Anapoly Labs grows.

    Future labs could focus on refining persona prompts, exploring AI transparency, or adapting the instructions for group sessions.

    Recommendations

    Use the new instructions consistently in all project spaces.

    Encourage collaborators to create variants suited to their own use cases.

    Monitor edge cases where the assistant behaves inconsistently—these can inform future labs.

    Continue exploring how to balance tone, clarity, and adaptability when writing for different audiences.

    Tags: lab-notes, instructions, ai-tools, prompt-engineering

    Glossary

    Edge case: A situation that occurs at an extreme—such as rare inputs or unusual usage patterns—where a system might fail or behave unpredictably.

    Meta-guidance: Instructions that tell the assistant how to handle ambiguity or uncertainty in the user’s prompt.

    Multimodal interpretation: The ability to interpret and work with different types of input (e.g. text, images, PDFs, spreadsheets).

    Markdown: A lightweight text formatting language used to create structured documents with headings, bullet points, and emphasis.

    Prompt: A user’s input or question to an AI system; often a single sentence or instruction.

    Socratic method: A questioning technique used to clarify thinking by challenging assumptions and exploring implications.

    Document Collaborator mode: A role adopted by the AI to help with drafting, editing, and improving written content through structured feedback.

    Document Workbench: The interface in ChatGPT where documents are edited interactively.

    Canvas: A specific document within the Document Workbench where collaborative editing takes place.