Ideas
The Deep Wealth Podcast by Jeffrey Feldberg
Author & AI Strategist David Dean: AI Won’t Fix Your Business But Will Reveal What’s Broken
Hosted by Jeffrey Feldberg
AI can expose the broken workflows, missing judgment, and leadership silence already costing your business. Author & AI Strategist David Dean reveals why organizational self-discovery must come before automation.
Original episode
Listen or watch
Conversation summary
A job description records the duties an organization can name. The last line, “other duties apply,” points to the ambiguity people handle every day. David Dean joins Jeffrey Feldberg to examine that unwritten contract of work: the decisions, informal coordination, and adaptations that keep a business operating beyond its official account of itself.
The conversation turns to leadership silence. David describes how communication records can show that someone received information but never acted on it. A stalled decision may reflect avoidance, or it may mean a person does not believe they have authority to proceed. In the latter case, a conversation that establishes trust and decision rights may resolve the problem without another tool.
Spreadsheets offer another view of work that formal systems miss. People create them to track responsibilities and move tasks forward, often without stopping to examine the habit. David also considers the risks of analyzing messages and meeting transcripts: information people already have permission to read can reveal new patterns when examined together. He raises questions about acceptable use, personality profiling, and the consistency of a leader’s public and private behavior.
Jeffrey asks whether AI informed by yesterday’s company could hold back a company trying to change. David responds with human-directed AI: people define the question, control the relevant knowledge, assess the result, and explain why they made a decision. Recording that reasoning gives future work more context than a thumbs-up or thumbs-down response.
The practical starting point is organizational self-discovery. David suggests spending a week dictating what you do, why you do it, and how you handle problems. AI can help identify signals in that account; people interpret those signals alongside their experience and communication records. Business domain experts need to lead this work with technology teams, and prospective advisors should be able to discuss the everyday behavior, constraints, and failures they have encountered inside organizations.
Key insights
- The official account of a business leaves out much of the judgment and adaptation required to deliver it. Those unwritten behaviors belong in the assessment of any proposed automation.
- Silence can leave evidence. A record of unanswered information can help a leader examine where a decision stalled and whether the person had authority to act.
- A spreadsheet can be a locally built system for tracking work and responsibility. Understanding why someone created it can reveal needs the formal process has missed.
- Access to individual messages does not settle how people should use AI to infer patterns across them. David argues that organizations need to confront acceptable use as well as technical access.
- Human-directed AI includes people’s reasoning before and after the AI interaction. Explaining a decision preserves context that a simple approval or rejection does not.
- Employees need room to apply judgment and creativity. David proposes explaining AI adoption in terms of the ambiguous work it can help them handle and the time it can return to them.
- Organizational self-discovery requires business expertise. Technology teams can enable the systems, while the people who understand the work help determine where and how to use them.
- When choosing an AI advisor, ask about time spent with the people doing the work, including the process failures and behavioral constraints the advisor has had to understand.
Questions answered
Why should organizational self-discovery come before automation?
Formal processes show only part of the work. David argues that leaders need to understand the adaptations and decisions people rely on before choosing what to automate. Otherwise, the project may carry existing problems into a faster process.
What can leadership silence reveal?
Communication records may show where information arrived and no decision followed. David distinguishes a failure to exercise judgment from uncertainty about authority. The appropriate response may be a direct conversation that gives someone permission and confidence to act.
What does human-directed AI mean in this conversation?
People guide the work at the beginning and the end: they define the question and relevant knowledge, assess the output, and explain the decision they make. David emphasizes capturing that reasoning so the system has more useful context for subsequent work.
How can someone begin organizational self-discovery?
Dictate a week of work notes describing what you do, why you do it, and how you handle situations. Ask AI to help identify challenges and opportunities in that account, then interpret the signals with your own knowledge and relevant communication records.
What should a business ask a prospective AI advisor?
Ask how much time they have spent with everyday business operations and the people carrying them out. David looks for an ability to connect process knowledge with human behavior, including practical challenges and failures.
Conversation guide
The unwritten contract of work
Why “other duties apply” points to the ambiguity and judgment that a job description cannot fully capture.
Silence, authority, and leadership consistency
How unanswered information can reveal a stalled decision, and why restoring authority may be the necessary intervention.
Spreadsheets and automatic behavior
The familiar tools people build to track work, take responsibility, and compensate for missing systems.
What behavioral data can reveal
Questions about inference, acceptable use, and the risks of interpreting communication across an organization.
Human-directed AI
Keeping people in charge of the question, the relevant knowledge, and the reasoning behind the final decision.
A week of self-discovery
Using a dictated account of daily work to find signals worth examining with the people who understand the business.
Choosing an advisor close to the work
Looking for experience with business behavior and operational failures when evaluating AI advice.
Timecodes are omitted where the supplied transcript did not contain verifiable timing.
Edited transcript excerpts
Speaker labels and punctuation were reconstructed from automated transcription and edited for readability.
“How well can they make decisions? How well can they deal with the ambiguity of everything that’s relating to this with it? And the only way to do that is lived experience and lived consequence.”
“But sometimes it’s not just a lack of judgment. Sometimes they don’t feel like they have the right authority to make this judgment call in there.”
“Not a thumbs up, not a thumbs down. I need context, with it.”
“Start telling your story, your business story, what you do every day, why you do it every day, and how you handle things every day with it.”
“And you say it’s one of the subtitles, “AI is not here to replace human judgment. It is here to make the absence of judgment visible.” And I thought, “Wow, that encapsulates what I’ve often wondered.” … Here for the first time, potentially, is something to take those blind spots, those inflection points right there in front of us, but they’re hiding, and make them visible.”
Keep exploring
Follow the idea without losing your place
Start with the closest connections. Open the additional links only when you want to go further.
Reading
Related articles
Source
Original episode
Continue the conversation
Explore the book behind these ideas, bring the conversation to an audience, or learn more about David’s work.