Ideas
Back to all appearances
Follow The Brand by Grant McGaugh
AI Is About to Expose How Your Company Really Works with David Dean
Hosted by Grant McGaugh
AI is not just a faster tool, and it is not only a threat to job security. David Dean joins Grant McGaugh to explain how AI can reveal the unwritten rules, side conversations, and invisible decisions that show how work actually gets done—and why that visibility makes governance, judgment, and human consequence more important.
Original episode
Listen or watch
Conversation summary
David approaches AI as a Business AI Realist, grounding the technology in the way people already do their jobs. He describes AI as a third presence in the relationship between employee and organization: something that can accelerate work, but also something that requires cultural adjustment while people and technology learn how to work with each other.
The episode’s central possibility is organizational self-awareness. Companies have a documented system of policies, values, process documents, and presentations, but they also have an undocumented system made of the steps and decisions people use to get the work done. AI can connect fragments across email, instant messages, meeting transcripts, document comments, and published material to make more of that second system visible.
David calls that communication and collaboration data the behavioral record. He uses a slow approval process as the practical example: looking across two years of correspondence might reveal that a decision-maker is afraid to act or is sitting in 30 meetings a week without the capacity to respond. The pattern does not require another AI tool to fix it. It may require a human conversation.
Visibility creates its own risk. David describes an employee comparing a leader’s response patterns—90% for one colleague and 20% for another—and explains why the signal can feel like evidence of favoritism without containing the whole story. Once everyday AI begins to function like a personal eDiscovery tool, the line between insight and surveillance depends on intent, perspective, and how people are educated to use what they find.
That is why David describes AI governance as 40% technical and 60% behavioral. Conventional security can constrain a known set of inputs and outputs; a large language model can accept and produce far more varied material. Organizations need technical controls, but they also need shared expectations about appropriate behavior, interpretation, authenticity, and accountability.
The conversation returns to why people matter. David identifies lived experience and lived consequence as qualities AI cannot carry in the same way, using a pilot’s six-second decision at LaGuardia to show how judgment includes responsibility for downstream human effects. His alternative to a replacement narrative is to use AI to give people breathing room for the critical and proactive work their jobs rarely leave time to do.
Key insights
- AI is a third presence in the relationship between an employee and an organization, not merely another workplace tool.
- Organizational self-awareness may be more valuable than another front-door AI application.
- Every company operates through a documented system and an undocumented system of actual work.
- Email, instant messages, meeting transcripts, document comments, and published material can be read together as a behavioral record.
- Two years of communication history can turn a recurring approval complaint into a pattern the organization can investigate.
- An operational bottleneck may come from fear, overload, or a calendar holding 30 meetings a week—and the response may be a conversation rather than another tool.
- Everyday AI can act like self-service eDiscovery, giving individuals the ability to assemble workplace stories from fragmented evidence.
- Insight and surveillance use similar capabilities; intent, perspective, and behavior determine how the capability is applied.
- AI governance is 40% technical and 60% behavioral because the possible inputs and outputs cannot be governed through technical controls alone.
- Lived experience and lived consequence remain central to human judgment, responsibility, and accountability.
- AI should give people breathing room for critical and proactive work instead of reducing them to the tasks listed in a job description.
Questions answered
Why does David describe AI as an organizational mirror?
Because AI can connect patterns across the communication and collaboration systems where everyday work leaves evidence. The reflection can show how the organization actually operates beneath its formal process descriptions.
What is the inbox as a behavioral record?
It is the accumulated evidence in email, chat, meetings, and document collaboration that records how people coordinate, make decisions, adapt processes, and move work forward.
What is the difference between the written and unwritten contracts of work?
The written contract contains official roles, policies, and processes. The unwritten contract consists of the relationships, workarounds, side conversations, and judgment through which the work actually gets done.
How can AI reveal organizational bottlenecks?
David uses approval delays as an example. Looking across two years of email, messages, and conversations may show that someone is afraid to decide or is overloaded by a calendar with 30 meetings a week. The technology creates the insight; the response may be a human conversation.
When does organizational insight become surveillance?
The boundary depends on why behavioral evidence is collected, who can see it, how much context is preserved, and whether people understand how the information will be used.
Why might leaders feel uncomfortable with this visibility?
Measurable patterns can expose gaps between stated policy and actual leadership behavior. David gives the example of an employee discovering that a leader responds to one colleague 90% of the time but responds to that employee only 20% of the time. The pattern is difficult to ignore even though it does not contain the whole context.
Why does David compare self-service AI to eDiscovery?
Legal teams use eDiscovery to reconstruct events by examining email and other records. When an everyday user connects AI to workplace communication, that person gains a similar ability to ask questions, connect fragments, and build a story from the behavioral record.
Why is AI governance 40% technical and 60% behavioral?
Traditional controls usually govern known inputs and outputs. Large language models allow far more varied questions and results, so technical guardrails cannot define every appropriate use. Behavioral education, expectations, intent, and accountability have to carry more of the governance burden.
What continues to make humans valuable in AI-enabled work?
People bring lived experience, understand the consequences of decisions, interpret ambiguous context, and remain accountable for what happens after an insight becomes an action. David illustrates the distinction with a pilot who had six seconds to choose how to respond to a fire truck on a runway at LaGuardia.
What should organizations do with time saved by AI?
Use it as breathing room for people to think, decide, and lead. The value is not only in making work move faster, but in creating more room for human judgment.
Conversation guide
The Business AI Realist
Why David grounds AI in the reality of everyday work instead of beginning with future-facing “what if” scenarios.
AI as an organizational mirror
Why creating organizational self-awareness may be a more useful starting point than building another tool.
The inbox as a behavioral record
How email, chat, meetings, and collaborative documents preserve evidence of real workplace behavior.
The written and unwritten contracts
Where official policy collides with the relationships, decisions, and workarounds that move work forward.
Two years of approval history
How correspondence can reveal whether a slow approval comes from fear, overload, or a calendar with 30 meetings a week.
From insight to surveillance
Why a 90% versus 20% response pattern can create a powerful signal without providing the whole story.
Self-service eDiscovery
How connecting AI to everyday correspondence gives individuals new power to assemble workplace stories.
Governance is behavioral
Why David frames AI strategy as 40% technical and 60% behavioral.
Lived experience and lived consequence
The human context, judgment, and accountability that remain essential when AI surfaces a pattern.
Breathing room to lead
Using AI to create capacity for thinking and decision-making rather than simply increasing output.
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.
“A lot of people are saying, “Will AI replace my job?” But today’s guest, David Dean, challenges us to ask an even deeper and more powerful question: What does AI reveal about how work actually gets done?”
“What if the biggest thing that AI can do is not about tools and acceleration, but create organizational self-awareness? That is really the opportunity: AI can help organizations truly understand themselves.”
“There are two systems in every company. There is a documented system and an undocumented system. I also refer to them as the written contract and the unwritten contract.”
“The inbox is the behavioral record of all the communication and collaboration data on how the organization behaves.”
“Use AI to discover yourself. Discover where your challenges are in your unwritten-contract behaviors, and then utilize that to address your challenges as a company.”
“It does not take an AI tool to solve it. It takes AI to help you create that insight, so then you can go have human conversations about addressing those issues and problems.”
“There is one thing that humans will always have that AI will not, and that is understanding of lived experience and lived consequence.”
“It all comes down to intent. It can be used in a very positive way, and it can be used in a very negative way. What is your goal around surveillance or insight?”
“When I talk about governing AI from a strategy standpoint, I always say it is 40% technical and 60% behavioral.”
“You are valued. You will always be needed.”
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.