The Rationale: When the Human Voice Is Ignored, Does It Matter Who Delivers the Message?
By Hiran de Silva
October 2026
I am about to launch a series of video presentations featuring AI-generated characters.
Some will be explainers. Others will be conversations, debates and dramatisations. They will include my Tim and Reg series and my forthcoming Excel Mission Impossible productions, featuring five characters with different personalities, professional backgrounds and perspectives.
I anticipate that some people will object.
They may call the presentations artificial, inauthentic, dishonest or even deceitful.
Indeed, given some of the recent commentary about AI-generated content, I wouldn’t be surprised if those criticisms came from people with whom I’ve already had disagreements about Excel education.
But before anyone dismisses these productions simply because the presenters are AI avatars, I would like to explain why I am using them.
And, more importantly, why I believe the information they will be communicating deserves serious attention.
Because there is an interesting history behind this decision.
And it raises a rather uncomfortable question.
If I present important information personally, backed by thirty years of professional experience, and that information is ignored, does presenting the same information through an AI avatar somehow make it less true?
Or should we be examining the information rather than the person—or avatar—delivering it?
1. A Video I Made Three Years Ago
In the summer of 2023, I made a video specifically for Excel educator Wyn Hopkins.
It concerned cascading dropdown lists, a technique that was receiving considerable attention in the Excel community at the time.
The video was never published.
I subsequently developed what I considered a more useful demonstration, the REG Call Handler, which illustrated the same underlying principles in a practical enterprise application.
But the original video made a point that I believe is fundamental to understanding Excel.
It was a distinction that has informed much of my professional career.
There is an enormous difference between using Excel as a personal productivity document and using Excel as part of an enterprise business process.
And the techniques appropriate to one are not necessarily appropriate to the other.
This distinction is at the heart of what I now call the 10%–90% divide.
The 10%: Excel as an Editable Document
Consider the spreadsheet created for one person’s use.
It contains data, formulas, perhaps some dropdown lists, charts and reports.
The user creates it, edits it, saves it and uses the results.
The spreadsheet is essentially the end of the process.
There are no significant downstream dependencies. Nobody else needs to interact with its data. It is not an operational component of a larger business system.
The spreadsheet might occasionally be sent to a manager as evidence that some work has been completed.
But nothing else depends on it.
This is perfectly legitimate Excel usage.
It is particularly appropriate for individuals, hobbyists, sole traders and many personal productivity tasks.
And a considerable amount of Excel education on social media is directed towards this type of usage.
There is nothing inherently wrong with that.
The problem begins when techniques designed for this environment are applied to enterprise processes without recognising the fundamentally different requirements.
The 90%: Excel as Part of an Enterprise Process
In an enterprise, a spreadsheet is often not the end of the process.
It is part of an ongoing business operation.
Other people need its information. Other departments depend on its outputs. Transactions and master data change continuously.
A spreadsheet may be one of hundreds or thousands participating in the same business process.
And this changes everything.
Data must be accessible, current, consistent and appropriately controlled.
The spreadsheets must work together.
The architecture must support the business process rather than simply the individual document.
Yet this distinction is rarely discussed in mainstream Excel education.
And that omission has enormous consequences.
2. Cricket, Football and the Art of Bending a Ball
In my 2023 video, I used an analogy to illustrate this distinction.
In cricket, a skilled spin bowler can make a ball turn dramatically.
In football, David Beckham became famous for bending a football around defenders and into the goal.
In both cases, we are watching a ball change direction.
Superficially, the techniques have something in common.
But nobody would suggest that a cricket spin bowler could automatically teach David Beckham how to take a free kick.
The sports are different.
The objectives are different.
The mechanics are different.
The surrounding conditions are different.
A superficial similarity does not establish functional equivalence.
Now consider cascading dropdown lists in Excel.
In a personal spreadsheet, the data supplying those dropdowns might reside in another worksheet within the same workbook.
Perfectly sensible.
But in an enterprise application, the data might represent products, inventory, customers, warehouses or other information that is constantly changing.
That information cannot sensibly be maintained independently in hundreds of spreadsheets.
It needs to be held centrally, where it can be updated and accessed by every participating spreadsheet.
The dropdown might look identical to the user.
But the architecture behind it is fundamentally different.
One is a self-contained document.
The other is a component of an enterprise information system.
Same appearance. Different purpose. Different architecture.
That was the message I wanted to communicate three years ago.
And it remains central to everything I am teaching today.
3. The Missing Transformation: From 10% to 90%
Here is where the problem becomes serious.
Excel educators teach techniques that are perfectly appropriate for personal spreadsheets.
Their audiences learn those techniques.
Some of those people then take the same techniques into their workplaces and attempt to apply them to enterprise-scale processes.
They build spreadsheets that depend on other spreadsheets.
They distribute copies.
They introduce external links.
They create increasingly elaborate mechanisms for gathering, reconciling and consolidating information.
Eventually, the organisation finds itself struggling with multiple versions, broken links, inconsistent data and endless reconciliation.
I call this the swamp.
The industry has another name for it.
Excel Hell.
And the Excel replacement industry has been exploiting Excel Hell for years.
Companies such as Anaplan, Workday Adaptive Planning, Planful and Datarails promote alternatives to Excel-based processes by highlighting precisely these problems.
Their argument is that Excel is unsuitable for enterprise-scale operations.
But I have spent thirty years demonstrating something very different.
The problem is not necessarily Excel.
The problem is the architecture.
And, more specifically, the failure to recognise that the methods appropriate to a standalone personal spreadsheet cannot simply be extended to an enterprise process.
Excel already possesses capabilities that enable it to participate in properly designed enterprise systems.
It can communicate with relational databases.
It can GET information from a central source.
It can PUT transactions and updates back into that source.
It can participate in processes involving hundreds or thousands of users and spreadsheets.
These are not new capabilities.
They have been available for decades.
The missing transformation is not a new Excel feature. It is a change in thinking.
And that change in thinking is almost entirely absent from mainstream Excel education.
4. What Happened When I Explained This Personally?
This brings me back to Wyn Hopkins.
I explained the distinction to him personally in 2023.
I created a video in which I appeared on camera.
There was no AI avatar.
No synthetic voice.
No fictional character.
It was simply me, drawing upon my professional experience, explaining why the difference between personal and enterprise spreadsheet techniques matters.
As I recall, he regarded the distinction as dismissive of the work being taught.
But I was not dismissing the techniques.
I was identifying their appropriate context.
There is a world of difference between those two positions.
My concern is that, from the publicly available educational material I have seen since then, I have not observed a corresponding emphasis on this crucial architectural distinction.
The teaching continues to focus predominantly on techniques for the standalone spreadsheet.
And the audience is left without a clear warning about what happens when those techniques are carried into enterprise processes.
Now, I cannot know precisely what another person has understood or why they choose to teach particular subjects.
But I can ask a reasonable question.
If this distinction is important enough to prevent Excel Hell, why is it not being taught?
And if I have already explained it personally, what exactly would be achieved by insisting that I must continue to present everything personally?
That is one reason I am experimenting with a different medium.
5. The Database That Apparently Takes a Year to Obtain
Let me offer another example.
Mark Proctor has argued that citizen developers face enormous obstacles when trying to obtain a database from their IT departments.
Requests can take months, perhaps a year.
Then there are permissions, change management procedures and other organisational barriers.
I understand that such obstacles exist in some organisations.
But the suggestion that a business user cannot have access to a relational database without going through this process does not reflect my experience.
In fact, it overlooks something rather significant.
Microsoft Access is a relational database application that has been part of the Microsoft Office family for decades.
And Excel can communicate with it.
In my demonstrations, I show how a database can be created and used with remarkable ease.
Creating a simple local database is not the same as obtaining approval for a production enterprise system, of course. Governance, security and deployment requirements must be considered.
But that is precisely why the distinction matters.
The technical capability and the organisational approval process are two different things.
One should not be confused with the other.
And yet, how often do we see this distinction clearly explained to Excel users?
I have demonstrated these capabilities.
I have explained the history.
I have shown practical examples.
But I have not seen a public correction from Mark acknowledging the broader availability of these options.
Again, the issue is not whether Mark or I should win an argument.
The issue is whether the audience is receiving an accurate understanding of the available technology.
6. The REG Call Handler: Apparently, Excel Cannot Do That
Another example concerns Paul Barnhurst.
In discussions about my REG Call Handler demonstration, Paul argued that the requirements called for a proper system rather than Excel.
But that was exactly the point of my demonstration.
I was showing that Excel could be the user-facing component of a properly designed system.
Not a standalone spreadsheet attempting to behave like an enterprise application.
Not hundreds of workbooks exchanging copies of data.
But Excel leveraging a relational database to extend its capabilities.
The distinction is fundamental.
And the solution was not hypothetical.
I had demonstrated it.
If somebody sees evidence that challenges an earlier statement, surely the appropriate response is to examine the evidence.
And, if necessary, revise the earlier statement.
There is no shame in learning something new.
There is considerable value in sharing that learning with an audience.
Yet I have not seen the corresponding public correction.
Which raises a question that goes far beyond Excel.
What responsibility does an educator have to correct information when presented with evidence that challenges what they have previously taught?
7. Power Query, VBA and Another Misunderstanding
The same problem appears in discussions about Power Query and VBA.
We regularly encounter statements along the lines of:
“Since Power Query arrived, I haven’t needed VBA.”
That may accurately describe someone’s personal working practices.
But it does not establish that Power Query replaces VBA.
One reason for the confusion, I suspect, is the superficial similarity between Power Query’s Applied Steps and the sequential instructions produced by Excel’s macro recorder.
Both can describe a series of operations.
But VBA is much more than a means of recording repetitive actions.
It provides access to Excel’s object model, supports application logic, responds to events, communicates with external systems and can orchestrate entire business processes.
Power Query has a different primary purpose.
It is principally concerned with connecting to, transforming and loading data.
These technologies overlap in some tasks, but they are not functionally equivalent.
In my generation, many of us learned Excel’s object model by studying the VBA generated by the macro recorder.
It was an extraordinarily useful learning technique.
But the macro recorder was never the definition of VBA.
And Power Query’s Applied Steps are not a substitute for the wider programming and automation capabilities that VBA provides.
Again, we are confusing two things because they look superficially similar.
Just as we confused the cricket ball with the football.
Just as we confuse a personal spreadsheet with an enterprise spreadsheet.
8. Power Query Is Not a Relational Database
I have encountered similar misunderstandings in discussions with Christopher de Fennell and others.
Power Query can work with tables.
It can perform joins.
It can transform and combine data.
But that does not make it a relational database management system.
Nor does it make a Power Query transformation a substitute for a transactional system of record.
In my enterprise Excel architecture, the relational database has a specific purpose.
It holds shared business data.
It supports transactions.
It provides a central source from which spreadsheets retrieve information and to which authorised users submit updates.
That is the foundation of the GET/PUT architecture I have used for decades.
Power Query can certainly participate in a wider solution.
But it does not replace the transactional role of the database.
I have explained this distinction repeatedly, including in direct discussions.
Yet I continue to see comparisons that blur these fundamentally different functions.
Which brings me to the central issue.
9. Human Authenticity Versus Factual Authenticity
For years, I have explained these concepts personally.
I have written articles.
I have responded to LinkedIn posts.
I have created demonstrations.
I have shown working solutions.
I have drawn upon thirty years of experience designing enterprise Excel systems, often for senior finance executives.
And I have repeatedly encountered situations where the information I presented appeared to make little difference to the public narrative.
Now I propose to communicate those same ideas through AI-generated characters.
And some people may object that the presentations are not authentic.
But what exactly do we mean by authenticity?
There are at least two separate questions.
First: Is the presenter a real human being?
Second: Is the information being presented accurate?
These are not the same question.
A real person can present misleading information.
An AI-generated character can present information that is accurate, demonstrable and supported by evidence.
Equally, AI-generated content can be wrong, misleading or fabricated. The technology provides no guarantee of truth.
The appropriate test is therefore the same in both cases.
Examine the claims.
Examine the evidence.
Examine the reasoning.
And, wherever possible, test the demonstration.
I am not proposing to use AI characters to impersonate real people or to deceive viewers about who is speaking.
They will be openly presented as fictional characters.
Their dialogue will be written and directed by me.
The ideas will be mine.
The case studies will be mine.
The demonstrations will be based on working solutions.
The AI will provide the voices, faces and dramatic presentation.
It is a production technique, not a substitute for evidence.
And that distinction deserves to be understood.
10. Why Tim and Reg? Why Mission Impossible?
There is another reason I am developing these productions.
Some subjects are more effectively explained through conversation than through a lecture.
Consider Tim and Reg.
Tim represents the capable Excel user who has learned the techniques commonly taught in the Excel community.
He is not incompetent.
He is not foolish.
He is working with the knowledge available to him.
Reg brings a different perspective.
He understands Excel as a component of a wider business process.
He introduces the concepts of shared data, relational databases, GET/PUT and enterprise architecture.
The conversation allows the audience to discover the difference gradually.
Tim asks the questions that a real Excel professional might ask.
Reg challenges assumptions.
And the audience follows the reasoning.
My forthcoming Mission Impossible series extends this approach.
Jim provides leadership and asks the practical questions.
Cinnamon understands Excel education and training.
Barney explains the technology.
Rollin draws upon experience and challenges conventional thinking, often rather forcefully.
Willy contributes the practical perspective needed to turn plans into action.
The contrasting personalities create opportunities for disagreement, clarification and discovery.
Instead of simply telling the audience what to think, I can dramatise the process of examining an assumption and testing it against reality.
And AI production tools make this possible without assembling a cast of professional actors, hiring studios or managing a conventional film production.
I can concentrate on the script, the reasoning and the evidence.
That is an extraordinary opportunity for an independent educator.
11. Who Should Decide Whether This Matters?
There is a bigger question behind all this.
Is the 10%–90% distinction important?
Does it matter that enterprise spreadsheets require a different architecture from personal spreadsheets?
Does it matter that businesses can use Excel with relational databases to build shared, controlled and scalable processes?
Does it matter that techniques which work beautifully in isolation may create serious problems when extended across an organisation?
I believe these questions should be considered by the people responsible for the consequences.
Finance directors.
CFOs.
Business leaders.
Senior managers.
People responsible for budgets, operations, compliance, reporting and information systems.
They are the people who experience Excel Hell.
They are the people who are told that Excel is the problem.
They are the people who may be asked to spend substantial sums replacing Excel-based processes with proprietary applications.
And they are the people who should be able to examine the alternatives.
If the architecture I am demonstrating can prevent the problems they face, or help them escape from those problems, that is commercially important.
If it cannot, then the evidence should establish that.
But the question should not be settled by the preferences of social media educators.
Nor should it be settled by whether they approve of AI-generated presenters.
The question is whether the information is correct, relevant and useful to the business.
That is the test that matters.
12. The Paradox of Misinformation
There is an irony in all of this.
The misunderstandings I have been describing are themselves part of the reason why enterprise Excel expertise can be so valuable.
When organisations are taught to think of Excel primarily as an editable document, they naturally attempt to solve enterprise problems using document-oriented techniques.
When those techniques fail at scale, the resulting problems are attributed to Excel.
The Excel replacement industry then offers expensive alternatives.
But a different approach to Excel architecture can remove many of the causes of those problems.
That is precisely the opportunity I discovered in my professional career.
In several organisations, introducing that different way of thinking resulted in extraordinary commercial outcomes, including substantial increases in my consulting rates.
The value did not come from knowing more spreadsheet functions.
It came from understanding the business process and using the available technology differently.
In that sense, the widespread misunderstanding creates an opportunity.
But it is an expensive opportunity for the organisations that remain trapped in Excel Hell.
And it is a missed opportunity for Excel professionals who have never been taught the enterprise techniques that could transform their careers.
I would much rather see those capabilities become common professional knowledge.
That is why I am creating these educational productions.
13. So, Is Using an AI Avatar Dishonest?
Let me return to the anticipated criticism.
Suppose somebody watches a Tim and Reg episode and says:
“Why are you using AI? Why don’t you explain this yourself?”
My answer is simple.
I already have.
I have explained it personally.
I have demonstrated it personally.
I have written about it.
I have discussed it with educators and industry commentators.
I have spent decades putting these principles into practice.
Now I am exploring another way to communicate the same information.
A way that allows characters to ask questions, challenge assumptions, disagree, explain and demonstrate.
A way that may make complex ideas easier to understand.
A way that can reach people who would never sit through a lengthy technical lecture.
If the objection is that viewers might mistake a fictional AI character for a real person, I agree that transparency matters.
I will make it clear that these are AI-generated characters.
But if the objection is that information becomes less credible simply because an AI avatar delivers it, I reject that argument.
The information does not change its truth value when the presenter changes.
A false statement does not become true because a respected human educator says it.
And a true statement does not become false because a fictional AI character says it.
The responsibility remains with the author to ensure that the information is accurate, and with the audience to evaluate it critically.
Conclusion: Judge the Message, Not the Messenger
The Excel community has spent decades teaching people how to work with spreadsheets.
Much of that teaching has been excellent within its intended context.
But the transformation from personal spreadsheet usage to enterprise spreadsheet architecture remains largely missing from mainstream education.
That missing transformation has consequences.
It contributes to the proliferation of disconnected spreadsheets, duplicated data, reconciliation problems and fragile processes.
It helps create the conditions that the industry calls Excel Hell.
And it sustains a market for products promoted as replacements for Excel.
Yet the capabilities needed to address many of these problems have existed within the Microsoft technology ecosystem for decades.
That is the story I want to tell.
It is a story about technology.
But it is also a story about education, professional responsibility, the spread of misinformation and the reluctance to challenge familiar assumptions.
I have told parts of that story personally for years.
Now Tim, Reg, Jim, Cinnamon, Barney, Rollin and Willy will help me tell it differently.
They are fictional characters.
The productions will use AI.
But the problems are real.
The demonstrations are real.
The professional experience behind them is real.
And the evidence can be examined.
So, before anyone dismisses the series because the presenters are not human, I would invite them to consider a more useful question.
If the information can help businesses escape Excel Hell, does it really matter whether the person explaining it is standing in front of a camera—or is an AI avatar?
And perhaps an even more important question:
If the same information has already been presented by a human being and ignored, shouldn’t we be asking why?
That, ultimately, is the rationale behind my AI productions.
The messenger has changed.
The message has not.
And the evidence still deserves to be heard.
Hiran de Silva
Former CFO | Enterprise Excel Consultant | Author of Triple Your Pay with Excel
Here is the complete article again, in ordinary text, without the special writing block.
Using AI Avatars to Explain the 10%–90% Divide
The Rationale: When the Human Voice Is Ignored, Does It Matter Who Delivers the Message?
October 2026
I am about to launch a series of video presentations featuring AI-generated characters.
Some will be explainers. Others will be conversations, debates and dramatisations. They will include my Tim and Reg series and my forthcoming Excel Mission Impossible productions, featuring five characters with different personalities, professional backgrounds and perspectives.
I anticipate that some people will object.
They may call the presentations artificial, inauthentic, dishonest or even deceitful.
Indeed, given some of the recent commentary about AI-generated content, I wouldn’t be surprised if those criticisms came from people with whom I’ve already had disagreements about Excel education.
But before anyone dismisses these productions simply because the presenters are AI avatars, I would like to explain why I am using them.
And, more importantly, why I believe the information they will be communicating deserves serious attention.
Because there is an interesting history behind this decision.
And it raises a rather uncomfortable question.
If I present important information personally, backed by thirty years of professional experience, and that information is ignored, does presenting the same information through an AI avatar somehow make it less true?
Or should we be examining the information rather than the person—or avatar—delivering it?
1. A Video I Made Three Years Ago
In the summer of 2023, I made a video specifically for Excel educator Wyn Hopkins.
It concerned cascading dropdown lists, a technique that was receiving considerable attention in the Excel community at the time.
The video was never published.
I subsequently developed what I considered a more useful demonstration, the REG Call Handler, which illustrated the same underlying principles in a practical enterprise application.
But the original video made a point that I believe is fundamental to understanding Excel.
It was a distinction that has informed much of my professional career.
There is an enormous difference between using Excel as a personal productivity document and using Excel as part of an enterprise business process.
And the techniques appropriate to one are not necessarily appropriate to the other.
This distinction is at the heart of what I now call the 10%–90% divide.
The 10%: Excel as an Editable Document
Consider the spreadsheet created for one person’s use.
It contains data, formulas, perhaps some dropdown lists, charts and reports.
The user creates it, edits it, saves it and uses the results.
The spreadsheet is essentially the end of the process.
There are no significant downstream dependencies. Nobody else needs to interact with its data. It is not an operational component of a larger business system.
The spreadsheet might occasionally be sent to a manager as evidence that some work has been completed.
But nothing else depends on it.
This is perfectly legitimate Excel usage.
It is particularly appropriate for individuals, hobbyists, sole traders and many personal productivity tasks.
And a considerable amount of Excel education on social media is directed towards this type of usage.
There is nothing inherently wrong with that.
The problem begins when techniques designed for this environment are applied to enterprise processes without recognising the fundamentally different requirements.
The 90%: Excel as Part of an Enterprise Process
In an enterprise, a spreadsheet is often not the end of the process.
It is part of an ongoing business operation.
Other people need its information. Other departments depend on its outputs. Transactions and master data change continuously.
A spreadsheet may be one of hundreds or thousands participating in the same business process.
And this changes everything.
Data must be accessible, current, consistent and appropriately controlled.
The spreadsheets must work together.
The architecture must support the business process rather than simply the individual document.
Yet this distinction is rarely discussed in mainstream Excel education.
And that omission has enormous consequences.
2. Cricket, Football and the Art of Bending a Ball
In my 2023 video, I used an analogy to illustrate this distinction.
In cricket, a skilled spin bowler can make a ball turn dramatically.
In football, David Beckham became famous for bending a football around defenders and into the goal.
In both cases, we are watching a ball change direction.
Superficially, the techniques have something in common.
But nobody would suggest that a cricket spin bowler could automatically teach David Beckham how to take a free kick.
The sports are different.
The objectives are different.
The mechanics are different.
The surrounding conditions are different.
A superficial similarity does not establish functional equivalence.
Now consider cascading dropdown lists in Excel.
In a personal spreadsheet, the data supplying those dropdowns might reside in another worksheet within the same workbook.
Perfectly sensible.
But in an enterprise application, the data might represent products, inventory, customers, warehouses or other information that is constantly changing.
That information cannot sensibly be maintained independently in hundreds of spreadsheets.
It needs to be held centrally, where it can be updated and accessed by every participating spreadsheet.
The dropdown might look identical to the user.
But the architecture behind it is fundamentally different.
One is a self-contained document.
The other is a component of an enterprise information system.
Same appearance. Different purpose. Different architecture.
That was the message I wanted to communicate three years ago.
And it remains central to everything I am teaching today.
3. The Missing Transformation: From 10% to 90%
Here is where the problem becomes serious.
Excel educators teach techniques that are perfectly appropriate for personal spreadsheets.
Their audiences learn those techniques.
Some of those people then take the same techniques into their workplaces and attempt to apply them to enterprise-scale processes.
They build spreadsheets that depend on other spreadsheets.
They distribute copies.
They introduce external links.
They create increasingly elaborate mechanisms for gathering, reconciling and consolidating information.
Eventually, the organisation finds itself struggling with multiple versions, broken links, inconsistent data and endless reconciliation.
I call this the swamp.
The industry has another name for it.
Excel Hell.
And the Excel replacement industry has been exploiting Excel Hell for years.
Companies such as Anaplan, Workday Adaptive Planning, Planful and Datarails promote alternatives to Excel-based processes by highlighting precisely these problems.
Their argument is that Excel is unsuitable for enterprise-scale operations.
But I have spent thirty years demonstrating something very different.
The problem is not necessarily Excel.
The problem is the architecture.
And, more specifically, the failure to recognise that the methods appropriate to a standalone personal spreadsheet cannot simply be extended to an enterprise process.
Excel already possesses capabilities that enable it to participate in properly designed enterprise systems.
It can communicate with relational databases.
It can GET information from a central source.
It can PUT transactions and updates back into that source.
It can participate in processes involving hundreds or thousands of users and spreadsheets.
These are not new capabilities.
They have been available for decades.
The missing transformation is not a new Excel feature. It is a change in thinking.
And that change in thinking is almost entirely absent from mainstream Excel education.
4. What Happened When I Explained This Personally?
This brings me back to Wyn Hopkins.
I explained the distinction to him personally in 2023.
I created a video in which I appeared on camera.
There was no AI avatar.
No synthetic voice.
No fictional character.
It was simply me, drawing upon my professional experience, explaining why the difference between personal and enterprise spreadsheet techniques matters.
As I recall, he regarded the distinction as dismissive of the work being taught.
But I was not dismissing the techniques.
I was identifying their appropriate context.
There is a world of difference between those two positions.
My concern is that, from the publicly available educational material I have seen since then, I have not observed a corresponding emphasis on this crucial architectural distinction.
The teaching continues to focus predominantly on techniques for the standalone spreadsheet.
And the audience is left without a clear warning about what happens when those techniques are carried into enterprise processes.
Now, I cannot know precisely what another person has understood or why they choose to teach particular subjects.
But I can ask a reasonable question.
If this distinction is important enough to prevent Excel Hell, why is it not being taught?
And if I have already explained it personally, what exactly would be achieved by insisting that I must continue to present everything personally?
That is one reason I am experimenting with a different medium.
5. The Database That Apparently Takes a Year to Obtain
Let me offer another example.
Mark Proctor has argued that citizen developers face enormous obstacles when trying to obtain a database from their IT departments.
Requests can take months, perhaps a year.
Then there are permissions, change management procedures and other organisational barriers.
I understand that such obstacles exist in some organisations.
But the suggestion that a business user cannot have access to a relational database without going through this process does not reflect my experience.
In fact, it overlooks something rather significant.
Microsoft Access is a relational database application that has been part of the Microsoft Office family for decades.
And Excel can communicate with it.
In my demonstrations, I show how a database can be created and used with remarkable ease.
Creating a simple local database is not the same as obtaining approval for a production enterprise system, of course. Governance, security and deployment requirements must be considered.
But that is precisely why the distinction matters.
The technical capability and the organisational approval process are two different things.
One should not be confused with the other.
And yet, how often do we see this distinction clearly explained to Excel users?
I have demonstrated these capabilities.
I have explained the history.
I have shown practical examples.
But I have not seen a public correction from Mark acknowledging the broader availability of these options.
Again, the issue is not whether Mark or I should win an argument.
The issue is whether the audience is receiving an accurate understanding of the available technology.
6. The REG Call Handler: Apparently, Excel Cannot Do That
Another example concerns Paul Barnhurst.
In discussions about my REG Call Handler demonstration, Paul argued that the requirements called for a proper system rather than Excel.
But that was exactly the point of my demonstration.
I was showing that Excel could be the user-facing component of a properly designed system.
Not a standalone spreadsheet attempting to behave like an enterprise application.
Not hundreds of workbooks exchanging copies of data.
But Excel leveraging a relational database to extend its capabilities.
The distinction is fundamental.
And the solution was not hypothetical.
I had demonstrated it.
If somebody sees evidence that challenges an earlier statement, surely the appropriate response is to examine the evidence.
And, if necessary, revise the earlier statement.
There is no shame in learning something new.
There is considerable value in sharing that learning with an audience.
Yet I have not seen the corresponding public correction.
Which raises a question that goes far beyond Excel.
What responsibility does an educator have to correct information when presented with evidence that challenges what they have previously taught?
7. Power Query, VBA and Another Misunderstanding
The same problem appears in discussions about Power Query and VBA.
We regularly encounter statements along the lines of:
“Since Power Query arrived, I haven’t needed VBA.”
That may accurately describe someone’s personal working practices.
But it does not establish that Power Query replaces VBA.
One reason for the confusion, I suspect, is the superficial similarity between Power Query’s Applied Steps and the sequential instructions produced by Excel’s macro recorder.
Both can describe a series of operations.
But VBA is much more than a means of recording repetitive actions.
It provides access to Excel’s object model, supports application logic, responds to events, communicates with external systems and can orchestrate entire business processes.
Power Query has a different primary purpose.
It is principally concerned with connecting to, transforming and loading data.
These technologies overlap in some tasks, but they are not functionally equivalent.
In my generation, many of us learned Excel’s object model by studying the VBA generated by the macro recorder.
It was an extraordinarily useful learning technique.
But the macro recorder was never the definition of VBA.
And Power Query’s Applied Steps are not a substitute for the wider programming and automation capabilities that VBA provides.
Again, we are confusing two things because they look superficially similar.
Just as we confused the cricket ball with the football.
Just as we confuse a personal spreadsheet with an enterprise spreadsheet.
8. Power Query Is Not a Relational Database
I have encountered similar misunderstandings in discussions with Christopher de Fennell and others.
Power Query can work with tables.
It can perform joins.
It can transform and combine data.
But that does not make it a relational database management system.
Nor does it make a Power Query transformation a substitute for a transactional system of record.
In my enterprise Excel architecture, the relational database has a specific purpose.
It holds shared business data.
It supports transactions.
It provides a central source from which spreadsheets retrieve information and to which authorised users submit updates.
That is the foundation of the GET/PUT architecture I have used for decades.
Power Query can certainly participate in a wider solution.
But it does not replace the transactional role of the database.
I have explained this distinction repeatedly, including in direct discussions.
Yet I continue to see comparisons that blur these fundamentally different functions.
Which brings me to the central issue.
9. Human Authenticity Versus Factual Authenticity
For years, I have explained these concepts personally.
I have written articles.
I have responded to LinkedIn posts.
I have created demonstrations.
I have shown working solutions.
I have drawn upon thirty years of experience designing enterprise Excel systems, often for senior finance executives.
And I have repeatedly encountered situations where the information I presented appeared to make little difference to the public narrative.
Now I propose to communicate those same ideas through AI-generated characters.
And some people may object that the presentations are not authentic.
But what exactly do we mean by authenticity?
There are at least two separate questions.
First: Is the presenter a real human being?
Second: Is the information being presented accurate?
These are not the same question.
A real person can present misleading information.
An AI-generated character can present information that is accurate, demonstrable and supported by evidence.
Equally, AI-generated content can be wrong, misleading or fabricated. The technology provides no guarantee of truth.
The appropriate test is therefore the same in both cases.
Examine the claims.
Examine the evidence.
Examine the reasoning.
And, wherever possible, test the demonstration.
I am not proposing to use AI characters to impersonate real people or to deceive viewers about who is speaking.
They will be openly presented as fictional characters.
Their dialogue will be written and directed by me.
The ideas will be mine.
The case studies will be mine.
The demonstrations will be based on working solutions.
The AI will provide the voices, faces and dramatic presentation.
It is a production technique, not a substitute for evidence.
And that distinction deserves to be understood.
10. Why Tim and Reg? Why Mission Impossible?
There is another reason I am developing these productions.
Some subjects are more effectively explained through conversation than through a lecture.
Consider Tim and Reg.
Tim represents the capable Excel user who has learned the techniques commonly taught in the Excel community.
He is not incompetent.
He is not foolish.
He is working with the knowledge available to him.
Reg brings a different perspective.
He understands Excel as a component of a wider business process.
He introduces the concepts of shared data, relational databases, GET/PUT and enterprise architecture.
The conversation allows the audience to discover the difference gradually.
Tim asks the questions that a real Excel professional might ask.
Reg challenges assumptions.
And the audience follows the reasoning.
My forthcoming Mission Impossible series extends this approach.
Jim provides leadership and asks the practical questions.
Cinnamon understands Excel education and training.
Barney explains the technology.
Rollin draws upon experience and challenges conventional thinking, often rather forcefully.
Willy contributes the practical perspective needed to turn plans into action.
The contrasting personalities create opportunities for disagreement, clarification and discovery.
Instead of simply telling the audience what to think, I can dramatise the process of examining an assumption and testing it against reality.
And AI production tools make this possible without assembling a cast of professional actors, hiring studios or managing a conventional film production.
I can concentrate on the script, the reasoning and the evidence.
That is an extraordinary opportunity for an independent educator.
11. Who Should Decide Whether This Matters?
There is a bigger question behind all this.
Is the 10%–90% distinction important?
Does it matter that enterprise spreadsheets require a different architecture from personal spreadsheets?
Does it matter that businesses can use Excel with relational databases to build shared, controlled and scalable processes?
Does it matter that techniques which work beautifully in isolation may create serious problems when extended across an organisation?
I believe these questions should be considered by the people responsible for the consequences.
Finance directors.
CFOs.
Business leaders.
Senior managers.
People responsible for budgets, operations, compliance, reporting and information systems.
They are the people who experience Excel Hell.
They are the people who are told that Excel is the problem.
They are the people who may be asked to spend substantial sums replacing Excel-based processes with proprietary applications.
And they are the people who should be able to examine the alternatives.
If the architecture I am demonstrating can prevent the problems they face, or help them escape from those problems, that is commercially important.
If it cannot, then the evidence should establish that.
But the question should not be settled by the preferences of social media educators.
Nor should it be settled by whether they approve of AI-generated presenters.
The question is whether the information is correct, relevant and useful to the business.
That is the test that matters.
12. The Paradox of Misinformation
There is an irony in all of this.
The misunderstandings I have been describing are themselves part of the reason why enterprise Excel expertise can be so valuable.
When organisations are taught to think of Excel primarily as an editable document, they naturally attempt to solve enterprise problems using document-oriented techniques.
When those techniques fail at scale, the resulting problems are attributed to Excel.
The Excel replacement industry then offers expensive alternatives.
But a different approach to Excel architecture can remove many of the causes of those problems.
That is precisely the opportunity I discovered in my professional career.
In several organisations, introducing that different way of thinking resulted in extraordinary commercial outcomes, including substantial increases in my consulting rates.
The value did not come from knowing more spreadsheet functions.
It came from understanding the business process and using the available technology differently.
In that sense, the widespread misunderstanding creates an opportunity.
But it is an expensive opportunity for the organisations that remain trapped in Excel Hell.
And it is a missed opportunity for Excel professionals who have never been taught the enterprise techniques that could transform their careers.
I would much rather see those capabilities become common professional knowledge.
That is why I am creating these educational productions.
13. So, Is Using an AI Avatar Dishonest?
Let me return to the anticipated criticism.
Suppose somebody watches a Tim and Reg episode and says:
“Why are you using AI? Why don’t you explain this yourself?”
My answer is simple.
I already have.
I have explained it personally.
I have demonstrated it personally.
I have written about it.
I have discussed it with educators and industry commentators.
I have spent decades putting these principles into practice.
Now I am exploring another way to communicate the same information.
A way that allows characters to ask questions, challenge assumptions, disagree, explain and demonstrate.
A way that may make complex ideas easier to understand.
A way that can reach people who would never sit through a lengthy technical lecture.
If the objection is that viewers might mistake a fictional AI character for a real person, I agree that transparency matters.
I will make it clear that these are AI-generated characters.
But if the objection is that information becomes less credible simply because an AI avatar delivers it, I reject that argument.
The information does not change its truth value when the presenter changes.
A false statement does not become true because a respected human educator says it.
And a true statement does not become false because a fictional AI character says it.
The responsibility remains with the author to ensure that the information is accurate, and with the audience to evaluate it critically.
Conclusion: Judge the Message, Not the Messenger
The Excel community has spent decades teaching people how to work with spreadsheets.
Much of that teaching has been excellent within its intended context.
But the transformation from personal spreadsheet usage to enterprise spreadsheet architecture remains largely missing from mainstream education.
That missing transformation has consequences.
It contributes to the proliferation of disconnected spreadsheets, duplicated data, reconciliation problems and fragile processes.
It helps create the conditions that the industry calls Excel Hell.
And it sustains a market for products promoted as replacements for Excel.
Yet the capabilities needed to address many of these problems have existed within the Microsoft technology ecosystem for decades.
That is the story I want to tell.
It is a story about technology.
But it is also a story about education, professional responsibility, the spread of misinformation and the reluctance to challenge familiar assumptions.
I have told parts of that story personally for years.
Now Tim, Reg, Jim, Cinnamon, Barney, Rollin and Willy will help me tell it differently.
They are fictional characters.
The productions will use AI.
But the problems are real.
The demonstrations are real.
The professional experience behind them is real.
And the evidence can be examined.
So, before anyone dismisses the series because the presenters are not human, I would invite them to consider a more useful question.
If the information can help businesses escape Excel Hell, does it really matter whether the person explaining it is standing in front of a camera—or is an AI avatar?
And perhaps an even more important question:
If the same information has already been presented by a human being and ignored, shouldn’t we be asking why?
That, ultimately, is the rationale behind my AI productions.
The messenger has changed.
The message has not.
And the evidence still deserves to be heard.
Hiran de Silva
Former CFO | Enterprise Excel Consultant | Author of Triple Your Pay with Excel
Using AI Avatars to Explain the 10%–90% Divide
The Rationale: When the Human Voice Is Ignored, Does It Matter Who Delivers the Message?
By Hiran de Silva
October 2026
I am about to launch a series of video presentations featuring AI-generated characters.
Some will be explainers. Others will be conversations, debates and dramatisations. They will include my Tim and Reg series and my forthcoming Excel Mission Impossible productions, featuring five characters with different personalities, professional backgrounds and perspectives.
I anticipate that some people will object.
They may call the presentations artificial, inauthentic, dishonest or even deceitful.
Indeed, given some of the recent commentary about AI-generated content, I wouldn’t be surprised if those criticisms came from people with whom I’ve already had disagreements about Excel education.
But before anyone dismisses these productions simply because the presenters are AI avatars, I would like to explain why I am using them.
And, more importantly, why I believe the information they will be communicating deserves serious attention.
Because there is an interesting history behind this decision.
And it raises a rather uncomfortable question.
If I present important information personally, backed by thirty years of professional experience, and that information is ignored, does presenting the same information through an AI avatar somehow make it less true?
Or should we be examining the information rather than the person—or avatar—delivering it?
1. A Video I Made Three Years Ago
In the summer of 2023, I made a video specifically for Excel educator Wyn Hopkins.
It concerned cascading dropdown lists, a technique that was receiving considerable attention in the Excel community at the time.
The video was never published.
I subsequently developed what I considered a more useful demonstration, the REG Call Handler, which illustrated the same underlying principles in a practical enterprise application.
But the original video made a point that I believe is fundamental to understanding Excel.
It was a distinction that has informed much of my professional career.
There is an enormous difference between using Excel as a personal productivity document and using Excel as part of an enterprise business process.
And the techniques appropriate to one are not necessarily appropriate to the other.
This distinction is at the heart of what I now call the 10%–90% divide.
The 10%: Excel as an Editable Document
Consider the spreadsheet created for one person’s use.
It contains data, formulas, perhaps some dropdown lists, charts and reports.
The user creates it, edits it, saves it and uses the results.
The spreadsheet is essentially the end of the process.
There are no significant downstream dependencies. Nobody else needs to interact with its data. It is not an operational component of a larger business system.
The spreadsheet might occasionally be sent to a manager as evidence that some work has been completed.
But nothing else depends on it.
This is perfectly legitimate Excel usage.
It is particularly appropriate for individuals, hobbyists, sole traders and many personal productivity tasks.
And a considerable amount of Excel education on social media is directed towards this type of usage.
There is nothing inherently wrong with that.
The problem begins when techniques designed for this environment are applied to enterprise processes without recognising the fundamentally different requirements.
The 90%: Excel as Part of an Enterprise Process
In an enterprise, a spreadsheet is often not the end of the process.
It is part of an ongoing business operation.
Other people need its information. Other departments depend on its outputs. Transactions and master data change continuously.
A spreadsheet may be one of hundreds or thousands participating in the same business process.
And this changes everything.
Data must be accessible, current, consistent and appropriately controlled.
The spreadsheets must work together.
The architecture must support the business process rather than simply the individual document.
Yet this distinction is rarely discussed in mainstream Excel education.
And that omission has enormous consequences.
2. Cricket, Football and the Art of Bending a Ball
In my 2023 video, I used an analogy to illustrate this distinction.
In cricket, a skilled spin bowler can make a ball turn dramatically.
In football, David Beckham became famous for bending a football around defenders and into the goal.
In both cases, we are watching a ball change direction.
Superficially, the techniques have something in common.
But nobody would suggest that a cricket spin bowler could automatically teach David Beckham how to take a free kick.
The sports are different.
The objectives are different.
The mechanics are different.
The surrounding conditions are different.
A superficial similarity does not establish functional equivalence.
Now consider cascading dropdown lists in Excel.
In a personal spreadsheet, the data supplying those dropdowns might reside in another worksheet within the same workbook.
Perfectly sensible.
But in an enterprise application, the data might represent products, inventory, customers, warehouses or other information that is constantly changing.
That information cannot sensibly be maintained independently in hundreds of spreadsheets.
It needs to be held centrally, where it can be updated and accessed by every participating spreadsheet.
The dropdown might look identical to the user.
But the architecture behind it is fundamentally different.
One is a self-contained document.
The other is a component of an enterprise information system.
Same appearance. Different purpose. Different architecture.
That was the message I wanted to communicate three years ago.
And it remains central to everything I am teaching today.
3. The Missing Transformation: From 10% to 90%
Here is where the problem becomes serious.
Excel educators teach techniques that are perfectly appropriate for personal spreadsheets.
Their audiences learn those techniques.
Some of those people then take the same techniques into their workplaces and attempt to apply them to enterprise-scale processes.
They build spreadsheets that depend on other spreadsheets.
They distribute copies.
They introduce external links.
They create increasingly elaborate mechanisms for gathering, reconciling and consolidating information.
Eventually, the organisation finds itself struggling with multiple versions, broken links, inconsistent data and endless reconciliation.
I call this the swamp.
The industry has another name for it.
Excel Hell.
And the Excel replacement industry has been exploiting Excel Hell for years.
Companies such as Anaplan, Workday Adaptive Planning, Planful and Datarails promote alternatives to Excel-based processes by highlighting precisely these problems.
Their argument is that Excel is unsuitable for enterprise-scale operations.
But I have spent thirty years demonstrating something very different.
The problem is not necessarily Excel.
The problem is the architecture.
And, more specifically, the failure to recognise that the methods appropriate to a standalone personal spreadsheet cannot simply be extended to an enterprise process.
Excel already possesses capabilities that enable it to participate in properly designed enterprise systems.
It can communicate with relational databases.
It can GET information from a central source.
It can PUT transactions and updates back into that source.
It can participate in processes involving hundreds or thousands of users and spreadsheets.
These are not new capabilities.
They have been available for decades.
The missing transformation is not a new Excel feature. It is a change in thinking.
And that change in thinking is almost entirely absent from mainstream Excel education.
4. What Happened When I Explained This Personally?
This brings me back to Wyn Hopkins.
I explained the distinction to him personally in 2023.
I created a video in which I appeared on camera.
There was no AI avatar.
No synthetic voice.
No fictional character.
It was simply me, drawing upon my professional experience, explaining why the difference between personal and enterprise spreadsheet techniques matters.
As I recall, he regarded the distinction as dismissive of the work being taught.
But I was not dismissing the techniques.
I was identifying their appropriate context.
There is a world of difference between those two positions.
My concern is that, from the publicly available educational material I have seen since then, I have not observed a corresponding emphasis on this crucial architectural distinction.
The teaching continues to focus predominantly on techniques for the standalone spreadsheet.
And the audience is left without a clear warning about what happens when those techniques are carried into enterprise processes.
Now, I cannot know precisely what another person has understood or why they choose to teach particular subjects.
But I can ask a reasonable question.
If this distinction is important enough to prevent Excel Hell, why is it not being taught?
And if I have already explained it personally, what exactly would be achieved by insisting that I must continue to present everything personally?
That is one reason I am experimenting with a different medium.
5. The Database That Apparently Takes a Year to Obtain
Let me offer another example.
Mark Proctor has argued that citizen developers face enormous obstacles when trying to obtain a database from their IT departments.
Requests can take months, perhaps a year.
Then there are permissions, change management procedures and other organisational barriers.
I understand that such obstacles exist in some organisations.
But the suggestion that a business user cannot have access to a relational database without going through this process does not reflect my experience.
In fact, it overlooks something rather significant.
Microsoft Access is a relational database application that has been part of the Microsoft Office family for decades.
And Excel can communicate with it.
In my demonstrations, I show how a database can be created and used with remarkable ease.
Creating a simple local database is not the same as obtaining approval for a production enterprise system, of course. Governance, security and deployment requirements must be considered.
But that is precisely why the distinction matters.
The technical capability and the organisational approval process are two different things.
One should not be confused with the other.
And yet, how often do we see this distinction clearly explained to Excel users?
I have demonstrated these capabilities.
I have explained the history.
I have shown practical examples.
But I have not seen a public correction from Mark acknowledging the broader availability of these options.
Again, the issue is not whether Mark or I should win an argument.
The issue is whether the audience is receiving an accurate understanding of the available technology.
6. The REG Call Handler: Apparently, Excel Cannot Do That
Another example concerns Paul Barnhurst.
In discussions about my REG Call Handler demonstration, Paul argued that the requirements called for a proper system rather than Excel.
But that was exactly the point of my demonstration.
I was showing that Excel could be the user-facing component of a properly designed system.
Not a standalone spreadsheet attempting to behave like an enterprise application.
Not hundreds of workbooks exchanging copies of data.
But Excel leveraging a relational database to extend its capabilities.
The distinction is fundamental.
And the solution was not hypothetical.
I had demonstrated it.
If somebody sees evidence that challenges an earlier statement, surely the appropriate response is to examine the evidence.
And, if necessary, revise the earlier statement.
There is no shame in learning something new.
There is considerable value in sharing that learning with an audience.
Yet I have not seen the corresponding public correction.
Which raises a question that goes far beyond Excel.
What responsibility does an educator have to correct information when presented with evidence that challenges what they have previously taught?
7. Power Query, VBA and Another Misunderstanding
The same problem appears in discussions about Power Query and VBA.
We regularly encounter statements along the lines of:
“Since Power Query arrived, I haven’t needed VBA.”
That may accurately describe someone’s personal working practices.
But it does not establish that Power Query replaces VBA.
One reason for the confusion, I suspect, is the superficial similarity between Power Query’s Applied Steps and the sequential instructions produced by Excel’s macro recorder.
Both can describe a series of operations.
But VBA is much more than a means of recording repetitive actions.
It provides access to Excel’s object model, supports application logic, responds to events, communicates with external systems and can orchestrate entire business processes.
Power Query has a different primary purpose.
It is principally concerned with connecting to, transforming and loading data.
These technologies overlap in some tasks, but they are not functionally equivalent.
In my generation, many of us learned Excel’s object model by studying the VBA generated by the macro recorder.
It was an extraordinarily useful learning technique.
But the macro recorder was never the definition of VBA.
And Power Query’s Applied Steps are not a substitute for the wider programming and automation capabilities that VBA provides.
Again, we are confusing two things because they look superficially similar.
Just as we confused the cricket ball with the football.
Just as we confuse a personal spreadsheet with an enterprise spreadsheet.
8. Power Query Is Not a Relational Database
I have encountered similar misunderstandings in discussions with Christopher de Fennell and others.
Power Query can work with tables.
It can perform joins.
It can transform and combine data.
But that does not make it a relational database management system.
Nor does it make a Power Query transformation a substitute for a transactional system of record.
In my enterprise Excel architecture, the relational database has a specific purpose.
It holds shared business data.
It supports transactions.
It provides a central source from which spreadsheets retrieve information and to which authorised users submit updates.
That is the foundation of the GET/PUT architecture I have used for decades.
Power Query can certainly participate in a wider solution.
But it does not replace the transactional role of the database.
I have explained this distinction repeatedly, including in direct discussions.
Yet I continue to see comparisons that blur these fundamentally different functions.
Which brings me to the central issue.
9. Human Authenticity Versus Factual Authenticity
For years, I have explained these concepts personally.
I have written articles.
I have responded to LinkedIn posts.
I have created demonstrations.
I have shown working solutions.
I have drawn upon thirty years of experience designing enterprise Excel systems, often for senior finance executives.
And I have repeatedly encountered situations where the information I presented appeared to make little difference to the public narrative.
Now I propose to communicate those same ideas through AI-generated characters.
And some people may object that the presentations are not authentic.
But what exactly do we mean by authenticity?
There are at least two separate questions.
First: Is the presenter a real human being?
Second: Is the information being presented accurate?
These are not the same question.
A real person can present misleading information.
An AI-generated character can present information that is accurate, demonstrable and supported by evidence.
Equally, AI-generated content can be wrong, misleading or fabricated. The technology provides no guarantee of truth.
The appropriate test is therefore the same in both cases.
Examine the claims.
Examine the evidence.
Examine the reasoning.
And, wherever possible, test the demonstration.
I am not proposing to use AI characters to impersonate real people or to deceive viewers about who is speaking.
They will be openly presented as fictional characters.
Their dialogue will be written and directed by me.
The ideas will be mine.
The case studies will be mine.
The demonstrations will be based on working solutions.
The AI will provide the voices, faces and dramatic presentation.
It is a production technique, not a substitute for evidence.
And that distinction deserves to be understood.
10. Why Tim and Reg? Why Mission Impossible?
There is another reason I am developing these productions.
Some subjects are more effectively explained through conversation than through a lecture.
Consider Tim and Reg.
Tim represents the capable Excel user who has learned the techniques commonly taught in the Excel community.
He is not incompetent.
He is not foolish.
He is working with the knowledge available to him.
Reg brings a different perspective.
He understands Excel as a component of a wider business process.
He introduces the concepts of shared data, relational databases, GET/PUT and enterprise architecture.
The conversation allows the audience to discover the difference gradually.
Tim asks the questions that a real Excel professional might ask.
Reg challenges assumptions.
And the audience follows the reasoning.
My forthcoming Mission Impossible series extends this approach.
Jim provides leadership and asks the practical questions.
Cinnamon understands Excel education and training.
Barney explains the technology.
Rollin draws upon experience and challenges conventional thinking, often rather forcefully.
Willy contributes the practical perspective needed to turn plans into action.
The contrasting personalities create opportunities for disagreement, clarification and discovery.
Instead of simply telling the audience what to think, I can dramatise the process of examining an assumption and testing it against reality.
And AI production tools make this possible without assembling a cast of professional actors, hiring studios or managing a conventional film production.
I can concentrate on the script, the reasoning and the evidence.
That is an extraordinary opportunity for an independent educator.
11. Who Should Decide Whether This Matters?
There is a bigger question behind all this.
Is the 10%–90% distinction important?
Does it matter that enterprise spreadsheets require a different architecture from personal spreadsheets?
Does it matter that businesses can use Excel with relational databases to build shared, controlled and scalable processes?
Does it matter that techniques which work beautifully in isolation may create serious problems when extended across an organisation?
I believe these questions should be considered by the people responsible for the consequences.
Finance directors.
CFOs.
Business leaders.
Senior managers.
People responsible for budgets, operations, compliance, reporting and information systems.
They are the people who experience Excel Hell.
They are the people who are told that Excel is the problem.
They are the people who may be asked to spend substantial sums replacing Excel-based processes with proprietary applications.
And they are the people who should be able to examine the alternatives.
If the architecture I am demonstrating can prevent the problems they face, or help them escape from those problems, that is commercially important.
If it cannot, then the evidence should establish that.
But the question should not be settled by the preferences of social media educators.
Nor should it be settled by whether they approve of AI-generated presenters.
The question is whether the information is correct, relevant and useful to the business.
That is the test that matters.
12. The Paradox of Misinformation
There is an irony in all of this.
The misunderstandings I have been describing are themselves part of the reason why enterprise Excel expertise can be so valuable.
When organisations are taught to think of Excel primarily as an editable document, they naturally attempt to solve enterprise problems using document-oriented techniques.
When those techniques fail at scale, the resulting problems are attributed to Excel.
The Excel replacement industry then offers expensive alternatives.
But a different approach to Excel architecture can remove many of the causes of those problems.
That is precisely the opportunity I discovered in my professional career.
In several organisations, introducing that different way of thinking resulted in extraordinary commercial outcomes, including substantial increases in my consulting rates.
The value did not come from knowing more spreadsheet functions.
It came from understanding the business process and using the available technology differently.
In that sense, the widespread misunderstanding creates an opportunity.
But it is an expensive opportunity for the organisations that remain trapped in Excel Hell.
And it is a missed opportunity for Excel professionals who have never been taught the enterprise techniques that could transform their careers.
I would much rather see those capabilities become common professional knowledge.
That is why I am creating these educational productions.
13. So, Is Using an AI Avatar Dishonest?
Let me return to the anticipated criticism.
Suppose somebody watches a Tim and Reg episode and says:
“Why are you using AI? Why don’t you explain this yourself?”
My answer is simple.
I already have.
I have explained it personally.
I have demonstrated it personally.
I have written about it.
I have discussed it with educators and industry commentators.
I have spent decades putting these principles into practice.
Now I am exploring another way to communicate the same information.
A way that allows characters to ask questions, challenge assumptions, disagree, explain and demonstrate.
A way that may make complex ideas easier to understand.
A way that can reach people who would never sit through a lengthy technical lecture.
If the objection is that viewers might mistake a fictional AI character for a real person, I agree that transparency matters.
I will make it clear that these are AI-generated characters.
But if the objection is that information becomes less credible simply because an AI avatar delivers it, I reject that argument.
The information does not change its truth value when the presenter changes.
A false statement does not become true because a respected human educator says it.
And a true statement does not become false because a fictional AI character says it.
The responsibility remains with the author to ensure that the information is accurate, and with the audience to evaluate it critically.
Conclusion: Judge the Message, Not the Messenger
The Excel community has spent decades teaching people how to work with spreadsheets.
Much of that teaching has been excellent within its intended context.
But the transformation from personal spreadsheet usage to enterprise spreadsheet architecture remains largely missing from mainstream education.
That missing transformation has consequences.
It contributes to the proliferation of disconnected spreadsheets, duplicated data, reconciliation problems and fragile processes.
It helps create the conditions that the industry calls Excel Hell.
And it sustains a market for products promoted as replacements for Excel.
Yet the capabilities needed to address many of these problems have existed within the Microsoft technology ecosystem for decades.
That is the story I want to tell.
It is a story about technology.
But it is also a story about education, professional responsibility, the spread of misinformation and the reluctance to challenge familiar assumptions.
I have told parts of that story personally for years.
Now Tim, Reg, Jim, Cinnamon, Barney, Rollin and Willy will help me tell it differently.
They are fictional characters.
The productions will use AI.
But the problems are real.
The demonstrations are real.
The professional experience behind them is real.
And the evidence can be examined.
So, before anyone dismisses the series because the presenters are not human, I would invite them to consider a more useful question.
If the information can help businesses escape Excel Hell, does it really matter whether the person explaining it is standing in front of a camera—or is an AI avatar?
And perhaps an even more important question:
If the same information has already been presented by a human being and ignored, shouldn’t we be asking why?
That, ultimately, is the rationale behind my AI productions.
The messenger has changed.
The message has not.
And the evidence still deserves to be heard.
Hiran de Silva
Former CFO | Enterprise Excel Consultant | Author of Triple Your Pay with Excel
Add comment