Digital Financial Reporting

IF we humans want machines to help us humans in our information exchange efforts; THEN we humans need to deliver meaningful, purposeful, actionable information to the point of need, at the right time, in an accessible, reusable format, and the entire process needs to be 100% controllable and understandable to the human system operators.

Whether the information exchange is one human exchanging information with another using a software application or one software application API exchanging information with another API; the above paragraph is true.

Fundamentally, accounting is an information exchange technology, in fact accounting was the world's first communications technology. Accounting has evolved before, it is evolving again; going through a great upheaval like many, most, others.

Note that we are talking about the exchange of information, not the exchange of data. Exchanging information is about the receiver being able to take action with that information.  The receiver wants to increase knowledge, insight, and wisdom.

Explainable artificial intelligence (XAI) helps humans automate tasks and processes.  Pacioli is explainable artificial intelligence.  You need Pacioli or something like Pacioli to make XBRL-based digital financial reporting to work.

The use case of financial reporting is such that the creators of reports are permitted to modify the report model of the information being exchanged within specific permitted boundaries. Software must be able to handle this business use case effectively, enabling business professionals to control the process and staying within those permitted boundaries.  Software needs to provide 'bumpers" as some people call it or "guard-rails" as it is called by others.

Auditchain's Pacioli does the above effectively. This fact is provable and can be demonstrated. Pacioli is the result of 20 years of effort.

I don't know exactly what to call Pacioli.  At first I called Pacioli a logic/rules/reasoning engine.  Then I referred to it as a knowledge engine.  Now I am contemplating that Pacioli is an insights engine.  For now, I consider Pacioli a logic/reasoning/rules/knowledge/insights engine in order to be complete I guess.

The rules and information Pacioli works with is declarative, global standard XBRL.  Article 4 of the Business Rules Manifesto states that rules should be declarative rather than procedural. Declarative is more flexible, reusable, and easier to maintain.

Merkle trees and Merkle proofs can be leveraged to make sure the declarative rules have not been tampered with and provide an audit trail.  IPFS provides reliability.  Pacioli leverages both Merkle trees and IPFS.

XBRL processors + XBRL formula processors are defecient when it comes to processing XBRL-based financial reports. Pacioli leverages the open source Arelle XBRL processor and XBRL formula processor; enhancing those capabilities and filling in the gaps between what is provided and what is needed for effective XBRL-based financial reporting.

The Standard Business Report Model (SBRM) of OMG provides a standard logical conceptualization of a business report.  A general purpose financial report is a specialization of a business report (a type of business report).  The Logical Theory Describing Financial Report specifically defines the logic of financial reports and is the basis for SBRM.  SBRM enables the creation of a digital alternative to the general purpose financial report which has been, historically, analog.  The Seattle Method, which is supported by Pacioli, implements the Standard Business Report Model.

But Pacioli, SBRM, XBRL, and the Seattle Method are not an end game; they are a beginning.  The starter set of rules provided for Pacioli are likewise only a beginning.  The XBRL-based global standard rule framework is an approach to implementing customizable but controllable financial reporting schemes.  The same approach can most likely also be used for general business reporting use cases.

There are two approaches to artificial intelligence and, as I have said before, the right approach should be used for the given job.  The two approaches are: 

• Rules-based systems (expert systems, three basic types)
  • Classification or diagnosis type: helps users of the system select from a set of given alternatives.
  • Construction type: helps users of the system assemble something from given primitive components.
  • Simulation type: helps users of the system understand how some model reacts to certain inputs.
• Patterns-based systems (machine learning which can be supervised or unsupervised, five basic type, this video explains the types)
  • Clustering algorithms: categorize or group things
  • Explanatory algorithms: explain the relationships between variables
  • Ensemble learning algorithms: use multiple models
  • Similarity algorithms: compute the similarity of pairs of things
  • Dimensionality reduction algorithms: reduces variables in a dataset

Currently, Pacioli has no machine learning capabilities.  I am no machine learning expert by any measure, but I suspect machine learning is coming at some point.  I do know that to make machine learning work you need training data and that the machine learning is only as good as its training data.

Who will put all these pieces together effectively?  Who will be first?  The answer to that question will revel itself over time.

My personal focus is getting a rules-based expert system built, see here, Expert System for Creating Financial Reports and Logical Schema of Financial Reports.  I know of at least 8 other individuals and groups that have an interest in this at different levels.  Some are focuses at the traction or business event level; others are focused on the complete record-to-report process.  Others are focused on providing an environment within which all this will operating including Auditchain and Twali.

Which path will you choose? Simple but wrong; or Complex but right?

Does all this seem like a mystery to you?  Well, it really is not that mysterious at all if one has been paying attention and has tried to make sense of all this (i.e. employ sensemaking).  I think the risk equation has flipped; it is more risky NOT to be paying attention to all this that it was to get in early and go down the wrong path.

###########################

Pesseract screen shots