John Holland developed genetic algorithms in the 1960s. This family of algorithms is based on natural selection and natural laws of genetics to solve optimization problems.
Accounting is fundamentally a binary system (debit-credit; asset-liability; plus-minus) – even the proposed Triple Entry system proposed by Ian Grigg is also binary.[i]
I think that my ‘triple entry – optimized entry’ is NOT binary system, and it is based on Yuji Ijiri’s paper (A Framework for Triple-Entry Bookkeeping, 1986).
“… framework under which details of triple entry bookkeeping are now being developed and presents a simple example that illustrates the interrelationship among dimensions in triple-entry bookkeeping.”[ii]
The recording of accounting transactions onto the blockchain is the easy part. The determination of the interrelationship among the transaction dimensions is the hard part that no one to date has addressed. I understand that when Grigg proposed his triple entry, he may not have this in mind, and I would love to ask him for comments.
This is the problem that I have with recording transactions onto the blockchain because I am doing it, and I don’t see it as a ‘killer app’ for accounting. It creates layers of complexity, all coming from the client-side with regards to (1) the public-private key, (2) unique address, and (3) ease of access.
There are other issues, for example:
- What if I don’t record every transaction on the blockchain, so how will the blockchain help? How often do you enter transactions onto QuickBooks?
- If I record all my transactions, and how will that help my vendor, staff, anyone?
- Ask yourself, if you can trace your bitcoin’s history, what is the purpose of the exercise?
- Another interesting question, how is triple entry an improvement over the closely held and secure tax and books being held today (Trump’s tax returns)?
I am a strong advocate of blockchain, and the blockchain is integrated into my automation system. Still, I think that the importance and utility of triple entry is the ‘interrelationship among dimensions’ laid out by Ijiri. Besides removing the mistrust, frauds, or manipulations by others, the interrelationship aspect is a prelude to machine learning, fraud detection, pattern recognition, and prediction. More to come.
Hint: Petri Net and convolutional neural network (CNN).[iii]
[i] https://iang.org/papers/triple_entry.html
[ii] https://www.gwern.net/docs/bitcoin/1986-ijiri.pdf
[iii] http://citeseerx.ist.psu.edu/viewdoc/download?rep=rep1&type=pdf&doi=10.1.1.218.2516