Category: Technology

Creating Accounting Nirvana

Posted on by Phil Garland

I’ve written before about the possibility of a blockchain-centered accounting system that would automatically pinpoint assets, liabilities, revenues and expenses. This system, utilizing a financial code called XBRL (or something similar), would be able to tag every transaction that occurs in the world and match it to the correct company’s books. Not only would this automate the mundane tasks of bookkeeping but it would also provide valuable anonymized datasets to set benchmarking for good or bad companies.

Today, I read an article on TechCrunch about a bookkeeping company called Pilot that just raised $15 million in their Series B. Pilot is run by three MIT graduates and thrh seem to have figured out a way to integrate the many data points from banks to accounting systems to credit cards and combine them into an accounting system. It seems like accounting nirvana.  

I wish it were true. As much as I want to hope they’ve found a viable solution, it’s nearly impossible to do. We live in an ever-changing world with APIs that fall apart without consistent upkeep. Even Xero, probably the best accounting system I’ve seen for bank feeds, has its limitations based on API plugins and pulling data from banks. On top of that, you have to be constantly monitoring changes companies make to their UI. It’s an extremely complicated task to create the accounting nirvana today with our current systems. 

Additionally, I’ve never seen a company with perfect books. There will always be an error that has tossed the books into chaos and getting down to zero reconciliation is nearly impossible. There are ways of correcting the books but simply adding data on top of bad data won’t fix everything at once. 

I believe the solution is to re-think our systems from the ground up. Everything needs to be built with a systematic approach where everyone’s accounting system is in communication with one another. Integrating a bunch of datasets may seem like a solid idea in the short term but it isn’t sustainable in the long term. There are too many moving parts.

Accounting isn’t a sexy field. There probably aren’t many people who care about what I’m writing. But it’s these things that I see on a daily basis that will need to be solved to make our monetary system more efficient. There’s opportunity here, hopefully we find a way to seize it.

The Overblown Belief in Technology

Posted on by Phil Garland

I feel like everyday I see ten articles exclaiming “the end of human jobs” or “robots are taking over”. Although robotic improvements are happening at an exponential rate, robots are nowhere close to being able to take control of human production. The major missing link is an understanding of prices. Robots don’t operate under scarcity or supply and demand and they don’t have millions of thoughts racing through their neurons every single day. Robots are as good as we program them to be and anyone who has ever tried programming knows how hard it is to get right.

The leaps from artificial intelligence and machine learning should make our programming far more efficient as robots can learn faster. But that doesn’t mean humans will be without work. Just because technology advances doesn’t make manpower obsolete. We’re on the edge of a new wave, the technology era so-to-speak, and that can lead to a lot of fearmongering. We’re not being replaced by machines. If anything, we’re going to start growing exponentially smarter alongside the machines.

Ludwig von Mises, in Human Action, has a great quote on the inability of machines to replicate the human mind.

Technology operates with countable and measurable quantities of external things and effects; it knows causal relations between them, but it is foreign to their relevance to human wants and desires…Technology tells how a given end could be attained by the employment of various means which can be used together in various combinations, or how various available means could be employed for certain purposes. But it is at a loss to tell man which procedures he should choose out of the infinite variety of imaginable and possible modes of production.

A machine is not a human mind. It doesn’t have wants and desires. A machine doesn’t have to choose between millions of different ways to spend it’s time. Machines are fundamentally different than humans. (Note: Brain emulation, when possible, is something that might make human production near obsolete). Although people will continually debate the merits of artificial intelligence and technological progress, the human workforce isn’t going anywhere. Most likely we’ll become even more efficient with our technology counterparts.