Remember when someone first told you that even-numbered highways generally run east-west and odd-numbered highways north-south?

A clever convention.

Without a compass, you instantly have a sense of direction.

Now think about the first chart of accounts you built in an ERP system—or even in QuickBooks.

1xxxx Assets.

2xxxx Liabilities.

3xxxx Equity.

Then some variation of 4, 5, 6, 7, 8, and 9 for revenue and expenses.

Again, clever.

Until it isn’t.

How many times have you carefully reserved number ranges, only to discover years later that you’ve run out of room? Another bank account. Another product line. Another legal entity. Suddenly your carefully organized cash accounts occupy three different ranges because the original block filled up years ago.

Or worse, committing the cardinal sin of reusing old account numbers, forever corrupting the historical record.

The longer a company lives, the harder it becomes to preserve the integrity of the original numbering strategy.

Anyone who has integrated acquisitions knows this pain all too well.

One acquired company numbers cash accounts in the 1000 series.

Another uses the 10000 series.

A third embeds departments, locations, or business units directly into the account number itself.

Each scheme made perfect sense—until they had to coexist. Suddenly you’re not managing one elegant numbering philosophy. You’re trying to reconcile three or four competing definitions of what those numbers were supposed to mean.

I wonder if the telephone companies imagined they’d someday exhaust area codes.

Or whether the designers of Social Security numbers assumed nine digits would always be enough.

The schemes were brilliant.

They were simply designed for a different world.

Those embedded meanings exist for us—not for the database.

A modern ERP system doesn’t need account 10100 to know it’s Cash. It already has a chart of accounts table describing the account type, financial statement classification, reporting hierarchy, entity, currency, and every other attribute that gives the account meaning.

The intelligence lives in the account attributes—what database designers call metadata—not in the account number.

For decades we’ve designed numbering systems to help humans remember information. Increasingly, our systems don’t need that shortcut. AI certainly doesn’t. It understands attributes, labels, and relationships far better than it interprets clever numbering conventions.

Perhaps AI’s greatest value isn’t replacing people. It’s freeing us from legacy design assumptions.

Perhaps it’s time we stopped embedding business logic inside our identifiers.

So the next time you’re sitting in a conference room designing a new chart of accounts, and someone says, “That’s how we’ve always done it,” stop.

Step outside the accounting box.

Drag the conversation into the world of database design.

The chart of accounts is no longer a memory aid.

It’s a table.

The meaning belongs in the attributes.

The number is just an identifier.

Welcome to the simpler world AI unlocks.