Artificial Intelligence Explained for the Small Business Owner
As a certified public accountant serving many different businesses, I’ve heard countless business owners say some version of, “I run my business on the back of an envelope.” And for many small, simple enterprises, that approach actually works; key numbers and other important information scribbled down on paper, the rest stored in the owner’s head.
But as a business grows, the “back‑of‑the‑envelope” method begins to break down. Add a second location, introduce additional products, hire more employees, expand operations; complexity multiplies. Eventually the enterprise reaches a point where informal mental bookkeeping is no longer sufficient or safe.
At that stage, the business transitions from envelope‑based thinking to a basic accounting information system. Maybe the owner buys QuickBooks or a similar basic bookkeeping system. Soon after, they realize not everyone should see everything in QuickBooks, so they add a dashboard tool like BlueIQ to give managers the information they need but without exposing information the owner prefers to keep private.
The first challenge is to get the right information into QuickBooks. That itself has challenges, but I don't want to go down that rabbit hole.
Then comes the next challenge: getting the right information out of QuickBooks and into BlueIQ, in the right form, at the right time, for the right people. As the organization grows; say 25 employees, two locations, five product lines; this becomes increasingly difficult.
This is why accounting systems, management systems, dashboards, org charts, workflow diagrams, spreadsheets, word‑processed documents, emails, PDFs, and other artifacts exist. They are not created for the sake of only “documentation.” They exist to extend and strengthen the envelope-based approach; so owners, managers, and employees can operate the enterprise effectively.
These artifacts, that documentation, improves shared understanding across the members of the organization. As the enterprise evolves, its information systems must evolve with it, shifting from static documents to living representations of how the business actually works.
Enter artificial intelligence.
AI can help owners, executives, managers, and employees stay aligned and operate the enterprise more effectively and in many different ways. But for AI to perform to its full potential or even to simply succeed, the information originally created for human interpretation must also now be correctly interpretable by machines; and perhaps often by humans and machines working together to perform their work tasks.
This raises the bar. AI requires a coherent, consistent representation of the enterprise; and preferably one that both humans and machines can interpret in the same consistent way. Getting two humans to agree on the “truth” of the business is hard enough. Now machines must also understand that truth, and do so reliably.
A faithful, justifiable representation of enterprise reality in machine interpretable form becomes a strategic capability. It strengthens operations, management, quality, other insights, and overall mission performance.
The informational artifacts of an enterprise are not the enterprise itself. For example, a financial statement of an enterprise is not actually the enterprise itself; it is simply a representation of some specific aspects of an enterprise per some specific model.
Historically, those documents were documents meant for human interpretation; enhancing the "back-of-the-envelope". But now those documents must also support machine interpretation. But because how humans represent information in the form of a document can be quite arbitrary; machines have a hard time properly interpreting the information within those documents.
So how do we maximize the likelihood that AI succeeds; maximizing the potential of AI?
Two primary approaches tend to emerge:
- Teach AI to read and correctly interpret all existing human‑oriented documentation.
- Redesign how documentation is created; structuring it so machines can interpret it easily, and then projecting that same information into human‑friendly formats.
A third path blends both approaches, evolving toward a hybrid that works in practice.
Think of it this way. What might be better; (a) a business operating just as it has for the past 10 years or (b) a business taking advantage or artificial intelligence to create an advantage over competitors who are not using artificial intelligence?
The "back-of-the-envelope" approach works fine for some things. Not so much for others.
Additional Information:
Comments
Post a Comment