Magic vs. Magical

People talk about AI like it’s magic. It’s not magic; but sometimes it seems magical. However, there is real work that goes into using AI to create value. The code may seem to come out of thin air, but the work that goes into getting to that code is real.

For example, I’ve been using AI to build the tools that bring the recommendations and outputs of my client work to life. Here are a few things I’ve done with AI recently:

• Before: Write a strategic plan in PowerPoint. Create an Excel workbook that calculates the metrics for the client to measure success over time.

• Now: Write the strategic plan. Build an interactive KPI dashboard that ingests the metrics from a central source and presents the data in an interactive dashboard.
  

• Before: Analyze the cost and time savings of better compliance by reducing manual error and increasing throughput of reviewing invoices against bid sheets.

• Now: Draft the requirements and develop an automated compliance tool.

The possibilities are so fundamentally game-changing, and yet, the game hasn’t completely changed. 

Don’t get me wrong; a lot has changed. A LOT. Like me, someone with no developer experience, now building functional tools for clients.

But, a lot has stayed the same, which is why many organizations still need support:

• Knowing how to put together a comprehensive set of requirements

• Planning for the downstream impacts of automation and adjusting workflows

• Mitigating risks (security, governance, operational) from the outset

• Understanding data quality and what good metrics are

• Having a neutral, independent view of best practice, and what good looks like

• Knowing that just because you can build it, doesn’t mean you should build it

AI isn’t doing my job for me. I’m not offloading any cognitive labor onto these large language models. But, it is enabling a whole new set of capabilities for me that previously would have required me to find a new vendor (or to get a new degree).

And that, my friends, is pretty magical.