Re: Your “Home Depot” Approach to AI/ML

Original article was published by Brian O’Neill on Artificial Intelligence on Medium


Re: Your “Home Depot” Approach to AI/ML

Sorry, returns not accepted.

Home Depot.

If you’ve bought a house or property before, chances are, you have probably spent some time at a Home Depot, Lowes or whatever your local home-improvement store/chain looks like. I know I have. Aisles and aisles of tools, gadgets, widgets, and honestly — a million things that to are solution to problems I either don’t have or don’t know I have.

Meanwhile, back at home, a leaking faucet, a broken shelf, a sagging tomato plant, or some percussion setup hardware issue say “fix me, please!”

Known problems, for which Home Depot and my brain/hands probably have a solution.

Do I need a 3/4 in. Lead Free Bronze FIP Reduced Pressure Backflow Preventer?

No idea. (The “lead-free” part also doesn’t help me any more than knowing your solution is “cloud-based.”)

That’s a solution to somebody’s problem. Does my customer have that problem? IDK. Does yours?

Maybe — but you and I certainly won’t know if we’re not doing the diligence to talk to them, empathize, and find out.

It would be helpful to first understand what my customer’s needs are, instead of starting with this backflow preventer and looking for a place to insert it — not really being sure whether anyone will appreciate it or find it valuable.

Sure, take a walk through Home Depot and get inspired. “What could we do with THAT?”

But, don’t forget about those projects and real customer problems staring you in the face right now.

The “ML/AI” aisle in Home Depot might end up costing you an arm and a leg, and you can’t easily return it for a refund.

Want more articles like this on designing human-centered data products, ML/AI applications, and analytics products? Subscribe here.

About Brian

Brian T. O’Neill is a designer and advisor to software leaders on turning machine learning and analytics into indispensable data products. He is the founder and principal of Designing for Analytics and hosts the Experiencing Data podcast. 🎧 | Website | LinkedIn | Twitter