Today’s [AI] mistakes frame AI’s future

AI makes mistakes. I’ve seen it. You probably have as well.

But it’s doing exactly what it should at this stage. As early AI adopters, we share the pain of the initial owners of the first Model T –

  • We are the ones with sore arms from endlessly hand-cranking our engine since there was no electric starter yet
  • We are the ones complaining about the bumpy ride since suspensions weren’t what they are today
  • We are the ones fed up with driving in the rain, and eating bugs as we drove because there was no windshield yet

How did all the creature comforts and enhancements come about to give us the automobiles of today that we know and love?

Feedback.

Users reported their experience and updates followed.

And while it took years for some of these automotive enhancements to come to fruition (e.g., electric starter), technological updates of today happen in weeks or months.

And while a select, elite few had the privilege of “beta testing” the Model T, we all get a shot at molding this new technology we call AI.

We all have the opportunity to leave our unique imprint on the future of AI – so when any platform makes a mistake, don’t complain.

Be patient.

Teach it something.

It’s learning.

How many days to replace a light bulb?

It took nearly a year and a half for a paper towel dispenser to be replaced in my office.

Because it was too expensive? Not likely – I’d guess under $100.

Because it was too complex and needed that amount of time? Nope – that type of task should conservatively take about 4 hours of actual work to complete.

Complex tasks need that amount of time to be completed.

Simple tasks – like this dispenser – are sidelined to focus on the complex, “more meaningful” tasks. Simple tasks could (and should) just be handled when they are identified, but they are delayed by complex project management processes.

Simple tasks are bogged down by process.

But without process, complex tasks inevitably fail.

What’s the solution?