The promise of AI comes with a high degree of risk which requires an appetite

AI

AI is an emerging field that appears to be on everyone’s lips and minds at the moment.  A common challenge is that the majority of projects undertaken require a fair bit of investigation and trial and error to produce valuable results.

There are some AI tools you can use that can be put in place straight out of the box.  These are tools like image recognition, chatbots, etc.

When vanilla just won’t do

 

If your needs are a bit more specific you will most likely need something more bespoke.

This means that when creating an AI framework, the objective you are trying to achieve will need a specific outcome.  For example:

  • increase profit margin
  • reduce customer churn
  • improve on-sell and up-sell rates

 

Custom AI solutions take you into the wilderness

 

There’s nothing wrong with building a custom AI solution.  In fact, I’d encourage you to do so as I believe it will provide a far better outcome than a canned solution.  However, you must be aware of the risks that you may introduce when embarking on such a project.  Common risks are not too dissimilar to the usual IT development projects, but the causes are poles apart.

  • The project becomes far more costly than planned
  • It may take more time than you originally allowed, to achieve reliable outcomes 
  • You develop a solution that is far more onerous to maintain than budgeted for
  • You can’t get meaningful results in the first 2-3 versions
  • Your solution comes with an overhead that requires a higher degree of maintenance and management than budgeted for
  • All of the above

Worse still is you end up creating something that is currently under development by an incumbent.  The last thing you want is to spend a tonne of cash and months of time building something and Google or IBM release an off the shelf version the following month.

Can you compete with Google?

 

So, whilst AI has fantastic possibilities, there are unique risks that should be considered early on in any AI project.  If you charge full steam ahead into the project with your eyes and mind wide open, the outcome will be a game changer.

Share this post