AI: The new digital Cookie-cutter

  I am a database and application developer.  At my age I suspect part of the reason why I was included among the estimated 40K employees that were terminated, in the year Larry Ellision decided to burn his house down for AI, is that the assumption would be that I might be resistant to AI, or not capable of learning new skills that would make me successful with AI.

  I’m not going lie and say that this is false; because I have a healthy skepticism of what impact AI can represent on the world today, some of which the world is now seeing and many of us who were replaced so that ai can be provided for over our past contributions.

My skepticism is as follows:

1.      The termination of ( a whole lot of) highly skilled workers in favor of what essentially becomes a modern cookie-cutter solution

2.      The AI cookie cutter is not necessarily a cost savings as

a.      Massive Data Centers consume physical Assets, Electricity on a “by powerplant scale”

b.     are being trained by lower wages, lower skilled workers, resulting in incomplete, insecure, and some flat-out wrong results

c.      In order to use the AI cookie cutter the “developer”, or more likely their organization, must also pay for the privilege.

d.     Debugging is often required, which requires some of those highly-skilled worker you just put back into the job market.

3.      The AI cookie cutter also stifles innovation. Sure you can tell it what you want but it is going to produce its “solution” based on a lower-skilled knowledge base and therefore you lose the true power of code innovation in the form of organization, performance optimization, security-minded constructs, and sheer confidence that what you produce will pass a regulatory review.

  That being said, I consider myself partially (time will tell how bad it can get) correct in these cases.  Now I also have a powerful need to make my mortgage payments and feed my family, and since I am not a billionaire that can buy and sell the media using the same dividends that could have bought a good AI foundation while still respecting his employees and the needs, I have begun updating my skillset to include AI concepts and specifics for projects both personal and production related.  My initial assessment is as follows…

You have to know what you want.

  In great detail.  I have been seeing all these training programs that tout a code assistant without needing to know code.  Please reference my above points.  This is the path to literally paying for production code that, if you are ignorant enough to publish without meaningful review, will create the largest amount of expensive technical debt we will ever see in the field.  If you as an AI user don’t understand the term technical debt, then you are the current and future problem in this space.  If you want to be a developer or an IT professional, you need to do the work and learn your trade.

You have to know how to describe what you want.

  Again, in great detail.  Of course, this also comes with the idea that you have to be knowledgeable in the end result, how it look, how it behaves, and especially how it fits the needs of the problem you are solving.  See my previous assertion that like the blacksmith apprentice of old,  learn your trade.

  I have had some interesting test interactions with AI, that I will bring out in a future post but the things I have leaned in them are “educational” it presents an opportunity of sorts that might just make the disillusioned, highly-skilled, and by necessity “petty” developer and IT professionals even more valuable in the destopian AI future that we find ourselves pursuing like Alice heading down the rabbit-hole.

EOL