Nothing I might write about generative AI could be particularly new. Or at least, there’s nothing obviously new and important in recent commentaries on the subject. Commentaries about the commentaries? That’s far more likely to be fruitful.
There is, for example, this popular cliché regarding AI adoption: that AI will automate mundane tasks, allowing humans to focus on more interesting work.
Maybe it will. But Dr. Yeahbut wonders whether this outcome will scale — whether, that is, the corporation’s key decision-makers will have any interest in paying for the kinds of more-interesting work AI will free up employees to perform.
So, let’s imagine AI rampages through a typical IT organization, freeing up employees to do something more interesting than what they’ve been up to. We’ll start with coding.
Once upon a time, programming — writing code — was the centerpiece of IT’s responsibilities. You might know CIOs who still look at the world this way, too.
But for oh so many reasons, your typical IT shop no longer expends much effort coding. Instead, IT buys when it can while building only when it has to. IT buys — well, licenses — when it can because licensing delivers pre-written, well-tested functionality. Even more important, it delivers pre-written integration.
From which we might deduce that, AI aside, the move away from custom development to licensed enterprise application suites should have already freed up coders to work on more interesting tasks. If that describes their current job descriptions, AI won’t have much impact. If it doesn’t, we can predict that freeing employees from their current responsibilities doesn’t end up with their working on more interesting stuff, regardless of the tasks they’re working on and what technology has supplanted what they used to work on.
Organizational realities
But never mind all that. Let’s agree, for the sake of argument, that AI might in fact free up some developers. Let’s further agree that some of them will have brilliant ideas and know how to write the code that can turn their brilliant ideas into brilliant realities.
(And let’s quietly ignore the question of what agreeing for the sake of argument even means.)
What happens next?
What happens next is that the Change Advisory Board has to certify that this new application has, in fact, been certified by Software Quality Assurance and can safely be deployed to Production. Which implies that SQA was willing to devote time and energy to testing it.
Nor are we done should the CAB give it the okay. Whoever in the enterprise represents the constituencies that would use the new software will have to give it their okay as well.
For example, the coder responsible for an interesting idea that could be deployed on the company’s mobile app will have to contend with Marketing, whose leaders might not be all that excited about an interesting idea that didn’t originate in the Marketing department.
Vanishingly few businesses are organized so that employees can ignore the organizational chart, their place on it, and the work assigned to them based on their place on it.
IT advancement is a team sport
Organizing like this might seem utopian when viewed from the position of an individual contributor. But, “more interesting” notwithstanding, for the most part important results come from collections of individuals who are organized into project teams and whose responsibilities are derived from the project plan, sprint plan, or some other vehicle for turning a big idea into assigned tasks that have scheduled start and end dates.
And then to assemble the individual results into team accomplishments, and from there to something that’s been implemented and put to successful use.
In case the point is still obscure, the more interesting work that’s being touted as our desirable future is unlikely to originate with the employees who have to do it.
No, from the individual contributors’ perspective the ideas they’ll be implementing will more likely be interesting to someone else.
The more likely result? If AI frees up employees to pursue even the most promising ideas, the most likely outcome will be for the organization to jettison both the hard-to-deploy ideas, and the employees who have them.
See also:
Read More from This Article: AI benefits don’t scale
Source: News