The Robots (AI) Are Already in Control (Part One)

The robots are already in control. Having witnessed several systems migrations, I’ve been saying this for at least a decade.

Robots Gone Mad

Remember how the Phoenix pay system wreaked havoc on public servant pay back in 2016? Did you know that system has cost more than $2.4 billion? The problems persisted well into 2022. A full timeline for implementation was nicely set out by the Ottawa Citizen.

Now think about Ontario’s Social Assistance Management system. It cost over $294 million to build and fix after being implemented in 2014. Major issues were experienced as a result. I knew someone who worked there before and after the migration. After the migration, their authority in the system was removed. So even if they saw something awry with a potential payment to a client, they had no ability to stop or adjust it. The human was removed from the loop. Whether this occurred intentionally, or a from bungled configuration – the impact and outcome were the same.

Everyone reading this post has customer service war stories to share. Think about how long you’ve sat on phone queues, waiting to discuss an airline booking / cancellation or credit, incorrect cellphone charges, credit card fraud, etc. Post a comment with your customer-service related horror story!

System Migrations Gone Awry

As a mid-level Department Head at a university, I’ve been intimately involved in system migrations. I’ve been involved to some degree in procuring / implementing / configuring and/or learning /training others on two library systems, one new finance system, one new academic CRM, migrating from Microsoft 365 from Lotus Notes and a migration to a new learning management system. I’m sure many of us have been there.

And having been there, we all know the routine: the organization decides it needs a new system. A procurement team is pulled together, sometimes consisting of people who know what is actually involved in migration (but many of whom don’t). A few big players in the marketplace put in their bids. They arrive to the office to give their dog and pony show – of course demonstrating all of the great shiny things that the system can do.

The organization puts up more cash than they intended to buy the thing, but quickly, the budget rears its ugly head. The deciders are left with the decision: proceed? and if so, where can we cut costs? Of course, the costs are cut in three key places: data migration, system configuration, and/or user (and IT team) training.

However, the promised Cadillac, will be a lemon if the organization does not put up the cash to pay for proper migration, configuration and user training.

Customers Suffer

As the staff users, IT personnel (and customers!) well-know, major headaches abound for months or years afterwards.

When the organization’s migration service contract runs out, the staff are left holding the bag. Processes haven’t been set up in the system. Configuration is a mess. The data needs way more clean-up than anticipated. Half-trained IT personnel with dubious system permissions are wearily left without all the tools in the toolbox. They are now trying to prop up the Cadillac just to get the basics done.

Organization that don’t have the resources to properly implement in the first place, can’t later intelligently assess the migration impacts on staff, productivity and customer service.

I’d propose a great research study to find out how many customer-service related complaints arise from problems with the org’s IT infrastructure.

How many times have you been told by a service agent that they can’t help you because the system won’t allow it?

The Robots are already in control.

Now, take this very real lived experience that I just laid out. Predict the future by adding AI on top of it…

Queue the AI-generated doom tunes…

Stay tuned for my next post “The Robots (AI) are Already in Control (Part Two)!


James Bagnall, “A timeline of the Phoenix pay debacle: 29 years and counting” (Ottawa Citizen, February 23, 2018), online:

Ashley Burke, “Social assistance software a headache for city staff, clients” (CBC News, March 19, 2015), online:

Priscilla Ki Sun Hwang, “Phoenix ‘nightmare’ still haunting public servants, more than 6 years on” (CBC News, May 24, 2022), online:

Keith Leslie, “Cost of problem-plagued computer that administers welfare soars to $294 million” (The Canadian Press, October 21, 2015), online:

The post The Robots (AI) Are Already in Control (Part One) appeared first on Slaw.

Related Posts