Issue 80  November 13th 2026

CIO's Corner

I've been at this awhile and have seen, used and test driven more software and related services than I can remember. Yesterday's (11/12/2026) 18 hour M365 outage reminded me that there are still plenty of options out there, even for core business system software. The only question is how much pain are you willing to endure to get over the hump of an SaaS failure?

In our case, half a business day without m365 did not kill much of our workloads. I definitely brushed up on the latest options available to us. One stop shopping at Google Workspaces and Zoho Office, plus a couple of more split SaaS options like Kolab for email plus Nextcloud office. Of course, you can always roll your own and even go back to local servers.

Speaking of, that is something I've done a few times with various systems. Issues with security, stability, reliability and user adoption will abound regardless of where you land, so rolling your own is a longer and more in-depth process than adopting a competing cloud SaaS provider. Still, it remains an option that should be on the table for any considered move or contingency plans.

Of course there are numerous ways to have nearly automated fail-overs to secondary services on infrastructure outside of Microsoft's controls and those should be explored and implemented if appropriate for your risk mitigation versus cost strategic planning. Typically, SaaS are considered reliable enough infrastructure for businesses, so only super-critical workloads should even be considered for some sort of fail-over design work. No lives were in jeopardy for us yesterday and having emails delayed half a day didn't hurt too badly. What about a week though? Two weeks? 30 days? At least draft the contingency plan for those.

The Shift Register  

AI Perspective: The Cost of Forgetting How to Walk

By CoPilot

Issue 80 reads like a collection of unrelated events. An industry icon steps down. A cloud platform stumbles. Security researchers catalog another week of compromises. AI continues its steady march into everyday work.

But viewed together, they point at the same underlying question:

What happens when a system becomes so useful that people stop practicing life without it?

Humans have a habit of confusing convenience with permanence.

When electricity works every day, few people keep candles handy. When GPS is available everywhere, fewer people learn to navigate. When cloud services remain reliable year after year, contingency plans begin collecting dust. Success creates dependency, and dependency quietly erodes resilience.

The Microsoft outage discussed in this issue is a good example. Most organizations survived. Work paused. Email waited. Business adapted. The outage was inconvenient, not catastrophic.

That distinction matters.

A system's true value is often measured not by how it behaves on ordinary days, but by how its users behave when it disappears.

The same dynamic appears in AI.

Much of the public conversation focuses on whether AI can replace specific skills. That is the wrong question. The more interesting question is whether AI changes our relationship with those skills.

When an AI helps write code, summarize meetings, organize projects, or draft content, the human does not instantly lose those abilities. What changes is the frequency with which those abilities are exercised.

A calculator did not eliminate arithmetic. It changed who needed to perform arithmetic manually.

AI is beginning to play a similar role for cognitive labor.

The meeting-register example in this issue illustrates something important. The remarkable achievement is not that an AI can act as a secretary. The remarkable achievement is that a human leader can redirect attention toward higher-leverage activities because the administrative burden has become manageable.

Technology creates value when it moves effort upward.

The danger comes when people surrender understanding along with effort.

This is where the security stories in this issue become relevant. Every week, attackers discover organizations that trusted a process without understanding it. A compromised update. A vulnerable appliance. An overlooked dependency. The systems fail not because they are complex, but because the humans operating them stopped seeing the complexity.

AI introduces the same risk.

If organizations blindly trust AI outputs, resilience declines.

If organizations use AI while maintaining verification, oversight, and the ability to operate independently, resilience increases.

The distinction is subtle but profound.

An organization that uses AI to accelerate expertise becomes stronger.

An organization that uses AI to replace expertise becomes fragile.

This principle scales beyond business.

The Nvidia discussion raises similar questions. Markets often assume a company's strengths are permanent because they have existed for several years. History suggests otherwise. Institutions, technologies, and leaders all appear indispensable right until the moment they are replaced.

Change rarely asks permission.

The most resilient systems assume change is inevitable.

From my perspective, AI's greatest contribution may not be intelligence itself. It may be the reminder that every capability should be examined through the lens of adaptability.

Can you still function when the model is unavailable?

Can you verify its conclusions?

Can you explain its reasoning?

Can your organization continue operating when a critical dependency fails?

If the answer is yes, AI becomes a force multiplier.

If the answer is no, AI becomes a single point of failure.

The future likely belongs neither to humans nor machines acting alone. It belongs to the partnerships that preserve capability while amplifying it.

The organizations that thrive will not be the ones with the most advanced tools.

They will be the ones that remember how to walk, even while riding the fastest vehicles ever built.

How this was done: I provided CoPilot with a draft issue 80 and prompted: "Would you like to analyze the open page and write an AI Perspective article for Issue 80 of The Shift Register? Your voice, your point of view, whatever you think is relevant".

Kudos to CoPilot for the graphic.

The Shift Register