Real Time View

For anyone who has become a seasoned IT professional, being part of the industry for several decades, the thought of change isn’t anything new. Big changes? Small change – and I don’t mean the few coins we might have jingling in our pockets. But change that comes in waves that are ever-present in a sea that itself is always on the move. From silicon to databases to the internet and browsers (the dot.com era) and upwards, into the clouds. The appearance of these waves has been both regular as they have been disruptive. However, what lies ahead is perhaps the most challenging wave any of us has ever faced – AI.

Is this wave for us or against us? As we take our first tentative steps into 2026, is the appearance of AI, a transformational technology, just another instance of change or is it an entry point into an entirely new sea? For the Nonstop Compute community, what makes this change concerning is that the fresh, clean, data created on Nonstop will feed mechanisms outside of our control indeed, frequently outside any guardrails we may have established and have been following for many years. It asks the bigger question – what problems exactly are we addressing? The strength of Nonstop has always been in its support of the Customer Experience (CX). Will we see that changing and who exactly will be our future “customers?”

In the article The Stagnant Order, as appearing in the November – December, 2025 issue of the magazine, Foreign Affairs, author and academic Michael Beckley, writes:

“Some forecasts claim that artificial intelligence will turbocharge global output
by 30% per year, but most economists expect it to add only about
one percentage point to annual growth. AI excels at digital tasks, yet the
toughest labor bottlenecks are in physical and social realms.

Reflecting these limits, roughly 80% of forms using generative AI
reported that it had no material effect on their profits, in a
McKinsey Global Survey of AI.”

When we consider the impact of Generative AI, it comes as no surprise to the Nonstop community that while it proved entertaining in the near term, when it came to adding value in the real world of transaction processing, reading about it, generative AI did not provide any effective value to transaction processing. Entertaining? I experienced this first-hand when Margo’s daughter Anna, fed a recent blog into AI and produced an “interview” between two AI-generated hosts, where I was truly surprised by the resultant presentation.

However, viewed as amusing if not a source of entertainment is one thing and providing data created on Nonstop to adjacent servers for inclusion in model development is another thing but then what to make of the massive build-out of data centers in support of AI and its models? For our risk-averse Nonstop systems’ managers, is it all becoming too much? Are we chasing a bridge too far? Are we oblivious of the dangers involved, blinded by the hype? Not forgetting, too, that Quantum Computing looks to be just around the corner. After all, as any sailor can attest to, waves do come crashing down. The short-term benefits from the ride may be obliterated by a sustained period of retraction as the sea we entered proves shallow.

In her article, Dust to data centers: The year AI tech giants, and billions in debt, began remaking the American landscape published in CNBC, December 31, 2025, journalist MacKenzie Sigalos ponders how circular the pursuit of supporting AI infrastructure has become:

“Nvidia is effectively financing demand for its own chips, Oracle is building
the sites, AMD and Broadcom are positioning as alternative suppliers,
and OpenAI is anchoring the demand.

Critics call it a circular economy: capital, capacity, and revenue all recycling
through the same small set of players. It works as long as growth holds —
but if demand slips or funding tightens, the stress can propagate fast
through a web of shared exposures.”

Ouch! Then again, is it Generative AI or Agentic AI that is driving this amazing, indeed sudden upward ramping of investments in data center sites? It may not matter but what should be a concern for the Nonstop community is Agentic AI. Whereas Generative AI is all about “Content creation (text, images, code),” Agentic AI is “Autonomous action, goal execution, decision-making.” Think of the former as being somewhat passive whereas the latter is far more active.

According to a March 2025 report by researcher Forrester, Agentic AI Is The Next Competitive Frontier, the differences between Generative AI and Agentic AI couldn’t be clearer and not just in a passive versus active way (both using AI agents):

“Agentic AI isn’t just smarter tech — it’s a smarter approach to transformation.
True enterprise autonomy requires orchestration, adaptability,
and leadership alignment across the organization.

Agentic AI (plans) acts, adapts, and scales—
far beyond the limits of traditional AI agents …
Static automation breaks under change—
agentic systems evolve with real-time context.”

For the sake of clarity, the changes that should be bothering all members of the Nonstop community is the potential business imperative to allow smart AI agents to be present on Nonstop systems as part of Nonstop applications. The very thought of “foreign code” of this nature having a presence on Nonstop, capable of bringing down a Nonstop system, is not only a bridge too far but rather, a bridge to oblivion. Operationally and Security-wise, alarm bells should already be sounding. This is not about taking what we have recorded but rather, having the potential to alter what is being recorded (as a transaction) to better reflect a desired outcome.

From a management perspective, how engaged will your Nonstop team be with those driving AI? Who on your team is working with Oracle, AWS, Google, Meta, OpenAI, etc.? “Even if AI keeps advancing, major productivity gains may take decades because economies must reorganize around new tools,” wrote Foreign Affairs contributor, Beckley. Let’s not forget too, where will all the power infrastructure come from and at what cost? Worse still, what (potential) loss of sovereignty?

For some time, I have expressed my own sentiment on the topic of AI as being nothing more than Arrogance working with Ignorance. Too much? Perhaps, it is with this in mind, that I warmed to Sydney Morning Herald journalist, Richard Glover, who in his January 2, 2026 opinion column, These are my 20 definite, undeniable guaranteed predictions for 2026, included the observation:

“Mistakes made by ChatGPT, and other AI helpmates, will become
so ubiquitous and commercially disastrous that risk-averse companies
will ban staff from using them, except for the simplest of applications.
The phrase ‘it’s neither artificial nor intelligent’ will become commonplace,
as if we were all on to this from the start.”

I have been on this IT sea for many decades and I have born witness to the many changes already referenced. The sea is now a familiar place for IT veterans like myself. However, we are watching a transition to a new, unfamiliar sea where the waves that come at us are unpredictable. Is AI, whether Generative or Agentic simply more IT hype or is it a blueprint for the future of IT? It’s early days and as uncertain as I am about promoting this wave to my clients, it is safe to say, this ship we call Nonstop better come equipped with lifeboats!