I have shunned new year predictions for the entire time I’ve been writing about technology and manufacturing. As has been said (attributed to physicist Niels Bohr and philosopher Yogi Berra), predictions are hard, especially about the future.
Usually these reflect either wishful thinking about the future or recognition of trends that may play out.
Marketing guru Seth Godin noted, “The Paradox of ‘On Trend’—By the time you get around to embracing the fashion of the moment, it’s almost certainly too late. The leading edge is defined by the fact that most of us aren’t on it.”
However, I have chosen to highlight two sets of predictions that came my way late last year. One is from Larry O’Connor, Founder and CEO, Other World Computing (OWC); the other from Michael Weller, Practice Leader for Manufacturing, Energy and Utilities at Verizon Business.
OWC makes high performance compute and networking equipment mostly targeted to markets not ours. The Verizon release was a bit surprising. I do think about networks. Usually not from the major carriers.
Their thoughts and my thoughts. Enjoy.
I looked at these OWC compute predictions mostly because I think that these have been happening. I don’t know to what degree adoption will occur, but it will be interesting.
2026 Prediction 1: On-Prem Comes Back, Not as a Rebellion Against Cloud, but as the Sensible Default for Performance, Cost, and Control.
“In 2026, more teams are going to rediscover the joy of having their data and workflows close to where the work actually happens. Not because the cloud is bad. The cloud is a great tool. It is just not the right answer for everything, especially when you are talking about performance, predictable costs, and keeping control of your own data.
It is easy to move a workflow up into the cloud, and then you wake up one day and realize you are paying for every little thing, and you are also at the mercy of a lot of services you cannot fix or influence. If your internet is flaky, or the provider has an outage, or you get hit with egress costs at the exact moment you need your data, that is not a strategy. That is a hope. In 2026, the smart shops will keep cloud as redundancy and reach, but they will rebuild the core on-prem so they can get their job done with less drama.”
I saw this play out last year by the company called 37 Signals. They looked at their monthly bills and decided their was a better way.
2026 Prediction 2: The Real Differentiator Will Be ‘Boring’ Infrastructure: High Performance Tech That Disappears into the Workflow.
“I think 2026 is the year more people stop buying ‘fancy numbers’ and start buying results. Everybody can show a chart. Everybody can promise the sky. But in the real world, what matters is whether the product is low overhead, dependable, and actually makes your day easier. The best compliment we can get is that someone forgets we are there, because they are too busy getting real work done.
More buyers are going to get tired of the enterprise pattern where you buy the thing, and then you learn you need ten other modules, another server, and a pile of add-ons to get what you thought you already purchased. That is not delight. That is aggravation. In 2026, the winners are going to be the companies that show up, evaluate the environment honestly, and deliver what the customer actually needs, with the least amount of fuss. Under promise. Over deliver. And make it work in the real workflow, not just in a lab.”
The real question—when will the big AI players realize this.
2026 Prediction 3: AI Becomes a Creative Partner – but the Creativity Remains Human.
“AI finally settles into its proper role for creatives, in 2026. It stops trying to be the artist and starts becoming the best assistant a cinematographer, editor, or photographer has ever had. The true creative spark still lives with the human, not in the machine. You can’t automate taste, timing, instinct, and storytelling. What AI can do is clear the runway so creators can spend more time making decisions that actually matter.
Those that rethink where AI lives in the workflow will be the teams that get this right. Instead of pushing raw footage and unreleased work into distant clouds, they will bring AI closer to the media and closer to the creator. When AI runs next to your storage, things happen at the speed of thought. You can test an idea, throw it away, try another, and never break your flow. That immediacy changes how people create. In 2026, the most successful creative teams will not be the ones chasing the biggest models. They will be the ones who build infrastructure that keeps humans in control, keeps their content private, and lets AI quietly do the heavy lifting in the background while the creativity stays exactly where it belongs.”
I think AI and AgenticAI have been way over-hyped. On the one hand, AI has been embedded in much technology we already use. Industrial companies are embedding Microsoft Copilot. We’ll see continued searching for ways to use it as a tool to help workers do a better job.
Where I pick at OWC’s comments a little, I think Michael Weller is a bit optimistic. Check his thoughts. As I just wrote above about AI, don’t buy the hype.
AI moves from paralysis to production: “Many manufacturers are worried about AI, and uncertain where to put compute power. Next year brings breakthrough deployments focused on computer vision for quality control and AI ‘shells’ that wrap legacy systems in protective security layers. This finally moving innovation ‘out of the drawer.'”
He should say that manufacturing executives are worried about AI. The hype has consumed so much media space, that they feel they must tell the board they’re working on it.
Factories will embrace visual technology in a larger way: “Manufacturing floors will become highly visual environments, driven by computer vision, digital twins, AR/VR headsets, and gamification. Humans learn visually far more effectively than through text, and visual content. From just-in-time training videos to 3D schematics, visual environments will transform worker engagement and productivity.”
I’m not with him here. I’ve tried out the headsets for 10 years. There are selected places (training?) where there is a bona fide use case. If Zuckerberg is dropping the metaverse from Meta, I’d take a hint.
Connected worker technologies will deliver on their promise: “After years of hype, 2026 is when connected worker technologies prove their value on the factory floor. Wireless-enabled tools – especially mobile equipment – finally deliver on promises of improved safety, real-time asset tracking, and operational flexibility. Manufacturers are realizing they can modernize without extensive infrastructure overhauls.”
This is the one I expected from Verizon. We’ve been through the 5G hype. Not sure what wireless he’s promoting, but actually workers have had mobile tools for a decade. And, I expect to see evolving applications.
Increased importance of wireless tech as a sustainability achievement: “Beyond operational benefits, eliminating copper cabling and reducing network power consumption represents measurable environmental progress they can actually quantify. The math is compelling: a single cellular antenna can displace 3-10 Wi-Fi access points, significantly reducing energy needs across large facilities.”
Cellular in place of WiFi? The jury is still out.
Click on the Follow button at the bottom of the page to subscribe to a weekly email update of posts. Click on the mail icon to subscribe to additional email thoughts.
