April 13, 2026
By Bob O'Donnell
If you’ve been following tech industry news at all lately, you can’t help but have noticed that everything is turning “agentic”. All the major AI platforms, the big operating systems, the browsers, and seemingly every software category now appears to be getting infused with AI-powered agents.
In truth, things like Claude Cowork/Claude Code, OpenClaw, NemoClaw, and other emerging agentic platforms are already demonstrating capabilities that would have sounded far-fetched even a year ago. They can automate tedious processes, generate high-quality content, and even let non-programmers build useful applications. For those with even longer perspectives, the memories of the original digital assistants and what they were supposed to offer now seem like simple childhood daydreams. We are truly at the dawn of a new era of computing power that is, quite frankly, a bit scary to behold.
And yet, we’re also still clearly in the early days of these developments. In part, much of this is due to the speed at which these advances are happening. As many others have pointed out, the pace of innovation happening in the world of agentic AI is astounding and certainly well beyond anything I’ve experienced in the many decades I’ve been part of the tech business. What’s fascinating is it’s not just the speed of technical advances that’s breathtaking to behold, but also the seeming disregard for any concerns about the potential impact that these advances might have. Move fast and break things might be a perfectly acceptable approach when it comes to a single product or service, but when the technologies involved have the potential for global disruption, a more mature, thoughtful approach is not just smart but essential. Yes, it’s cliched, but it’s still true: with great power comes great responsibility.
But regardless of your philosophical take on some of these advancements, the reality is they’re happening and will continue to do so for many years to come. What matters now is making sense of where this agentic wave will have the most immediate impact. In the near term, I see three areas that deserve the closest attention: how we interact with computing devices, what infrastructure is needed to support these workloads, and how organizations adapt to the resulting workforce and security challenges.
First, on a practical level, the way we interact with our devices and, indeed, the types of devices we need are starting to evolve. One of the key changes I’ve perceived in the growth of agentic tools is the more direct types of interactions we’re having with our technology. We’re very quickly moving away from thinking about how to perform a task or achieve something and simply requesting an outcome and expecting it to be done without even worrying much about what tool will do the work. The implications of this on the traditional model of opening an application, navigating an operating system, and hunting for a specific feature are profound. They are also likely to shape how all of these tools progress (and which ones even survive) for many years to come.
At the same time, the types of devices people are using to run agentic applications and workloads is evolving quickly. The idea of leveraging small desktop computers like Nvidia’s Spark and its equivalents or Mac Minis as a second personal computer to run tools like OpenClaw and local versions of open source LLMs was unheard of even a few months ago, yet it’s now becoming a notable trend—at least for those who can afford to purchase machines with the large amount of RAM necessary to run these workloads. As these tools become more dependent on cloud services, continuous connectivity becomes more valuable, which could strengthen the case for 5G cellular-connected PCs.
Similarly, on the data center side, we’ve seen a surprisingly large and rapid shift towards providing new types of silicon and infrastructure capable of running these agentic workloads. From Arm’s recent AGI CPU launch to Nvidia’s new CPU-only and Grok technology-powered LPU infrastructure, there’s been an impressive degree of diversification in data center silicon recently. The bigger point is that agentic AI is widening the range of viable compute architectures. GPUs remain central, but the market is clearly opening up to CPUs, LPUs, and other specialized designs optimized for different parts of the AI workflow.
An important corollary to this, however, is that much of this currently applies to a relatively small—though very vocal—group of people. One societal aspect of the agentic onslaught is that it’s creating a significantly larger gap between people who have varying levels of experience and understanding of the latest AI developments. Think of it as the digital divide on steroids (or should I say peptides?!). People who are closely monitoring and trying out all the latest developments are waxing poetic about how much more they can achieve and even how much more they’re being inspired to work because of what they can do with them. Most workers and consumers, on the other hand, have little to no idea what’s even possible with these tools, let alone how to use them. This is partly due to the still limited training that most organizations provide on AI tools. However, it’s also because integrating agentic-powered processes into someone’s job requires radical rethinking about how people work, redefining of what’s expected of them, and other bigger picture questions that can’t (and won’t) be solved quickly.
In enterprise computing environments, concerns about agentic “bloat,” rogue agents and security are quickly rising to the fore. Not surprisingly, traditional enterprise infrastructure and security companies such as Microsoft, Cisco, Palo Alto Networks, IBM and many others are quickly jumping onto what they see as both a huge threat and huge opportunity for them to help businesses. Many of these new agentic security focused products look to be strong entrants, but figuring out which pieces an organization actually needs, and how those pieces fit together, is likely to be a major challenge for enterprise IT teams. Also, the “agentic AI divide” is becoming a serious concern within organizations as many companies are having a hard time trying to accommodate both the super eager early agentic AI adopters and the agentic AI laggards. Toss in the fact that a surprising number of companies are blaming extended AI use for recent layoffs and the environment for agentic AI adoption for many regular workers is starting to get murky at best.
As history teaches us, all major technology transitions bring with them uncertainty and even unrest and so it is with the agentic AI paradigm shift. We’re unquestionably starting to see some truly impressive, productivity enhancing capabilities that are finally fulfilling the dream of what “intelligent” computing devices and services could be. At the same time, a number of potential negative impacts are also starting to become clear. Ultimately, there isn’t necessarily an easy right answer, but there’s no doubt that more thought needs to be given to the implications of agentic AI developments by businesses, workers, developers, and societal leaders as they’re being built and released. The opportunities created by agentic AI are real, but so are the disruptions. The organizations that benefit most won’t be the ones that move fastest blindly—they’ll be the ones that pair experimentation with discipline. And finally, the question remains: Are our organizational cultures ready to stop managing tasks and start managing outcomes?
Here’s a link to the original column: https://www.linkedin.com/pulse/outcome-economy-surviving-agentic-blitz-bob-o-donnell-o748c
Bob O’Donnell is the president and chief analyst of TECHnalysis Research, LLC a market research firm that provides strategic consulting and market research services to the technology industry and professional financial community. You can follow him on LinkedIn at Bob O’Donnell or on Twitter @bobodtech.
| 
