A generation entering the workforce under economic pressure is not waiting for organisations to modernise. It is building alternative operating models where AI agents replace structure, compress teams, and redefine how work is executed.
The assumption that work must be organised before it can be executed is beginning to fracture. For decades, organisations have treated structure as a prerequisite for productivity, layering management, process, and systems before anything meaningful could be delivered. What is emerging now, quietly but decisively, is a cohort that has dispensed with that sequence entirely. They are not reorganising work. They are bypassing it.
Jill Kenney, CEO of Sundae_Bar, describes a shift that is less about technology adoption and more about behavioural divergence. The company operates as an AI agent marketplace, connecting developers building task-specific agents with businesses looking to deploy them into real workflows. Positioned between rapid innovation on the supply side and growing demand from enterprises, it provides a view into how these systems are being adopted in practice. The narrative that AI is being introduced into workflows does not hold up when confronted with how younger operators are actually using it. There is no onboarding phase, no structured rollout, and critically, no waiting.
“They’re the generation that grew up with apps, with iPhones, coding in classrooms, learning how to build and play their own video games,” Kenney says. “So, where I see older generations nervous or hesitant about using AI, this generation is just more intuitive. It’s a new technology that they understand, and it’s easier for them to prompt and leverage.”
That intuition is translating into something more disruptive than early adoption. It is producing a mindset that aligns almost perfectly with how AI agents operate, modular, task-based, and indifferent to organisational boundaries. The result is not simply faster execution. It is a redefinition of what constitutes a team.
Kenney frames the contrast starkly. “Back in the day, if a company introduced something like Salesforce, you would have a team training on it for months, and you would still be figuring out how to integrate it,” she says. “Now, what I see is people going straight to the outcome. They will use whatever technology is in front of them to get there. They are not interested in the process for its own sake.”
This emphasis on outcome over process exposes a fault line that runs through most enterprise AI initiatives. The technology is often deployed without clarity on what it is meant to achieve, resulting in systems that exist but do not operate. In contrast, the emerging generation is working backwards from the result and assembling whatever combination of tools will deliver it.
“They understand how to prompt and use the tools,” Kenney continues. “But what’s happening already is that it’s becoming more intuitive. We’re moving towards a point where you won’t need to think about prompts or APIs. You’ll just say what you want, and the system will understand the outcome you’re trying to reach.”
Replacing the structure not optimising it
The most uncomfortable implication of this shift is not that AI agents enhance productivity. It is that they are beginning to displace roles that were previously considered structural necessities. The idea of hiring a team before building a business is being quietly dismantled.
Kenney offers a practical example in the form of a recent hire, Damien Player, a Gen Z operator who had already built a functioning business around AI agents before joining her organisation. His model was not built on developing proprietary technology, but on assembling and extending what already exists.
“He built a library of a thousand templates,” she explains. “People would use them, but they couldn’t get to the end. So they would hire him to finish the job. That’s a business built entirely around understanding how to deploy and complete workflows using AI.”
The significance lies in what is absent from that model. There is no traditional engineering team, no layered management, and no dependency on long development cycles. Instead, there is a capability to orchestrate tools, agents, and workflows in a way that delivers outcomes without the organisational overhead that would previously have been required.
This is where the idea of agents replacing co-founders, interns, or middle managers begins to move from speculation to operational reality. The replacement is not absolute, but it is targeted. It removes the need for roles that exist primarily to coordinate, transfer, or process information.
Kenney is explicit about where she sees the opportunity. “Why can’t we take off all that repetitive, computerised work that slows down creativity and independent thinking?” she says. “Why can’t we build a business where humans are in command, but the agents are doing the work that doesn’t need human intelligence?”
The consequence is not simply efficiency. It is a compression of organisational structure. When coordination can be automated, hierarchy becomes harder to justify.
The rise of plug and play execution
What distinguishes this moment from previous waves of automation is the accessibility of the tools involved. The current generation is not building systems from first principles. They are assembling them from components that already exist.
This is evidence of a shift away from building towards deploying. “We’re in the business of an AI agent marketplace,” Kenney says. “There are people building agents all over the place. The innovation is incredible. But for businesses, the challenge is finding and using them. What we are doing is creating a one-stop shop where developers can host and scale their agents, and businesses can search and deploy what they need.”
The implication is that capability is no longer constrained by internal resources. It is constrained by the ability to identify, connect, and orchestrate external components. This is a fundamentally different problem.
“From a user perspective, you don’t need to know how it was built,” Kenney explains. “You just need an agent that gets the job done. And increasingly, agents are starting to talk to other agents. If one can’t complete the task, it can find another that can. It’s like hiring a contractor who brings in other specialists to finish the job.”
This emerging agent-to-agent interaction, often referred to as A2A protocols, signals a move towards systems that are not centrally designed but dynamically assembled. It introduces a level of fluidity that traditional enterprise architectures struggle to accommodate. Yet the reality, for now, remains uneven. The promise of fully autonomous agents is not yet realised, and the current generation of tools still requires significant setup and oversight.
“People think they can just say, create me a campaign, post it, and report back,” Kenney says. “That’s not how it works today. You still need to connect systems, set things up, and guide it. But we’re seeing rapid progress. Agents that were solving 20 percent of a task are now solving 70 or 75 percent. That change is happening in months, not years.
The trajectory is clear even if the destination is not fully defined.”
A workforce that looks different from within
The deeper shift is not technological but organisational. If agents take on the work that once defined entry-level roles, the structure of work itself begins to change. Kenney reflects on her own early career to illustrate the point. “I started in communications, and for years I was just building lists, journalist names, emails, phone numbers,” she says. “What could I have been doing if that work was automated? I could have been understanding what those journalists cared about and building something meaningful for them.”
This reframing challenges the assumption that early career roles are inherently low value. It suggests that they have been constrained by the nature of the tasks assigned to them, rather than the capability of the individuals performing them. “What happens when you take that work off people’s plates?” Kenney asks. “Do they become more creative, more strategic? I think they do, but we don’t yet know what that fully looks like.”
The uncertainty is not a weakness. It is a recognition that the system is changing faster than the frameworks used to describe it.
The gap between intention and execution
For enterprises, the challenge is not a lack of interest but a gap between ambition and capability. Many organisations recognise the potential of AI agents, but struggle to operationalise them. “I talk to a lot of organisations that have the budget and want to implement AI,” Kenney says. “But they don’t necessarily want to be the ones driving it. They want someone else to set it up and show the efficiency.”
This creates a dependency on a new class of operators, individuals who understand not just the technology, but how to assemble it into functioning systems. At present, many of those individuals are emerging from the same generation that is driving the behavioural shift. “There will be a usage gap,” Kenney notes. “The people who are building and using these systems will not always be the same people who are asking for them.”
The gap is likely to persist until the tools become sufficiently intuitive to be used without specialised knowledge. Kenney suggests that this may happen sooner than expected. “When a 91-year-old can use it through voice and it just works, that’s when you know it has become mainstream,” she says.
A future without waiting
Looking ahead, Kenney resists the temptation to define a fixed endpoint. Instead, she focuses on the direction of travel, a movement away from constant interaction with systems and towards a model where work is executed in the background. “We started thinking about it like this,” she says. “Right now, you’re always looking down at your phone, working. What if your agents are doing the work while your head is up? You’re at a conference, you’re present, and the tasks are being handled in the background.”
The image is deceptively simple, but it captures the essence of the shift. Work becomes less visible, less manual, and less tied to specific moments of interaction. It becomes something that is initiated and then delegated, rather than continuously managed.
Whether this results in better outcomes, greater satisfaction, or new forms of dependency remains an open question. What is already clear is that a segment of the workforce is no longer waiting for those questions to be answered.
They are building systems that assume the answers will emerge through use.
And in doing so, they are redefining what it means to work before most organisations have decided how to respond.
