Speed Is the Consolation Prize

What four hours of freed time inside a CPA firm reveals about where the next wave of advantage actually comes from.

I was recently talking with the CEO of a startup that is building AI agents for complex, process-heavy work. In their case, it’s tax preparation – one of the most time-bound, deadline-driven functions inside CPA firms. During peak season, their agents were freeing up, on average, roughly four hours per day per tax preparer. Not at the margins, but at the center of the work. And yet, the most interesting part of the conversation wasn’t the performance of the technology, but rather the response from the organizations adopting it.

They weren’t struggling to implement the agents. They were struggling to absorb the time.

In many cases, that newly available capacity didn’t translate into a rethinking of the work itself. Instead, it triggered a kind of organizational friction (yes, politics) because when time gets freed up at that scale, it begins to challenge deeply embedded assumptions about productivity, utilization, and value. In some firms, the instinct was to quietly refill the time with more of the same work. In others, it created internal tension because the system wasn’t designed to accommodate that level of slack. What surfaced wasn’t a technology gap, but a mindset and culture gap.

Contrast that with another firm I spoke with – similar agent deployed with similar time savings – where leadership refused to let the freed hours get reabsorbed into more returns. They made the trade explicit: every preparer owed a minimum of five hours a week to learning client advisory work, sitting in on CFO calls, and understanding what insights showed up in real conversations. Same efficiency gains from the AI implementation, but dramatically different approaches with their portfolio of time.

This is the broader shift that is just starting to come into focus. AI is not simply making work faster or more efficient – it is collapsing the time required to perform it. And that collapse is happening unevenly across functions, roles, and industries, which makes it harder to see as a single, coherent trend. But at the organizational level, the implication is consistent: the relationship between time and output is breaking down.

For decades, most operating models have been built on the assumption that more output requires more time applied to known processes. AI is disrupting that equation. When a meaningful portion of the day is no longer required for execution, the question is no longer how to optimize the work, but what the work should become. And this is where many organizations stall, because the default response is to treat freed-up time as excess capacity to be redeployed into the existing system. It feels rational, it preserves predictability, and it aligns with how performance has historically been measured. But it misses the larger opportunity.

The more useful way to think about this is through the lens of core and edge:

Core: the work that sustains the business as it exists today – repeatable, measurable, necessary.

Edge: the work that shapes what the business becomes next – exploratory, undefined, harder to quantify in the near term.

Historically, the core has consumed almost all the organizational oxygen, leaving the edge to a small subset operating on the periphery. What AI introduces is the chance to rebalance the equation – not by eliminating the core, but by collapsing the time it requires, and in the process opening up space to build for the future.

That space is where the real leverage sits, but it is not automatically captured. In fact, most organizations will default to reinvesting that time back into the core, driving incremental gains in efficiency or volume. There is nothing inherently wrong with that, but it is unlikely to create meaningful separation. The organizations that begin to differentiate will be the ones that deliberately redirect a portion of that freed time toward the edge – activities that expand capability, deepen customer understanding, rethink products/services, and build entirely new ways of creating value.

Of course, most organizations aren’t built for this. Performance systems, incentives, and management practices are all wired to optimize known processes. So when time is freed up, the system naturally pulls people back toward the core, because that is where success is defined and measured. This is why the friction shows up as “politics” or resistance.

The real gift of AI isn’t speed... it’s choice. Organizations are being given, perhaps for the first time at scale, the ability to decide what to do with time that was previously non-negotiable. That choice will shape not only how work gets done, but what kinds of capabilities are built and where differentiation emerges over time.

Most will respond by doing more work, quicker.

A smaller group will respond by doing different work altogether.

The gap between those two approaches is where the next wave of advantage will be created.

@Kacee