For years, workplace technology has been sold on a promise that feels almost
impossible to argue with: remove friction from work and stress will disappear
with it.
The logic has repeated itself with every new wave of tools.
Email was supposed to eliminate paperwork and speed up decision-making.
Laptops promised freedom from the office. Collaboration platforms arrived to
streamline communication, while the cloud was positioned as the cure for the
operational bottlenecks that slow teams down.
Each shift made perfect sense at the time, and yet the lived reality of work
has become more complicated rather than less.
AI is now being positioned as the most complete expression of that idea. It
is being introduced not simply as another productivity tool, but as a
structural correction to how knowledge work is organised.
The expectation is that AI will absorb repetitive tasks, reduce cognitive
load, and return time to employees so they can focus on more meaningful work.
But as organisations begin deploying these systems at scale, the reality that
is emerging is more complicated than the promise suggests.
Work is becoming faster and more automated, but not necessarily lighter, and
in some cases it is becoming more intense in ways that are difficult to
capture through traditional productivity metrics.
The Architecture of Frictionless Work
Inside enterprise systems, AI is often described as a way of removing
friction that has accumulated over years of layered tools, processes, and
organisational complexity.
IT tickets, manual coordination, repetitive workflows, and administrative
overhead are increasingly treated as inefficiencies that should be eliminated
entirely.
Phil Lenton, Head of Product Management for SaaS Services and AIOps at
Riverbed Technology
,
describes this shift as a move away from reactive systems and towards
environments that resolve problems before users are even aware of them.
“IT tickets still dominate how workplace issues are resolved. Employees lose
time, context and focus, while IT teams are flooded with repetitive tasks
that pull them away from higher-value work.”
“A better employee experience starts with removing that friction altogether.
This means moving toward a zero-disruption model, where agentic AI can detect
and resolve issues before users even notice them.”
The intent is clear – if disruption disappears, attention should improve, and
if attention improves, burnout should decrease.
But that assumption depends on whether organisations use recovered capacity
to reduce pressure or simply to increase output.
Collaboration as a Mixed Human–Machine System
Work is also being reshaped at a structural level.
It is no longer defined purely by human interaction, but increasingly by
systems that actively participate in how work gets done.
Cisco’s Head of Collaboration, Snorre Kjesbu,
describes this as a fundamental shift in the fabric of work itself.
“Collaboration is not only people to people. It’s also people to AI and AI to
AI.”
He expands this idea further, describing how AI is no longer peripheral to
meetings or workflows, but embedded inside them.
“In all meetings now we actually have an AI agent or two or three or four or
five that can come in and help and assist.”
This changes the nature of collaboration from something episodic and
human-led into something continuous and distributed across both people and
systems. Meetings, decisions, and workflows no longer sit neatly within
human boundaries. They are increasingly co-produced in real time.
The benefit is obvious in terms of coordination speed.
The implication is less obvious but more important: when coordination becomes
easier, organisations rarely choose to do less of it.
The Cognitive Load Case for AI
One of the strongest arguments being made for AI adoption is not simply
productivity, but cognitive relief. In knowledge work, burnout is often
associated with constant context switching, fragmented attention, and the
accumulation of low-value tasks that interrupt deeper thinking.
Russell Tilsed, VP at RingCentral,
frames this as a structural issue in how modern work consumes attention.
“Burnout is frequently driven by low-value, repetitive tasks, tools
switching, and the cognitive load of constant communication.”
“AI can remove much of that friction. When AI handles meeting summaries,
drafts responses or surfaces relevant information at the right time, it
preserves mental energy and prevents the slow drain that compounds over
time.”
He also links this directly to business outcomes rather than just employee
wellbeing.
“When people feel their time is respected and their work is meaningful,
engagement rises, and with it, retention. Given the cost of replacing skilled
knowledge workers, even modest improvements here translate into significant
financial returns.”
The logic is compelling, but it assumes that time saved through automation is
returned to employees in the form of reduced load, rather than absorbed into
new expectations.
Meetings, Attention, and the Compression of Work
Meetings remain one of the most persistent sources of cognitive fatigue in
modern organisations, particularly in hybrid environments where coordination
costs have increased rather than decreased.
Daniel Johansson, Regional President for North at Jabra,
highlights both the scale and the opportunity embedded in this shift.
“Over a third of working weeks are filled up by meetings. And with around 80%
of meetings now taking place in virtual or hybrid formats, there is
opportunity to embed AI directly into collaborative workflows and improve
employee experience.”
AI is increasingly used to transcribe conversations, generate summaries, and
extract actionable insights so that participants do not need to process
everything manually in real time.
At the same time, Johansson points to a broader shift in how interaction
itself may evolve.
“Voice AI takes this a step further. In tasks like brainstorming or outlining
ideas, speaking to an AI assistant or LLM adds a more flexible way of getting
things done.”
The direction of travel is toward less friction in communication and faster
movement from idea to output. But again, the system-level effect is not
necessarily reduced workload. It is often increased velocity.
When Efficiency Becomes the New Baseline
Across operational environments, a consistent pattern emerges. When systems
become more efficient, organisations rarely use that efficiency to reduce
workload. Instead, they use it to increase throughput.
Lenton captures this dynamic in simple operational terms.
“By offloading routine, draining tasks and reducing context switching, AI
allows employees to stay focused and productive.”
Over time, however, what was once “extra capacity” becomes redefined as
normal capacity. The baseline shifts upward, often without any explicit
decision being made.
The Expectation Effect and Burnout Dynamics
For people working in organisational psychology and workforce wellbeing, this
pattern is already well understood. Each productivity wave expands what
organisations expect from individuals.
Stephanie Lemek, Founder & CEO of HR consultancy group The Wounded
Workforce
describes this shift as an almost automatic recalibration of expectation.
“When a task that once took hours now takes twenty minutes, the implicit
message is: you now have time for more. More projects. More output.”
“The capacity AI creates does not become rest or recovery. It becomes
expectation.”
This is where burnout begins to form structurally. Not only through workload,
but through the speed at which systems absorb efficiency gains into new
demands.
The result is a moving target for what “enough work” actually means.
The Human Cost of Acceleration
Alongside organisational dynamics, there is a more personal layer to this
shift that is increasingly visible among knowledge workers.
Candice Thompson, a Silicon Valley-based therapist working with technology
professionals, describes a growing emotional and psychological strain in her
clients.
“Tech workers are being asked to produce more, while layoffs and
organisational instability are increasing – this is leading to overworked,
more stressed employees who are questioning the longevity of their careers.”
Even when tools reduce effort at the task level, the wider environment can
still feel unstable and accelerating. The experience of work is shaped as
much by security, expectation, and cultural pressure as it is by efficiency.
The Unresolved ROI Question
Taken together, these perspectives reveal a structural tension in how AI is
reshaping work.
On one side is a compelling business case built on efficiency, reduced
friction, and improved utilisation of attention. On the other is a lived
experience shaped by rising expectations, compressed recovery time, and
increasing uncertainty about what sustainable performance actually looks
like.
Both realities can exist at the same time. AI can reduce effort while
increasing intensity. It can improve productivity while also increasing the
pace at which that productivity is demanded.
The unresolved question is not whether AI works. In many contexts, it clearly
does. The question is whether organisations are structurally capable of
converting that efficiency into reduced pressure rather than increased
output demand.
Because if productivity gains are consistently reinvested into higher
expectations rather than human recovery, then the return on investment will
not be measured in time saved, but in strain redistributed across the
workforce.
And that is where the paradox now sits – not in what AI enables, but in what
organisations decide to do with what it unlocks.
