2026: The year of the fire hose

I recently came across a line from an article by Ann Handley about 2026 being the Chinese Year of the Fire Horse, but she mentioned that she’d initially misread it as the “Year of the Fire Hose”, which seems to accurately describe how this year is shaping up. It’s a phrase most people have used at some point when things start piling up, whether it’s projects, information, or requests, it feels like drinking from a fire hose.

AI makes it easier and faster to generate something that looks complete. Documents arrive more fully formed and emails carry more detail and structure. I use it regularly in my own work for things like summarizing meetings, organizing requirements, and analyzing data, and sometimes just thinking through how to approach something before it’s shared more broadly. It’s made parts of the process more efficient, but at the same time, it’s also made it easier to produce something quickly that still needs another pass of thinking.

More input, faster output

As a project manager, a large part of my day is spent reviewing inputs, pulling together context, synthesizing information, and refining it before it gets passed along to my production team or clients. As more people start using AI tools, things move much faster. There’s an expectation that work can be produced more quickly, that responses come sooner, and that everyone should keep that pace.

Because it’s easier to generate and move things forward, it becomes easier to move past the thinking that gives the work direction.

Communication is still the work

Communication can make or break a project. That’s where communication becomes the differentiator. Joe Wilson (The Esoteric Techie), a former manager of mine who shaped how I approach project management, recently wrote about different styles of communication. People process information differently, with some preferring concise direction and others needing more context or structure.

AI allows us to generate all those versions quickly. The challenge is choosing the right one, and resisting the urge to send all of them “just in case.” The type, format, and level of detail matters, and when those don’t align with the audience, it causes unnecessary friction and more work to process the information.

The real constraint: attention

The attention and focus challenge has been building for a while. Smartphones changed how often we check in and how quickly we respond. Messaging apps multiplied. At work, it’s rarely just one channel, with Slack, Teams, text messages, email, plus multiple project management tools layered in. We’ve been managing fragmented attention for years and AI now accelerates it.

At a certain point, the line between productivity and overload starts to blur, and how it impacts both us and others becomes harder to ignore.

Instead of just messages competing for our attention, it’s now fully formed documents, ideas, and outputs arriving faster than before. AI makes it easier to generate work, but it doesn’t create the necessary space to think through that work.

A familiar project management principle, Parkinson’s Law, states that work expands to fill the time available. What we’re seeing now is a related dynamic: as capability increases, the volume of what gets created expands with it. This compounds further as multi-agent AI systems come into play. There’s also a growing observation, highlighted in a LinkedIn post by Todd Anstis, that AI doesn’t just accelerate output, it expands the number of possibilities at any given moment. Ideas that once required teams or long timelines can now be explored quickly by a single person. That’s powerful, but it also means more directions to consider at once, more things in motion, and more work started in parallel. It becomes easier to initiate than to finish.

The cognitive cost of more

A recent article by the Harvard Business Review highlighted a trend of cognitive fatigue (or “brain fry”) associated with AI use. The article mentions that some organizations are beginning to measure productivity through AI usage and output from generated code to the number of systems being managed. As more multi-agent tools are introduced, the day-to-day work often involves moving between platforms with the expectation of greater efficiency. Yet it often results in more intense coordination and context-switching, which leads to mental fatigue and burnout. It’s no surprise that employee burnout is detrimental to business.

One of the more immediate effects is decision fatigue. Working intensively with AI introduces a constant stream of choices on what to prompt, what to refine, what to keep, and what to discard. Research shows that workers experiencing “AI brain fry” report significantly higher levels of decision fatigue. When that happens, the quality of decisions tends to decline. Even small drops in decision quality can have measurable business impact at scale.

The same cognitive strain shows up in the form of mistakes. Workers experiencing AI-related fatigue report making both minor and more significant errors more frequently. Small issues like formatting or code errors become more common, but so do mistakes with broader implications. When the volume of work increases and attention is spread thinner, it becomes easier for things to slip through.

Many of the people using AI most heavily are high performers. They’re often the ones exploring what’s possible and pushing work forward. At the same time, those same workers are more likely to report an intention to leave. The added cognitive load and pace of work can make the experience harder to sustain over time.

A more intentional approach

None of this suggests stepping away from AI. The benefits are clear, and it’s already part of how work gets done. What it does require is a more intentional approach to how we use it.

At an organizational level, that starts with recognizing that attention is a finite resource. Not every output needs the same level of depth, and faster production doesn’t always mean faster or better decisions. As AI becomes more embedded in workflows, there’s a growing need to build the skills around managing that output – knowing what to refine, what to move forward, and what to pause on. It also requires clearer expectations. If everything can be produced faster, it becomes even more important to define what “done” actually looks like, and to measure success based on impact rather than volume.

At the individual level, many of the same ideas show up in smaller ways. The principles are familiar: minimize context switching, reduce the number of channels you’re operating in, and be intentional about what you’re creating in the first place. Just because something can be generated quickly doesn’t mean it should be. More output doesn’t remove the need for clarity. It often increases it.

The first step is simply recognizing where friction can start to build. AI makes it easier to produce and share, and just “throw it over the fence”, so to speak.

That’s where a bit more ownership goes a long way.

Using AI thoughtfully means not just asking, “Can I generate this?” but also, “Is this ready for someone else to use?” It means taking the extra step to refine the output so it’s clear, relevant, and aligned to the person receiving it.

Ultimately, this comes down to how we work together.

If this really is the year of the fire hose, the goal isn’t to turn it off. It’s to get better at managing the flow, being intentional about what we take in, what we put out, and what we ask others to carry forward.