Gavin Fudge

27/07/2026

4 min read

When the machine does everything, what is human work for?

AI may automate skill. But meaning, judgement, and direction remain profoundly human.

The strange future of work

For most of modern history, work has meant one thing.

Doing things.

Producing things; calculating things; building things; analysing things.

Human effort has always been tied to execution. Even knowledge work - writing reports, designing systems, analysing markets - ultimately involved applying skill to produce an output. Artificial intelligence disrupts that assumption.

As AI systems become increasingly capable, they begin to perform many of those tasks themselves. Writing, analysing, modelling, coding, summarising, designing. The list grows almost weekly.

Which creates a strange situation.

For the first time in history, humans are building machines that can perform large parts of knowledge work faster, cheaper, and often more reliably than we can.

This raises a question that goes beyond productivity: If machines increasingly perform the tasks that define work, what is work for?

Work was never only about output

It is tempting to think of work purely as an economic mechanism. Organisations need outputs, people provide labour, and markets coordinate the exchange. But historically work has always carried deeper meaning.

Work is how people exercise agency in the world. It is how we test ideas, develop judgement, build communities, and shape the environment around us. Work has always been one of the primary ways humans interact with reality.

The philosopher Hannah Arendt famously distinguished between labour, work, and action. Labour sustains life. Work creates durable things. Action shapes the shared world we live in.

For most modern organisations these distinctions blurred together. The industrial economy valued efficiency, repeatability, and scale. Skills were rewarded because they increased output.

But AI introduces a new possibility.

If machines increasingly perform the execution layer of work, humans may shift toward something closer to Arendt’s idea of action: shaping direction, meaning, and purpose. In other words, the human role becomes less about doing the work and more about deciding what work should exist in the first place.

AI systems learn from the past. Events create the future.

‍

The problem with predictive systems

This shift becomes clearer when we examine how AI systems function.

Modern AI systems, including large language models, operate primarily through pattern recognition. They learn statistical relationships within enormous datasets and use those patterns to generate plausible outputs. This makes them extraordinarily powerful within environments that resemble the past. But the world does not always behave that way.

In The Black Swan, Nassim Nicholas Taleb argues that many of the most consequential events in history were inherently unpredictable because they fell outside the patterns visible in prior data.

Financial crises
Technological breakthroughs.
Scientific discoveries.

These events reshape systems precisely because they were not expected.

Taleb’s argument highlights a fundamental limitation of statistical systems: they struggle with novelty that has no precedent.

Alain Badiou and the Event

A similar idea appears in the work of Alain Badiou.

Badiou describes what he calls the Event - a moment when something genuinely new enters the world, something that cannot be explained by the existing order of knowledge.

An event is not simply a change - it is a rupture.

Something appears that the existing system could not predict or categorise. And from that moment onward the world reorganises around a new possibility. In Badiou’s philosophy, truth emerges when people recognise an event and remain faithful to it - when they pursue the implications of something new even before the rest of society understands its significance.

Technological revolutions often look like this in hindsight. The internet; the smartphone; artificial intelligence itself - each began as something strange, incomplete, and poorly understood. Only later does society reorganise around their consequences.

AI cannot recognise its own event

This is where the philosophical thread intersects with the technical one.

AI systems learn from the past. Events create the future.

An AI model trained on historical data cannot fully recognise the significance of something genuinely new. It can describe possibilities, simulate outcomes, and extrapolate trends. But recognising that an event changes the structure of reality requires a different kind of thinking.

It requires judgement.

● The judgement to recognise when the old model no longer fits.

● The judgement to explore a possibility that appears uncertain.

● The judgement to pursue a new direction before it becomes obvious.

In other words, the very moments that shape the future are precisely the moments where human cognition remains essential.

Organisations will need to change too

This shift has implications not only for individuals but for organisations.

Many organisations today are structured around the efficient execution of known processes. Workflows, reporting structures, and incentives often assume that value comes from consistent output. But if AI increasingly performs the execution layer of work, organisations may need to adapt in several ways.

First, they may need to place greater emphasis on judgement and system design rather than task execution.

Second, they may need to create environments where experimentation and exploration are possible. If genuine innovation emerges through unexpected events, organisations that are too rigid may struggle to recognise those opportunities.

Finally, organisations may need to reconsider how people contribute value. The most valuable individuals may not be those who produce the most output, but those who recognise new possibilities and shape systems around them. This is a cultural shift as much as a technological one.

A more human view of the future of work

Seen through this lens, the future of work may not be a story about machines replacing humans. It may be a story about machines changing what human work means.

As skills become increasingly automated, humans move further upstream in the process. Instead of executing tasks, they shape the environments in which intelligent systems operate. This involves judgement, creativity, and curiosity. It involves recognising when something unexpected appears and deciding what it might become.

In philosophical terms, it means remaining open to the possibility of the event.

The first steps are already visible

None of this requires waiting for some distant technological future. The first steps are already visible in how people use AI today.

People are beginning to design systems rather than simply perform tasks. They structure workflows, define goals for agents, and decide how machines interact with real-world processes. These activities look less like traditional work and more like system design.

Which brings us back to that central idea - artificial intelligence may automate skills, but judgement remains the human layer that defines intent, checks reality, and inspires the creative leaps.

And those creative leaps - the moments when something genuinely new appears - are still the moments that shape the future.

References

Arendt, H. (1958; 2018 ed.). The Human Condition. University of Chicago Press.

Taleb, N. N. (2007). The Black Swan: The Impact of the Highly Improbable. Penguin.

Badiou, A. (2005). Being and Event. Continuum / Bloomsbury.

Badiou, A. (2009). Logics of Worlds. Continuum / Bloomsbury.

Brindisa turned 30 years of experience into a single source of product truth

Computational Thinking: The human skill behind AI agents

The ops foundation you build before the shock arrives

Explore More