There is a point where tech assists in running a business to become a liability when we increase our dependence on it. I think it can be difficult to spot that point when we’re busy chasing goals. There are so many different perspectives.
Two things during the week caught my attention:
AWS
Two conflicting perspectives on AWS arose in a conversation during the week, and are sitting with me. The first - my own - was that it represented a demonstration of the danger of creating ‘single points of failure” in critical supply chains owned by private companies that could leverage that dependency.
The other, that an organisation carrying a third of the world’s internet commerce could suffer a major failure and be back up and running in twelve hours, is remarkable and that the capability relied on the very economics of scale that worried me.
The best way I know to resolve paradoxes is through conversations in small, trusted groups where you can argue the point and not the person. To argue from the standpoint of a theory rather than a hill to die on.
Perspectives are like that. They are a kaleidoscope that we can use to create different images using the same materials. Those different perspectives can lead us down really beneficial side roads. For instance, how much of the AWS problem was caused by the fact that they had probably got rid of many of the people who understood the origins of the problem, making recovery of it more difficult. The other side to that argument is that they have the systems that have captured that information. And so, although recovery was slower, it was robust and could be repeated.
The Games Industry
Widening the thought beyond AWS, I’ve long been fascinated by the games industry because it sits at the intersection of creative craft and technological infrastructure, making it an ideal lens for understanding how financial logic is reshaping creative work.
The video game industry is experiencing a paradox that reveals something profound about our economic moment. In 2023, the industry generated an estimated $406 billion in revenue, with projections reaching $666 billion by 2029. Yet since 2022, approximately 45,000 gaming workers have lost their jobs. Microsoft laid off nearly 12,000 gaming employees across multiple rounds while posting record profits. Embracer Group collapsed from over 15,000 employees to fewer than 8,000, closing 44 studios. Sony shuttered successful studios immediately after launching their games.
This has nothing to do with failure or economic necessity. These are profitable companies making strategic choices. The question is: why?
I think the answer lies in understanding three converging business dynamics that are transforming not just gaming, but all creative and knowledge work.
First, financial time horizons have collapsed.
Since the 1980s, shareholder value maximisation and quarterly earnings pressure have created a fundamental bias: proven projects always beat unproven ideas when you’re optimising for next quarter’s profits. Sequels are predictable; innovation is risky. Established IP can be modelled; experimental work cannot. It creates a systematic preference for extraction over exploration, and refinement over revolution.
To maximise short-term profits from proven formulae, companies need workforce flexibility: the ability to scale up for production and cut immediately afterwards. This requires projectification, converting permanent employment into temporary project-based work, eliminating “inefficiencies” like institutional knowledge, training programmes, and exploratory research.
Second, this creates a fractured structure
The emerging structure has fundamentally different realities. There’s a permanent layer that controls financial capital, relationships, and IP. These are the executives and platform owners who operate with portfolio logic and capture asymmetric value. Then there’s a temporary layer providing only operational skills: the producers, developers, artists, and creators who are hired project-by-project, bear full downside risk, and experience constant precarity. This bifurcation exists across both large corporate entities and independent creators, creating four distinct quadrants of work, each with its own economic logic and lived experience.
The consequences are already visible: systematic knowledge loss as experienced workers leave the industry, innovation deficits as only safe bets get funded, and quality degradation from constant team reassembly.
Third, AI enters as an accelerant, not a cause.
AI enables cheaper execution of proven formulae, provides ideological cover for cuts (through claims of “efficiency gains”), and concentrates power in those who control the infrastructure. It creates a vicious cycle: projectification leads to knowledge loss, which increases reliance on proven formulae, which are perfect for AI optimisation, which justifies more projectification.
The Nature of the Revolution
Economist Carlota Perez’s work identified that major technological revolutions follow a predictable two-phase pattern.
First comes the Installation Period, where financial capital dominates and new technologies trigger speculative investment booms, asset inflation, and a decoupling between financial markets and the real productive economy. This phase is characterised by short-term thinking, creative destruction, and instability, culminating in a crisis or bubble burst.
Then, if institutional reforms occur (through labour organising, regulatory intervention, and new social contracts), comes the Deployment Period, where production capital takes the lead, the technology matures and diffuses widely, stable organisational forms emerge, and the potential exists for a “golden age” of broadly shared prosperity.
The transition between these phases is the Turning Point, a crisis moment that forces societies to choose: either reform institutions to align with the new technology’s potential, creating conditions for productive growth and shared benefits, or fail to reform and face prolonged turbulence with persistent inequality and instability.
Crucially, Perez demonstrates that technology itself doesn’t determine outcomes. Political choices about institutional structures, labour power, and wealth distribution determine whether a technological revolution produces golden ages or lost decades.
Perez’s framework suggests we’re in the late ‘Installation Period’ of the broader digital revolution that began in the 1970s. AI isn’t a separate revolution; it’s the latest wave within this longer transformation, accelerating dynamics that have been building for decades..
Where we find ourselves is a phase characterised by financial capital dominance, short-term thinking, speculation, and instability. This phase always ends in crisis and restructuring. The question is what comes next: a productive “Deployment Period” with broadly shared prosperity, or prolonged turbulence with deepening inequality.
History shows the difference depends on institutional reform through labour organising, government intervention, and new social contracts.
The technology doesn’t determine the outcome. Politics does.
The games industry isn’t an isolated case. It’s a leading indicator of where all creative and knowledge work is headed under current financial logic. What we’re witnessing isn’t inevitable technological disruption. It’s organisational choices driven by Installation Period incentives, amplified by AI tools.
This raises urgent questions
About the future of creative work, the sustainability of innovation, and the structures we need to build.
In future posts in The Athanor, I’ll explore the nuances of this transformation: the specific mechanisms of each quadrant, the psychology of permanent precarity, the strategic options for
and organisations, the policy interventions that could enable different outcomes, and the historical parallels that show us both the dangers and the possibilities of this moment.The crisis is real, but the future remains unwritten.
That is our job, not technologys’.
These are interesting times, full of opportunity, as well as the perils of complacency.
From the start of November, I am changing the balance of my writing. Here at ‘Outside the Walls’, I will be focusing on the bigger changes taking place around us that impact our relationship with work, and will be writing a combination of short observations and longer form pieces.
At I will focus on the impact on us as individuals, again via short observations and longer form pieces.
And at The Athanor, I will be working with a small group as we develop specific individual strategies for thriving in the changing world of work, and sharing what we learn with subscribers.
I fully agree that politics, not technology, will determine how things evolve. The issue I see: the technology we have is global. Our supply chains are global. Company structures are global. All not equally distributed globally, but with global-ish reach. So, what needs to be managed is global.
In politics, US and China have global-ish reach. Nobody else has. German politics, Brazilian politics, South Korean politics: No way they can manage global-ish things. But local is still the level on which most of politics is made, and to some extent needs to be: We still have many and strong language and cultural barriers, and these determine many limits in politics, while companies, technology, and supply chains have overcome them.
So what happens if local-ish politics tries to rein in global-ish trends? Not much. And I do not see politics becoming more global. Looks like the opposite to me.
I therefore find it rather likely that those with global-ish reach (two governements, a number of companies, maybe some individuals) will determine the outcomes - to their benefit, obviously. I would be very happy to be shown believable pathways to more balanced influences and outcomes. Any takers?