THE AGE OF EFFICIENCY: MACH ADVOCATE KELLY GOETSCH ON AI'S GREAT REWRITE

Miya Knights
Oct 24, 2025
7 min read

The founder of nonprofit composable software advocacy group, the MACH Alliance, argues that AI is rewriting our socioeconomic contract with technology, reports Miya Knights, Publisher, The Promethean

When Kelly Goetsch took to the stage in London this week, he didn't bring a manifesto. He presented data points that, when taken together, mapped the tectonic shifts now reshaping enterprise software and, by extension, the digital foundations of modern industry and commerce.

Of all people, Goetsch should know. As the former Chief Strategy Officer for ecommerce software platform provider for large businesses, commercetools, he co-founded the nonprofit members group, the MACH Alliance , which advocates for a 'composable' approach to software integration, using microservices, application programming interface (API)-first, cloud-native, and headless systems.

Now, as Chief Executive of ecommerce logistics startup Pipe17, Goetsch addressed business executives, technology strategists, and business architects attending the inaugural MACH X event dedicated to building AI-ready organisations, with a clarion call to action.

His message was both pragmatic and provocative: we are living through one of those rare, career-defining inflexion points where the very code of business is being rewritten by artificial intelligence (AI).

THE INVESTMENT SINGULARITY

Goetsch opened with a stark metric: in 2025, 64% of all US venture-capital deals flowed into AI-native vendors. Not legacy software-as-a-service (SaaS) tools with a sprinkling of machine learning, but companies built from the ground up around AI architectures.

"Every dollar is basically being invested into AI," he noted, citing McKinsey's forecast that enterprise AI software could reach $180 billion of annual spend within a decade—roughly a fifth of today's trillion-dollar SaaS market.

This concentration, he argued, signals the beginning of a winner-take-most cycle. "It's not like the world needs fifty AI-enabled CMS [content management systems]," he said dryly. "They probably only need one or two."

Such software market consolidation suggests that a handful of AI-native super apps will dominate the next era of digital infrastructure, each absorbing functions that once required dozens of discrete platforms.

REWRITING THE STACK

The enterprise software ecosystem, long defined by disconnected, monolithic, and closed-loop systems that stifle business innovation and agility, is coalescing into a more intelligent and modular future, characterised by greater choice, where the only best-in-class solutions are those that offer superior capabilities and the best value.

The implication is profound. Across the trillion-dollar enterprise software market, Goetsch demonstrated how entire categories, such as content creation, data aggregation, commerce, and analytics, are being rebuilt from the ground up. Yet he cautioned against premature definitions: "We don't even know what an AI-native CMS is yet."

That uncertainty is itself revealing. It indicates a field still undergoing an evolutionary explosion, where experimentation precedes the establishment of taxonomy. The emerging systems will not simply automate old workflows; they will redefine value chains through prediction, generation, and autonomous orchestration.

In The Promethean's terms, this is the essence of AI as an evolutionary catalyst—a force accelerating the mutation of digital life forms rather than simply replacing their human hosts.

THE AGE OF EFFICIENCY

Goetsch's next theme, the "Age of Efficiency," served as both diagnosis and warning. The exuberant era of free money and "growth at all costs" has come to an end. Today's SaaS firms must prove they can generate profit, not just buzz.

He illustrated this with a metric familiar to investors: annual recurring revenue per thousand dollars of cost. Over the past few years, that ratio has tightened dramatically. Companies that once employed 208 people per $1 million in annual recurring revenue (ARR) now require 331; a jump reflecting both leaner operations and the productivity tailwinds of AI.

The shift, he suggested, is as much cultural as it is economic. "We've gone from in-office massages and talking about feelings all day to working on Saturdays and getting stuff done," he joked. Beneath his quip lies a serious point: automation is not freeing humans from work; it is intensifying the expectation of output.

Through The Promethean's lens, this is the paradox of progress: the same dynamic that powered the last Industrial Revolution. Efficiency breeds consumption. Goetsch invoked Jevons's Paradox, the 19th-century British economist's observation that improvements in energy efficiency lead to higher overall energy use. As coal-fired engines became cheaper to run, Britain burned more coal, not less.

So too with AI. As the cost per token plummets, usage is expected to skyrocket. "Soon, the price of anything in tech approaches zero," he said. "We're getting very close to that with AI." The likely result: an economy where computation becomes a near-free commodity, and imagination—what to ask, what to build—becomes the new scarce resource.

CODING AT THE SPEED OF THOUGHT

Goetsch's data from GitHub underscores the rapid pace at which this transformation is spreading through the creative layers of the enterprise. Developers using Copilot completed coding tasks in 1 hour and 11 minutes, compared to 2 hours and 41 minutes without it. "It's safe to assume most code today should be written with AI," he said, urging executives to challenge any IT department that blocks such tools on copyright grounds.

These gains are not a minor productivity boost; it is a redefinition of what authorship means. Code—the generative grammar of the digital economy—is becoming co-authored by machines. Junior engineers risk producing unreadable code if left unsupervised, but senior developers, armed with AI copilots, are multiplying their creative throughput.

Here, this magazine's central question bites: is this the evolutionary catalyst for a new era of human–machine symbiosis, or the unintended consequence of abdicating craftsmanship to algorithms? Goetsch's answer, implicitly, is pragmatic: adaptation is mandatory. "We are very firmly in the age of AI," he said. "Those who are not using it are operating to the detriment of their organisation and their own productivity."

"We are very firmly in the age of AI. Those who are not using it are operating to the detriment of their organisation and their own productivity."

THE NEW GROWTH CURVE

If the efficiency data signalled realism, the set of statistical information he shared radiated optimism. AI-native SaaS companies are reaching $1 million ARR in a single quarter—a velocity that dwarfs traditional SaaS providers' growth timelines. Investors are chasing these trajectories because they compound like algorithms, not factories.

But Goetsch drew a sharp line between AI-enabled and AI-native. Adding a "copilot" widget to an old platform, he argued, "does not make you SaaS. Adding a clippy does not make you AI-native." He clearly believes fundamental transformation requires re-architecting around intelligence, not accessorising with it.

AI-native firms allocate 57% of their headcount to R&D, compared to 35% in conventional software companies. They invest in products that "sell themselves," he said, relying less on marketing and more on intrinsic capability. Even their pricing models differ: consumption-based, metered, and dynamic, charging by tokens, words, or compute cycles rather than fixed contracts.

This economic grammar, which equates usage with value, is reshaping not only software pricing but also the nature of digital trust. When value is metered in real time, relationships between vendors and users become continuous conversations, rather than static transactions. It is, arguably, a more organic and evolutionary model of commerce that is adaptive, responsive, and agile.

FROM SAAS TO SUPER APPS

Looking forward, Goetsch imagined a world where businesses log into a "Business ChatGPT" instead of a patchwork of specialist tools. Within that interface, every enterprise function—CMS, production information management (PIM), analytics, customer relationship management (CRM)—renders as an AI-generated micro-application through modular component protocol (MCP) user interfaces (UIs).

Such a paradigm would invert decades of enterprise architecture. Instead of humans navigating systems, systems would materialise around intent. AI becomes the interface itself; the membrane between thought and execution.

For The Promethean, this vision resonates with the magazine's broader thesis: that the human-AI seam is becoming invisible. Where once the machine served as instrument, it now acts as interpreter and interrogator. The challenge is no longer access to information but governance of agency, which is what happens when the interface begins to decide.

"This is one of those handful of periods in an entire 40-year career when you can make your mark. The foundation for the next few decades is being laid right now — so go build something. Go experiment. Go play."

A CULTURAL REBOOT

Yet amid the data and diagrams, Goetsch's most human message came at the close. "This time we're living through," he said, "is one of those handful of periods in an entire 40-year career when you can make your mark. The foundation for the next few decades is being laid right now. So build something. Go experiment. Go play."

It was a call to reclaim enterprise software authorship at the very moment machines threaten to eclipse it. In The Promethean lexicon, this is the difference between evolution and consequence: whether we design the next systems or they redesign traditional ways of working.

THE PROMETHEAN PERSPECTIVE

Across his talk, Goetsch traced three intertwined trajectories:

Concentration of capital – AI has become the gravitational centre of investment, collapsing the software multiverse into a few dominant, intelligence-first constellations.
Compression of time – Development, scaling, and adoption are accelerating toward real-time cycles, demanding adaptive organisations and composable architectures.
Convergence of labour and code – Human work and machine generation are merging into a shared creative process, redefining efficiency and expertise alike.

Together, these forces generate an evolutionary process under pressure: a moment when digital organisms either mutate or go extinct.

For all his rational empiricism, Goetsch's closing sentiment carried a positive undercurrent of awe. The "AI bubble" he predicted will burst, reset, and rise again, like every technological wave before it. But what emerges after the trough of disillusionment may not resemble the software of today, or even the notion of software itself.

If AI truly becomes the substrate of business cognition, then the next frontier is not coding, but co-evolving: designing systems that learn, adapt, and create alongside us. The risk, as The Promethean's founding question warns, is that we mistake acceleration for advancement.

CATALYST OR CONSEQUENCE?

So, is AI an evolutionary catalyst or an unintended consequence? Goetsch's data suggests both. The catalyst lies in the efficiency gains, creative multipliers, and architectural revolutions already underway. The consequence lies in our response: whether we use those gains to deepen human ingenuity or to outsource it entirely.

As AI rewrites the world's software, it is also rewriting our socioeconomic contract with technology. We are no longer just users of tools, but the training data for systems that, increasingly, design themselves.

Goetsch urged his audience to participate—to "get technical" and leave a legacy. The Promethean would extend that invitation to all: to engage with this generative epoch not as spectators but as co-authors.

The story of AI will not be written by machines alone; it will be written by those who understand, as Goetsch does, that every line of code carries a question older than computing itself—whether invention is our greatest act of creation, or our most beautiful mistake.