The AI disrupting the technology industry doesn’t look like a monster. It isn’t smashing buildings or declaring independence. It shows up quietly, generates enormous amounts of output, and never sleeps. Like Peter Boyle’s unexpectedly civilized creature in Young Frankenstein, it’s polite, useful, even endearing. And it is endowed with immense strength. Not the strength to destroy the lab, but the strength to run it.
This matters because AI isn’t beating the best engineers at their own game. It isn’t writing better code than elite developers. What it is doing is writing vastly more code, vastly faster, at a cost structure that no human workforce can match. In many organizations, quantity is beginning to outrun comprehension. The bottleneck is shifting from production to oversight.
That single shift is enough to reorder an entire industry.
Software Development is Becoming Output Management
For decades, software engineering was constrained by human typing speed, attention span and headcount. AI removes those constraints. Code, documentation, tests, refactors, migrations, prototypes and feature variants can now be generated in torrents.
But software doesn’t become valuable when it is written. It becomes valuable when it works reliably in production.
That gap is widening. Engineers increasingly act as reviewers, integrators, and risk managers rather than primary authors. Companies are discovering that accelerating output without equal advances in validation can create fragile systems at scale. After outages tied to AI-assisted changes, some firms now require senior engineers to sign off on certain deployments. When productivity gains require additional governance to keep systems stable, you are no longer simply improving a process. You are changing its physics.
The “one-person unicorn” narrative emerges from this environment. A tiny team can generate the output that once required an army. Whether they can safely operate it is a different question.
The Disruption is Moving Up and Down the Stack
The market is reacting not to a single innovation but to a cascade of second-order effects.
At the application layer, AI threatens traditional revenue engines such as search advertising and seat-based SaaS licensing. If an agent can perform the work of several employees, the logic of charging per employee starts to erode. Buyers begin asking not how many users they have, but how much work needs to be done.
At the infrastructure layer, demand for compute, power, land, and water is exploding as data centers scale to support AI workloads. These facilities are no longer abstract “cloud” resources. They are industrial projects with environmental footprints and geopolitical implications.
At the capital layer, investors are beginning to question the return on massive AI spending even as they fear being left behind if they underinvest. Weak enterprise demand for traditional software products sits uneasily alongside record spending on AI capacity. The industry is simultaneously consolidating and expanding.
At the labor layer, hiring patterns are shifting dramatically. Entry-level roles are tightening while demand grows for senior talent capable of supervising complex systems, managing risk, and translating AI capabilities into real business outcomes. The golden age of abundant Big Tech employment is giving way to a smaller, more specialized workforce.
This is not a localized disruption. It is systemic.
Enterprise Buyers are Changing Behavior
Perhaps the clearest signal comes from customers rather than vendors.
Enterprise software buyers are reassessing priorities, delaying commitments, renegotiating contracts, and demanding proof of value in an environment where the ground is moving under their feet. Why lock into multi-year agreements for tools that may be reshaped or replaced by AI-native alternatives within a short window?
In effect, uncertainty is becoming a purchasing factor. When technology cycles accelerate, the safest decision is often to wait.
That hesitation reverberates back through the supply chain. Vendors see softer demand. Sales cycles lengthen. Financing conditions tighten. Even industry leaders must tap debt markets under less favorable conditions. What looks like a slowdown in one segment is often a shockwave from structural change elsewhere.
Business Models are Being Cannibalized From Within
AI is unusual in that it does not simply create new markets. It directly undermines existing ones.
Software companies are embedding AI into their own products to remain competitive, even when doing so reduces revenue tied to traditional licensing models. Service firms deploy automation that shrinks billable hours. Platform providers introduce capabilities that make third-party ecosystems less essential.
In other words, the industry is forced to dismantle parts of itself to survive.
This is why layoffs framed as “AI transformation” are occurring alongside record investment. Organizations are reallocating resources from human-intensive operations to capital-intensive infrastructure. The total spending may not decline, but its distribution changes dramatically.
The Monster Isn’t Revolting. It’s Advancing.
The most striking aspect of this moment is how undramatic it looks on the surface.
There is no singular breakthrough event that clearly marks the transition. Instead, thousands of incremental decisions — adopting AI tools, automating workflows, delaying hires, restructuring teams, building data centers, revising pricing — accumulate into a transformation that only becomes obvious in hindsight.
Like the civilized monster who can read, speak, and perform parlor tricks, AI integrates itself into daily operations until removing it becomes impossible. It does not need to overthrow the doctor. It simply becomes indispensable.
And once a creation can operate the lab more efficiently than its creator, authority shifts, whether anyone intends it to or not.
Tech is the First Domino
The technology sector sits at the epicenter because it produces both the tools and the infrastructure of AI. It also depends heavily on digital labor, scalable processes, and rapid innovation cycles — exactly the conditions where automation has the greatest leverage.
Other industries will follow, but more slowly, constrained by physical assets, regulation, and cultural inertia. Manufacturing plants, hospitals, transportation networks, and government systems cannot be rewritten overnight.
Software can.
What we are witnessing in tech is therefore not a special case but an early preview. The patterns emerging now — output outpacing oversight, capital replacing labor, buyers hesitating amid uncertainty, infrastructure straining to keep up — are likely to repeat across the broader economy.
The Doctor is Still in the Room — For Now
None of this means humans are obsolete or that AI will autonomously dominate the industry. The systems still require direction, judgment, and accountability. The best engineers, architects, and leaders remain indispensable.
But the balance has changed. The creation amplifies some roles while compressing others. It rewards those who can harness it and sidelines those whose value was tied to the scarcity of output.
Frankenstein’s mistake was not creating the monster. It was failing to anticipate what would happen once the creature could operate independently.
Silicon Valley is now confronting a similar realization. The tools it built to accelerate innovation are accelerating everything, including the disruption of the industry itself.
And because tech sits upstream of nearly every other sector, what happens here rarely stays here.
The monster is not on a rampage. It is doing exactly what it was designed to do — applying enormous capability to real problems at scale.
That may be even more consequential.
