Everyone's debating whether AI will replace workers. They're missing the revolution: it's already augmenting millions of us.
Edward Zitron’s "The Case Against Generative AI" makes a case that the industry is a financial bubble, executives chasing automated knowledge work are delusional, and LLMs can’t do real software engineering … but he’s watching the wrong revolution. While fixated on “corporate replacement”, he’s missing massive augmentation on millions of laptops and mobile phones.
Zitron argues that LLMs cannot perform true software engineering, and that they fail at the complex process of maintaining large systems. This is true for senior engineers maintaining massive systems, but … irrelevant for everyone else. Creating a simple web app, or a custom script, or even a functional website used to require specialized knowledge. That mountain is now a “manageable hill”.
My experience isn't about telling an AI to "build me a new SaaS company." It's a dynamic, iterative partnership. I start with an idea. The AI provides the scaffold: the boilerplate code, the syntax I’ve forgotten, the initial structure. I then step in to refactor, to add the unique logic and a creative spark. When I hit a bug, I have a Socratic dialogue with a partner who spots my error and explains the fix. This collapses the barrier between idea and execution, creating a new generation of builders. Zitron’s analysis completely omits this creative agency.
The same is true for intellectual work. The article dismisses AI's writing capabilities by arguing it cannot replicate the human process of fighting with structure, style, and intent. It's the wrong lens. AI isn’t the author; it is a cognitive prosthesis. I've used it to explore connections between disparate fields, finding patterns and linkages I wouldn't have spotted on my own.".
I can ask it to adopt the persona of a critic to find flaws in my own arguments. I can feed it a complex theory and ask it to generate analogies until I find one that clicks. It acts as a tireless research assistant and an intellectual sparring partner, helping me to synthesize information and deepen my own thinking. This isn't automating the essayist; it's augmenting the individual's capacity for curiosity and insight.
This brings us back to Zitron's strongest point: the broken economics. He rightly questions how a business model that loses money on its most engaged users can ever succeed. The current push to sell AI as a replacement for enterprise seats is the wrong model entirely.
The value isn't in firing 10% of your marketing department. The ROI is in the cumulative value of millions of augmented individuals. It’s the small business owner who can finally build their own inventory management tool. It’s the scientist who can analyze data and spot correlations faster. It’s the student who can grasp complex topics more deeply. This value is diffuse, hard to quantify on a quarterly earnings call, and looks nothing like the traditional SaaS model.
The generative AI bubble may very well burst. The current valuations are indefensible, and the corporate obsession with "full automation" is a dead end. But the technology's fundamental value is real. The revolution isn't that the machines are replacing us. The revolution is that they are making us more capable, more creative, and more productive than ever before. The failure isn't AI; it's the extremely unimaginative way today's executives are trying to deploy it.