Everyone Is Going To Become An Engineer - Kevin Siskar

TEXT START: For the last two years, almost every conversation about AI and work has started with the same question: Are software engineers going away?

The Dissection

The text reframes AI displacement as universal technical upskilling. Its real message is that knowledge workers should stop merely executing tasks and start designing the systems that execute them.

That diagnosis correctly identifies the immediate transition layer: workflow automation, agent orchestration, API integration, measurement, and human oversight. The Stripe and Box roles are presented as evidence that companies will embed technical operators inside every department.

But the article quietly changes the subject. It begins with whether jobs disappear and ends by describing how jobs are converted into automation systems. The labor is not preserved; the labor process is compressed. A high-paid AI operator can redesign workflows that previously required entire teams. The article treats the person building the machine as proof that the people consumed by the machine still have a future.

The Core Fallacy

The central error is confusing a new skill layer with mass productive participation.

“Everyone becomes an engineer” is not the same as everyone remains economically necessary. Under the Discontinuity Thesis, the decisive sequence is:

P1: AI gains durable cost and performance superiority across cognitive work.

P2: Institutions cannot preserve stable human-only domains at scale.

P3: The majority lose access to economically necessary labor.

The article recognizes P1 in miniature but refuses to follow it through P2 and P3. It assumes the economy will contain enough system-building work for everyone who learns the engineering mindset. That is the exact assumption the automation trend destroys. Once agents can build, debug, connect, document, and improve workflows, the system-builder role also becomes a target for automation.

The text mistakes occupational mutation for occupational survival. It is a clean description of how the guillotine is being redesigned, not evidence that there will be enough people left to operate it.

Hidden Assumptions

• Every knowledge worker can become a credible builder of AI systems, despite unequal access to technical ability, data, authority, capital, and organizational trust.
• Companies will retain large numbers of human operators after workflows become automated.
• The amount of system-design work will expand faster than AI reduces the need for human labor.
• Human judgment, coaching, and domain knowledge will remain scarce rather than being encoded into agents and playbooks.
• A role with a salary of $145,000–$218,000 demonstrates broad opportunity rather than the concentration of leverage in a small technical elite.
• The organization will reward employees who automate their own work instead of using the automation to eliminate positions or reduce headcount.
• Human handoffs, oversight, and “AI maturity” will remain economically necessary rather than becoming temporary transition scaffolding.
• The individual can control the systems they build. In reality, ownership, permissions, compute, proprietary data, and distribution remain concentrated with Sovereigns.
• The distinction between “using AI” and “building with AI” will remain stable. As tools improve, that boundary collapses.
• The future is a larger population of augmented workers rather than a smaller ownership class supervising increasingly autonomous productive systems.

Social Function

This is partial truth wrapped in transition management and ideological anesthetic.

It is partial truth because the article accurately identifies the first-order workplace shift: manual computer work is being reorganized into automated workflows, and people who can implement that transition are temporarily valuable.

It is transition management because it gives incumbent knowledge workers a role they can recognize: learn the tools, redesign the workflow, coach the team, and remain relevant.

It is ideological anesthetic because “everyone becomes an engineer” converts a distributional crisis into an individual competency problem. If displacement arrives, the implied failure belongs to workers who did not become technical enough. The ownership question disappears. So does the possibility that one engineer, one agent fleet, or one platform replaces dozens or thousands of employees.

The article is also prestige signaling for the emerging AI operator class. It presents proximity to tools such as Claude Code, Codex, agents, and automation frameworks as a path to altitude. That path exists, but it is narrow, competitive, and temporary. Most participants become servitors of AI capital, not Sovereigns of it.

The Verdict

The article correctly sees that AI will turn departments into software systems. It is catastrophically wrong about what follows.

The future is not universal engineering. It is selective ownership, concentrated system control, and broad loss of economically necessary labor. The “engineer mindset” is a temporary survival credential for a shrinking layer of transition operators. As AI begins engineering the engineering layer, the article’s proposed escape route becomes another queue at the same disappearing jobs market.

Under DT logic, this is not a rebuttal to obsolescence. It is a polished description of obsolescence arriving through workflow redesign instead of a single dramatic layoff.