Everyone Is Going To Become An Engineer
For the last two years, almost every conversation about AI and work has started with the same question:
Are software engineers going away?
I think that is the wrong fear.
The more interesting possibility is almost the exact opposite.
Software engineering does not disappear. Instead, the engineering mindset spreads into every other computer-based job. Marketing becomes more like engineering. Sales becomes more like engineering. Customer support, finance, recruiting, operations, and strategy all start to become more technical, more automated, and more systems-driven.
The old version of these jobs was built around doing the work manually. Write the email. Pull the report. Update the spreadsheet. Respond to the ticket. Build the campaign. Copy information from one SaaS tool into another. Follow up with the customer. Chase the internal approval. Move the workflow forward by hand.
The new version is different. The job is no longer just to do the work. The job is to understand the work deeply enough to build the system that does the work.
That is the shift.
Stripe just showed us the new job description
The clearest signal I have seen so far came from Stripe.
Stripe opened a role called Forward Deployed AI Accelerator, Marketing. The title sounds a little abstract, but the job description is one of the most important AI labor market signals I have seen. Stripe says its marketing organization is undergoing a “fundamental transformation” and that it is building a team embedded directly with marketers to make AI the default mode of work, not an occasional tool.
That sentence matters.
Stripe is not hiring someone to make more marketing content. It is hiring someone to rebuild how marketing works.
The role embeds with roughly 20 marketers, studies their day-to-day workflows, identifies the highest-leverage work worth transforming, builds custom tools, agents, automations, and skills for each person, coaches the marketers toward self-sufficiency, documents the playbooks, and tracks progress against an AI maturity model.
That is the whole future of knowledge work hiding inside one job description.
It is one part marketer, because the person has to understand the actual work. It is one part AI engineer, because the person has to build agents, automations, workflows, and tools. It is one part consultant, because the person has to study messy human processes and turn them into repeatable systems. And it is one part coach, because the real outcome is not just a tool being built. The real outcome is the team learning to operate differently.
The requirements make the shift even clearer. Stripe wants someone with a hands-on track record building AI-powered tools, agents, automations, or workflows that changed real work processes. Not someone who merely uses AI as a chatbot. Someone who builds with it. The preferred stack includes Claude, Claude Code, Codex, custom agent frameworks, API integrations, and workflow automation tools.
The salary range is serious too. Stripe listed the US base salary range for the role between roughly $145,000 and $218,000. Business Insider separately reported another version of the role with a range up to about $198,000.
That is not an intern experimenting with prompts. That is a serious operating role at one of the most sophisticated technology companies in the world.
This will not stay in marketing
The reason this role matters is not because Stripe needs a better marketing operations person. The reason it matters is because every department has the same hidden structure underneath it.
Every function is full of repetitive workflows trapped inside inboxes, spreadsheets, SaaS tools, Slack threads, CRM fields, support queues, calendars, dashboards, and human memory. Most companies are held together by thousands of tiny manual handoffs that nobody has had the time, technical ability, or organizational permission to automate.
That is what AI changes.
Once this cohort + coach + build model works in marketing, the same job description gets copied into sales, customer support, finance, legal, recruiting, and operations. The surface area changes, but the pattern is the same. Embed with the people doing the work. Understand the workflows. Find the leverage. Build the tool. Coach the person. Turn the one-off solution into a reusable playbook.
Box is already showing another version of this pattern with its AI Business Automation Engineer role. That role is designed to embed with subject-matter experts across finance, legal, people, go-to-market, customer success, and other functions to identify business processes worth reinventing and rebuild them with an AI-first mindset.
The titles will vary. Forward Deployed AI Accelerator. AI Business Automation Engineer. Agent Operations Lead. Workflow Engineer. Internal AI Systems Builder. Function-specific AI Operator.
But the pattern will be the same.
The most valuable person in the department will be the person who understands the function deeply enough to automate it.
The org chart becomes a work chart
Microsoft’s Work Trend Index points in the same direction. Microsoft describes the rise of “Frontier Firms,” companies that organize around human-agent teams instead of only traditional job functions. It also says leaders are already considering new AI-specific roles like AI trainers, AI agent specialists, AI ROI analysts, and AI strategists across marketing, finance, customer support, and consulting.
That is the transition.
Companies are not deleting jobs. They are deleting the old shape of jobs.
The person who only does a single task becomes less valuable. The person who can turn a series of tasks into a system becomes more valuable. The person who waits for engineering to build them a better dashboard becomes less valuable. The person who can build the dashboard, wire the workflow, train the agent, measure the outcome, and teach the team becomes more valuable.
This is why “every person becomes an engineer” matters.
I do not mean every person needs to become a traditional backend engineer. I do not mean every marketer needs to understand distributed systems, memory management, or database internals. I mean every knowledge worker is going to need an engineering mindset.
They will need to understand the raw materials of modern work:
Workflows
Inputs and outputs
APIs and integrations
Data quality
Automation
Agents
Permissions
Failure states
Feedback loops
Measurement
Security
Human handoffs
That used to sound like engineering language. Now it is becoming operating language.
Coding is the leading indicator
Software development is not just another job category. It is where the future of computer work usually shows up first.
Developers are already learning how to manage agents, review AI-generated work, give feedback, debug outputs, orchestrate tools, and ship faster through human-machine collaboration. Anthropic’s research on Claude Code found much higher automation patterns in coding-agent usage than in normal chatbot usage. That should not be dismissed as a niche developer behavior. It is a preview.
The rest of the company is about to learn the same pattern.
The marketer becomes someone who can build a campaign machine. The support person becomes someone who can build an AI triage and response system. The salesperson becomes someone who can build a prospecting, enrichment, and follow-up engine. The finance operator becomes someone who can automate reporting, reconciliation, and variance analysis. The recruiter becomes someone who can build a sourcing and screening workflow.
The function does not disappear overnight.
But the old version of the function starts to look slow.
My prediction for the next two to three years
Over the next two to three years, I think we will see a clean split inside companies.
On one side will be people who use AI as a slightly better search box or writing assistant. They will prompt ChatGPT, copy the answer, paste it somewhere else, and keep doing most of their work the old way.
On the other side will be people who use AI as a systems layer. They will build small tools around themselves. They will connect agents to workflows. They will automate the annoying parts of their jobs. They will create reusable playbooks. They will teach their teammates how to operate differently. They will turn their own role into leverage.
That second group is going to compound.
They will not just be better at their jobs. They will redesign their jobs.
This is the real story hiding underneath all the noise about AI replacing workers. The future does not belong to people who merely use software. It belongs to people who can shape software around their work.
Every person is going to become an engineer, not because software development goes away, but because software keeps eating the rest of the org chart.
