Coding agents now write a growing share of the world’s software. Yet the AI impact on IT jobs looks nothing like the doom headlines. Software developer job postings in the United States are up almost 15 percent since February 2025, Indeed Hiring Lab found.
New here? I publish one agentic AI governance post every weekday. Subscribe to the blog and it lands in your inbox the moment it goes live.
Still, the jobs did not stay the same. Senior roles drove 71 percent of the gain, and roles with AI in the title drove 37 percent. The work moved, and history says that is the usual pattern. Typists, bookkeepers, bank tellers and cartoonists all watched a new machine arrive. In each case, the skill changed shape, and the workers who adopted the new tool early kept working.
The typing pool went digital
In the 1980s, typing was a trade. It lived in the typing pool, on carbon paper and correction tape. But then WordPerfect, Lotus 1-2-3 and desktop publishing software landed on office desks. By 2004, the Labor Department projected a loss of 93,000 word processor and typist jobs over the next ten years. It was the third-largest drop on the agency’s list.
Still, the skill itself did not die. It spread. Instead, typing became something nearly every office worker did on a screen. The people who learned the new software early ran the modern office. Those who waited for the typewriter to come back lost the chance.
Spreadsheets hit the bookkeepers
VisiCalc and then Lotus 1-2-3 did the same thing to accounting. Software replaced many bookkeepers and accounting clerks, NPR’s Planet Money reported. Yet the number of accountant jobs went up, not down.
Cost explains why. When a “what if” took seconds instead of a weekend, firms asked for far more of them. So the work shifted from adding columns to judging what the numbers meant.
Bank tellers lived through a similar turn, and the AI impact on IT jobs echoes it. Cash machines cut the cost of running a branch, so banks opened more branches. Teller jobs grew about 2 percent a year in the late 1990s, economist James Bessen found. Tellers spent less time counting cash and more time solving customer problems.
Cartoonists learned the computer
Animation went through the same wringer two decades ago. By 2004, studios were moving from hand-drawn films to computer graphics, Animation World Network reported. The animators’ union did not wait for the layoffs. It opened a training lab and taught about 800 members new tools over four years.
DreamWorks ran up to 250 classes, and retraining was its largest track. Background painters moved into lighting, because their eye for light still mattered. Timing, weight and performance carried over. Only the pencil changed.
More on the Agentic AI SDLC
- What Changes: How the lifecycle shifts in my comparison of agentic AI SDLC and traditional SDLC.
- What to Unlearn: The agile habits I would keep, retire or rework when agents write the code.
- Step by Step: My five-stage guide for delivery managers moving teams from prompted to autonomous.
- AI-Native Work: Why the AI-native SDLC puts intent, not code, at the center.
Coding agents are the next typewriter
Today, coding agents are doing to software what WordPerfect did to the typing pool. The new demand is for people who review AI output, set the architecture, wire tools together and check security. That fits the Indeed data, which showed employers paying up for experience.
The pain is real at the entry level. Anthropic’s study of labor market impacts found no broad rise in unemployment for exposed workers since late 2022. But job finding for workers aged 22 to 25 in exposed fields fell about 14 percent. So the AI impact on IT jobs lands hardest on people who have not yet built the judgment employers want.
Adoption decides who wins
In short, every shift above rewarded the same thing. It was not raw talent. It was the speed of adoption. The people who learned WordPerfect, Lotus or digital lighting early had the best shot at the next job.
So here is my advice for anyone worried about the AI impact on IT jobs. Pick one agent tool this month and ship something real with it. Then learn to review, test and govern what it builds.
Instructions in, results out was IT; intent in, outcomes out is agentic AI. In my view, the AI impact on IT jobs will favor people who can state intent and judge outcomes. That skill is learnable, and it is yours to take.
What skill have you already moved from the old tool to the new one? Tell me in the comments.
Was this post useful? One click tells me what to write next.
Subscribe to my LinkedIn Newsletter
- Agentic AI Architect: control design for enterprise agents.
- Agentic AI Case Studies: deployment evidence, one teardown at a time.
- Agentic AI P&L: cost, payback, and risk in a CFO’s voice.
- Agentic AI SDLC: the lifecycle that builds the agents.
- Subscribe to the Blog: I publish a new post every weekday on one agentic AI governance question. Enter your email and each post arrives in your inbox the moment it goes live.
© Dr. Harish Kotadia, Ph.D., All Rights Reserved, 2026
Dr. Harish Kotadia, Ph.D., is an Enterprise AI Architect with 20+ years of IT consulting experience serving Fortune 100 clients, specializing in agentic AI systems built on Anthropic Claude, AWS Bedrock, and Google Vertex AI.
Disclaimer: This blog post is based on publicly available academic publications, vendor documentation, open standards, and news items from reputed media sources linked above. This post is intended for educational purposes, to help the enterprise agentic AI community build a shared vocabulary from public, authoritative sources.
Views and opinions expressed here are my own and do not represent those of any employer or client, past or present. The analysis presented is my independent interpretation of the published sources linked above and does not constitute legal, financial, or consulting advice of any kind.