Skip to main content

AI Is Eating the Job Market and Nobody Is Ready

00:08:11:10

AI Is Eating the Job Market and Nobody Is Ready

I have been working with AI tools every day for over two years now. I use them to write code, debug infrastructure, prototype ideas, and sometimes just to think through problems. AI has made me meaningfully faster at my job. But every time I see another round of layoffs where a company quietly replaces a team with an AI system, I get this uncomfortable feeling that most people are not paying attention to what is happening.

This is not a post about whether AI is good or bad. It is both. What I want to do is lay out the data, show you what is actually happening in the job market right now, and share what I think this means for anyone working in tech or really any knowledge work.

The numbers are hard to ignore

Let me start with the big picture. The World Economic Forum published their Future of Jobs Report in January 2025, surveying over 1,000 leading employers representing more than 14 million workers across 55 economies. Their projection: 170 million new jobs will be created by 2030, but 92 million will be displaced. That is a net gain of 78 million jobs, which sounds positive until you realize that 92 million people need to find completely new careers in a five-year window.

McKinsey's November 2025 report found that 57% of US work hours could already be automated with existing technology. Not future technology. Existing. They identified 40% of US jobs as "highly automatable," concentrated in legal and administrative services, routine data processing, and physically demanding repetitive work.

Goldman Sachs is more conservative, estimating that 6-7% of the US workforce could be displaced if AI is widely adopted, but they expect the unemployment impact to be temporary, with displacement effects typically disappearing after about two years. Their argument is that new jobs emerge to replace the old ones. Historically that has been true. But the speed of this transition is different from anything we have seen before.

What is already happening

The abstract numbers become real when you look at specific cases.

Salesforce cut 4,000 customer service jobs in 2025. AI agents now handle roughly 50% of their customer interactions. Their support workforce went from 9,000 to 5,000 employees. CEO Marc Benioff openly said "I need less heads." Support costs dropped 17%.

Klarna reduced their workforce from 5,000 to 3,000 employees, a 40% cut, largely through AI implementation. Their AI assistant handles the work of what they claim is equivalent to 700 full-time customer service agents.

IBM laid off 200 HR employees and replaced them with AI chatbots.

CNBC reported that over 50,000 layoffs in 2025 were explicitly tied to AI across multiple industries. And that is just the ones where companies actually admitted it. Most layoffs get framed as "restructuring" or "optimization." The real number is almost certainly higher.

The copywriting industry has been hit particularly hard. Writers report being forced to train AI tools on their own work, only to be laid off once the system was good enough. One support operations manager described being let go "the week before Thanksgiving" after his team was replaced by AI chatbots he had helped set up.

But the other side of the data is just as striking

Here is where it gets interesting. While some jobs are disappearing, the demand for people who know how to build and work with AI is exploding.

AI engineers now earn an average of $206,000 per year, a $50,000 increase from the previous year. Developers with AI skills earn 20-45% more than traditional counterparts. Senior LLM fine-tuning specialists make $195,000-$350,000. Prompt engineering demand surged 135.8% in one year.

90% of engineering teams now use AI in their workflows, up from 61% a year before. GitHub Copilot has over 15 million users, 4x what it had the year prior. 90% of Fortune 100 companies have adopted it. Developers report being 51% faster with these tools.

88% of organizations are regularly using AI in at least one business function, up from 78% the prior year. 92% plan to increase their AI investment over the next three years. Every dollar invested in generative AI yields an average return of $3.70.

ChatGPT alone hit 800 million weekly active users by late 2025. Over a quarter of US workers use it for their job. Among people with postgraduate degrees, that number is 45%.

The message from the market is clear: AI is not optional anymore. Companies are adopting it, paying a premium for people who understand it, and cutting the roles it can replace.

My take on all of this

I think learning AI, understanding how it works, and knowing how to use it effectively is going to be the single biggest career skill of the next five years. Not because AI is a fad (the $206k average salary and 20 million monthly SDK downloads tell you it is not) but because it is reshaping what it means to be productive.

I use AI every single day. I write code faster. I debug faster. I prototype faster. I think through architecture decisions by bouncing ideas off a model. My output has genuinely increased, not because I am working more hours but because the tools are that good. And they are getting better every few months.

But I am also concerned. Genuinely concerned.

The jobs that are getting automated first are not just repetitive factory jobs like in previous waves of automation. They are knowledge work jobs. Customer service. Copywriting. Translation. Basic programming. Legal research. Data entry. Accounting. These are roles that millions of people spent years training for, and the transition is happening fast enough that many will not have time to adapt.

The WEF report says 78 million net new jobs will be created. But "net new" does not help the 92 million people whose jobs are being displaced if the new jobs require skills they do not have. A customer service representative whose position was eliminated by an AI chatbot cannot just become an AI engineer next month. The skills gap is real, and most education systems are not moving fast enough to close it.

What worries me most is the speed. Previous technological shifts played out over decades. The transition from horse-drawn carriages to cars took 30+ years. The shift from physical retail to e-commerce took 20 years. The AI transition is happening in 3-5 years. McKinsey says 57% of work hours are automatable today, not in ten years. The technology exists right now.

What I think you should do

This is not a doom-and-gloom conclusion. I genuinely believe AI makes the world better on balance. It accelerates research, makes information accessible, helps with education, assists doctors with diagnoses, and enables small teams to build things that used to require hundreds of people. That is incredible.

But you cannot afford to sit this one out and hope it does not affect your field. Here is what I would focus on:

Learn how AI works at a conceptual level. You do not need to understand transformer architectures in detail, but you should know what an LLM is, what tokens are, what context windows mean, what tool calling does, and what the limitations are. Understanding the technology helps you see where it actually adds value and where it is just hype.

Get hands-on with the tools. Use ChatGPT, Claude, Cursor, Copilot, or whatever fits your workflow. The gap between someone who uses AI tools effectively and someone who does not is already large and it is growing. In software development specifically, the difference is dramatic. Developers using Copilot report 51% faster coding. That compounds over months and years.

Build things with AI APIs. If you are a developer, learn the AI SDK, learn how to build agents, learn how to do structured output and tool calling. These are the building blocks of every AI product being built right now. The salary data backs this up: AI-skilled developers earn 20-45% more.

Develop skills AI cannot easily replace. Complex problem solving, system design, understanding user needs, making judgment calls in ambiguous situations, building relationships, leading teams. AI is great at execution but still bad at figuring out what to execute. The people who can direct AI toward the right problems will be the most valuable.

Do not wait for your employer or your school to train you. The pace of change is too fast for institutions. The people who are thriving right now are the ones who started experimenting early, on their own time, out of curiosity. That advantage compounds.

The uncomfortable truth

The data tells a split story. AI is simultaneously the greatest career accelerator and the greatest career threat, depending on which side of it you are on. 78 million new jobs and 92 million displaced. $206k salaries for AI engineers and mass layoffs for customer service reps. 51% faster development and entire copywriting teams let go.

I do not think this is a future scenario anymore. It is the present. The question is not whether AI will change your industry, it is whether you will be ready when it does.

Sources