Over the last few years, there’s been no shortage of headlines warning that AI is coming for our jobs. At first, I brushed some of it off as hype — social media feeding me content based on one or two searches I made. But by 2022, I started seeing the writing on the wall. By 2023, I was telling colleagues: this technology is going to start replacing jobs.
Some of them nodded, but didn’t see it yet.
Now, in 2025, we’re seeing it play out in real time. The signs are undeniable.
The Changing Shape of Work
Junior programming roles are quietly disappearing. Mid-level engineering teams are shrinking. Middle management is being re-evaluated or cut. Large companies like Google, Meta, and Microsoft report that over 30% of their new code is already written by generative AI — and analysts are projecting that number could surpass 50% by 2026.
I’ve seen it first-hand. As someone who adopted GenAI early, what got me excited wasn’t just the novelty — it was the power. You could generate working code, spin up infrastructure, draft entire workflows, and automate time-consuming parts of your job in minutes. But there’s a catch: building your own AI models is expensive. The cost barrier is real. That’s why most innovation is happening either inside well-funded companies or scrappy startups leveraging existing models.
Over the last few years, there’s been no shortage of headlines warning that AI is coming for our jobs. At first, I brushed some of it off as hype — social media feeding me content based on one or two searches I made. But by 2022, I started seeing the writing on the wall. By 2023, I was telling colleagues: this technology is going to start replacing jobs.
Some of them nodded, but didn’t see it yet.
Now, in 2025, we’re seeing it play out in real time. The signs are undeniable.
The Changing Shape of Work
Junior programming roles are quietly disappearing. Mid-level engineering teams are shrinking. Middle management is being re-evaluated or cut. Large companies like Google, Meta, and Microsoft report that over 30% of their new code is already written by generative AI — and analysts are projecting that number could surpass 50% by 2026.
I’ve seen it first-hand. As someone who adopted GenAI early, what got me excited wasn’t just the novelty — it was the power. You could generate working code, spin up infrastructure, draft entire workflows, and automate time-consuming parts of your job in minutes. But there’s a catch: building your own AI models is expensive. The cost barrier is real. That’s why most innovation is happening either inside well-funded companies or scrappy startups leveraging existing models.
And that brings me to another major shift: the rise of AI-native startups.
More people I know are walking away from traditional 9–5 jobs and launching businesses powered by GenAI. Instead of working for a company, they’re using AI to become the company. Startups today are leaner, faster, and in many cases, more effective because they don’t need bloated teams — they need a few highly skilled people who know how to harness AI tools.
The Talent Companies Need Now
To compete in this environment, companies must evolve. That means:
- Reducing bloated team structures
- Hiring talent that knows how to work with GenAI
- Redesigning workflows around automation and intelligent tooling
The companies that thrive will be the ones that slim down operational inefficiencies and double down on capability. If you’re not streamlining with AI, someone else is — and they’re doing it faster and cheaper.
Let’s take a concrete example: DevOps.
A few years ago, deploying an app to a remote server with CI/CD would’ve taken me (or a DevOps engineer) hours. You had to:
- Write YAML files or bash scripts by hand
- Debug errors at every touchpoint
- Stitch together infrastructure by trial and error
Now, with GenAI, I can just prompt it:
“Deploy this app from local to ECS with CI/CD.”
It asks a few follow-up questions, then:
- Generates the infrastructure-as-code files
- Writes the pipeline
- Handles deployment
- Troubleshoots along the way
What used to take days now takes 10 minutes. This is the new normal.
So, as a manager, I have to ask: do I still need a DevOps engineer for that task? Maybe not. But that doesn’t mean the role is obsolete — it’s evolving. The next version of this job might be called AiOps Engineer — someone who manages multiple cloud platforms, automates workflows, and keeps the AI running safely, securely, and efficiently.
AI Isn’t Just Replacing Jobs — It’s Creating New Ones
Here’s the part people forget in these conversations: AI is changing what a “job” even means.
Yes, some roles are shrinking or going away. But new ones are emerging — jobs we couldn’t have imagined just a few years ago:
- Prompt engineers
- AI workflow designers
- LLM systems integrators
- Creative technologists
- Solo founders launching AI-native tools
What matters now isn’t just whether you can do the work, but whether you can partner with AI to amplify it. GenAI gives you the “how” and sometimes the “why” — but you still need to know what to do with that output.
For example, it can help you design a DevOps pipeline, but you still need domain knowledge to deploy it securely. It can help generate product copy, but you still need to understand your audience and strategy.
The Future of Work Is Closer Than You Think
Between 2023 and 2025, we’ve already seen AI go from a decent junior coder to a competent mid-level engineer. And that trajectory hasn’t slowed down. In some industries — like behavioral health IT, where I currently work — GenAI is already delivering output at a senior level.
The bottom line? You need to reskill now.
- If you’re a manager, start learning what these tools can do — not just conceptually, but hands-on.
- If you’re a developer, understand that coding is becoming a baseline, not a superpower. Your edge will come from what you can build with AI.
- If you’re an entrepreneur, this is your moment. The tools are here, the gatekeepers are fewer, and the cost of building something new has never been lower (if you use off-the-shelf models).
The way we work is changing — not in 2030, but now.
Let’s stop asking, “Will AI take my job?”
And start asking, “How can I take control of where AI takes me?”
More people I know are walking away from traditional 9–5 jobs and launching businesses powered by GenAI. Instead of working for a company, they’re using AI to become the company. Startups today are leaner, faster, and in many cases, more effective because they don’t need bloated teams — they need a few highly skilled people who know how to harness AI tools.
The Talent Companies Need Now
To compete in this environment, companies must evolve. That means:
Reducing bloated team structures
Hiring talent that knows how to work with GenAI
Redesigning workflows around automation and intelligent tooling
The companies that thrive will be the ones that slim down operational inefficiencies and double down on capability. If you’re not streamlining with AI, someone else is — and they’re doing it faster and cheaper.
Let’s take a concrete example: DevOps.
A few years ago, deploying an app to a remote server with CI/CD would’ve taken me (or a DevOps engineer) hours. You had to:
Write YAML files or bash scripts by hand
Debug errors at every touchpoint
Stitch together infrastructure by trial and error
Now, with GenAI, I can just prompt it:
“Deploy this app from local to ECS with CI/CD.”
It asks a few follow-up questions, then:
Generates the infrastructure-as-code files
Writes the pipeline
Handles deployment
Troubleshoots along the way
What used to take days now takes 10 minutes. This is the new normal.
So, as a manager, I have to ask: do I still need a DevOps engineer for that task? Maybe not. But that doesn’t mean the role is obsolete — it’s evolving. The next version of this job might be called AiOps Engineer — someone who manages multiple cloud platforms, automates workflows, and keeps the AI running safely, securely, and efficiently.
AI Isn’t Just Replacing Jobs — It’s Creating New Ones
Here’s the part people forget in these conversations: AI is changing what a “job” even means.
Yes, some roles are shrinking or going away. But new ones are emerging — jobs we couldn’t have imagined just a few years ago:
Prompt engineers
AI workflow designers
LLM systems integrators
Creative technologists
Solo founders launching AI-native tools
What matters now isn’t just whether you can do the work, but whether you can partner with AI to amplify it. GenAI gives you the “how” and sometimes the “why” — but you still need to know what to do with that output.
For example, it can help you design a DevOps pipeline, but you still need domain knowledge to deploy it securely. It can help generate product copy, but you still need to understand your audience and strategy.
The Future of Work Is Closer Than You Think
Between 2023 and 2025, we’ve already seen AI go from a decent junior coder to a competent mid-level engineer. And that trajectory hasn’t slowed down. In some industries — like behavioral health IT, where I currently work — GenAI is already delivering output at a senior level.
The bottom line? You need to reskill now.
If you’re a manager, start learning what these tools can do — not just conceptually, but hands-on.
If you’re a developer, understand that coding is becoming a baseline, not a superpower. Your edge will come from what you can build with AI.
If you’re an entrepreneur, this is your moment. The tools are here, the gatekeepers are fewer, and the cost of building something new has never been lower (if you use off-the-shelf models).
The way we work is changing — not in 2030, but now.
Let’s stop asking, “Will AI take my job?”
And start asking, “How can I take control of where AI takes me?”