AI Will Write 90% of Code (And Why That’s a Good Thing)

Recently, Anthropic’s CEO made a bold prediction that, in the next 3 to 6 months, AI will be writing 90% of all code.

And not only that, but he claimed that in just 12 months, AI would be writing essentially all of the code.

To be honest, I’m not surprised.

I’ve been using a lot of AI tools myself, and it’s clear that AI-generated code will soon become the norm.

I can’t say whether it will happen in the next few months or in a couple of years, but it will happen.

And you know what? That’s actually a good thing — not just for programmers but for everyone.

But why?

Because AI will eliminate the manual coding process, speed up the software development process, and save countless hours wasted on boring, repetitive tasks.

That means programmers can finally focus on what actually matters. And that’s what we will be talking about in this post.

Let’s start.

The End of “Coding” as We Know It

Let’s be honest — the process of traditional programming is clunky, complex, and time-consuming.

Say I want to build a game using Go. Even today, it’s not that simple.

First, I have to plan everything — come up with ideas, design the system, and map out the logic.

Then comes the actual coding, testing, deployment, and then the maintenance.

And before I write any code, I need to read the documentation, learn Go’s syntax, and understand its main concepts.

And then, once I start coding, errors will pop up. What’s the next step? Googling solutions — again and again. It’s a loop we programmers are stuck in.

And the worst part?

We’re so deep into writing and debugging that we rarely step back to focus on big-picture problem-solving or high-level design thinking.

But that’s about to change.

How AI Is Changing the Way We Code

For most of us, building a website or app simply means writing code. And if you get stuck, you have Stack Overflow to solve the issues.

This process is almost the same in every company, even in startups.

But today, AI is changing it completely.

You can easily generate the boilerplate, solve errors with just prompts, automate the code review and testing process, and even understand complex logic and concepts in simple English.

Let me be more specific to give you an idea.

• First off, you can use LLMs like ChatGPT, Claude, and more that can easily help you build your dream app within minutes.
• Then we have AI-powered tools like Cursor AI, Replit, and Lovable that make the process much easier.
• And then there are AI-powered tools like CodeRabbit AI that can make the code review process simply automated.
• On top of that, we can see the launch of new AI agents, which can further simplify the complete process by automatically creating what you need.

What Happens When AI Writes 90% of the Code

Well, I can’t make a bold claim that in the next 3 to 6 months, AI will write 90% of the code.

And I don’t actually believe it will be that fast.

But sure, it could be possible within a year or two, and that’s what will lower the barrier to building software.

Let me break this down further.

• Software development will be faster — Projects that used to take months will be built in weeks. Companies will ship products at lightning speed.
• More people will build software — The barrier to entry will drop. Non-technical entrepreneurs, marketers, and business owners will be able to build MVPs without hiring expensive dev teams.
• Creativity will skyrocket — With AI handling the tedious work, developers can focus on strategy, innovation, and high-level problem-solving.
• Demand for great developers will rise — I think AI will replace average programmers. But top developers, those who understand systems, logic, and business needs, will be more valuable than ever.

And guess what, programmers will always be needed since:

• AI may write complex logic in the future that makes your app unique, but programmers will still be needed to verify the code.
• AI will not be smart enough to make high-level architectural decisions.
• And surely, AI won’t understand business needs or customer behavior as programmers (more precisely, humans) do.

In short, AI will simply handle the tedious coding process, freeing you to focus on what really matters.

AI Taking Over Coding Is Actually Good for Developers

Now, the main question that may come to your mind is why this is good for developers.

Well, think about it.

Who would have imagined that we’d be able to build websites and apps in seconds? Yet, here we are.

And then, we have AI-powered tools like CodeRabbit AI that can automatically give you summaries, code reviews, walkthroughs, and more — right inside your pull request.

Just imagine how much time AI will save and how insanely productive it will make you.

But it doesn’t just make you more productive. AI can help in countless ways, such as:

• Less tedious or boring work for you — Since AI can handle writing code, building apps, and even reviewing the code, you don’t need to do the same boring work or get stuck in the never-ending loop.
• You can learn faster — We all know that AI can help you learn faster. You just need to ask what you want to learn, and it can provide you with a roadmap, the best resources, and even explain complicated concepts in simple terms.
• You can build bigger projects with smaller teams — A solo developer can now build what once took an entire team. And that’s what makes it easy for anyone to build a business and grow to millions.
• You can focus on real value — Since AI can complete most of the tedious work, you don’t need to spend your crucial time doing the same. Instead, you can focus on solving real business problems and making more.
• And more.

Here’s How to Adapt and Win

Now, I know most of you are worried about this recent AI development.

And at the same time, you’re probably thinking: “How can I actually benefit and grow from this AI revolution?”

Because whether we like it or not, AI is here to stay and will further grow, and those who learn to use it smartly will have a huge advantage.

Well, no one is entirely sure about what is going to happen.

But here’s what you can do:

• You motive should be to become a problem-solver. AI can handle syntax, and generate code, and you know programming to review the code. So your job is to focus on system architecture, creative problem-solving, and real-world application.
• Learn about prompting. I mean, get very good at prompting — and it’s easy. You just need to describe what you want, be specific, and learn a few prompting techniques.
• Master AI development workflows & learn about AI. Learn how you can seamlessly combine AI-generated code with your own, automate the process, and save your time.
• Build faster, ship more. Your focus should be on using AI to handle the tedious work, so you can deliver projects at lightning speed. I believe that people will focus on speed in the near future.
• Focus on AI tools that get your work done. Don’t go into the AI hype and use tons of AI tools when not in use.
• If you are not a programmer then also you need to learn traditional coding. Yes, you heard that right. I think everyone should learn the basics so they can easily write specific prompts, and review the code generated by AI.

Hope you like it.

That’s it — thanks.