2025 Is The Best Time to Learn Coding (Even with AI)

Hey there, I’m Deric. A few years ago I was an accounting grad scratching my head, wondering where I belonged.

Now I’ve become a builder and entrepreneur – coding websites and AI tools – and I’ve taught over 100 people to code. 

I even started a coding school that promises “get a job or it’s 100% free” – and we’ve seen 100+ success stories of students landing tech jobs across Southeast Asia.

Why does this matter to you? Because if someone like me can make that pivot, so can you.

And believe me, 2025 is actually the best time to dive into coding, AI and tech – not despite AI, because of it.

The Problem: Career Stagnation, Economic Shifts, Rising AI

Let’s face it – things feel shaky right now. Maybe you’re 25–30, early in your career or thinking of a switch, but you feel stuck.

Jobs that once promised stability are changing. Businesses are automating tasks, AI buzz grows every day, and you hear headlines about tech layoffs and economic uncertainty.

It’s natural to worry, “Is my job even safe? Should I be learning something else?”.

Indeed, by 2025 the world of work looks very different. Automation and artificial intelligence are reshaping industries; repetitive tasks and basic data crunching can now be handled by machines.

Many roles are shifting toward higher-skill, human-centered tasks. That sounds intimidating – but it actually points to a key truth: the skills you develop now are crucial.

Don’t fall for the myth that you’ve missed the train. I hear it all the time: “AI will write all the code, why learn?”

In fact, that’s a common myth – many non-technical voices claim developers are “automating themselves out of jobs”.

But leading tech thinkers say the opposite. As AI tools get better, it becomes an even better time to learn coding .

Even though 2023–24 saw some tech layoffs and salary cooling, remember: the world got used to crazy growth in 2021/22. (Engineers enjoyed 30% raises back then!).

Sure, we aren’t in hyperdrive right now, and yes, global economies have their ups and downs.

But smart observers still see opportunity. For example, companies in Singapore and Southeast Asia are still innovating – adopting AI, expanding fintech, cybersecurity and digital services.

These industries need skilled people. The problem right now is a talent gap, not a talent glut.

So the real problem isn’t too much technology or not enough jobs. It’s a mismatch: many jobs want digital and coding skills, but too few people have them.

This is exactly why coding and AI-savvy people will be in demand. In short: the landscape is shifting, and learning to code is a way to move from being a passenger to being the driver of your career.

First Principles: Why Coding is Valuable in 2025

Let’s break it down from first principles. What is coding? At its core, coding is simply telling a computer exactly what to do, step by step.

Think of it as learning a new language – the language of machines. Once you speak that language, you can create apps, automate tasks, analyze data – anything software can do.

Why does that matter in 2025? Because the world runs on software. Every company needs websites, apps, analytics dashboards, automated processes.

All of this is made by code. From ordering food delivery to running factories and financial systems, software is the invisible engine.

If you know how to code, you have the keys to that engine.From a first-principles view, a fundamental truth is: Humans have ideas and problems; computers do work on instructions.

Coding bridges that gap. You think of a solution or product (an idea), you break it into logical steps, and you write code to execute it.

No programming? Then you’re stuck outsourcing or hoping someone else solves it. Knowing code means you don’t have to wait. You can actually build the solution yourself. That’s why experts argue coding is like a new kind of literacy.

As one tech leader put it,“Just as literacy went from an esoteric skill among elites to the global standard for education, coding is becoming essential for everyone — if we are to understand, maintain, and wield the AI that will define the future”.

In other words, to shape this future where AI is everywhere, we all need basic coding fluency.

At a more granular level: coding teaches you computational thinking – breaking problems into chunks, solving them step by step.

That skill applies even if you don’t end up writing thousands of lines of code. In fact, even front-line workers and managers benefit from automating and optimizing parts of their jobs.

And for any tech or data role, you’ll need these fundamentals. It’s not about memorizing syntax or frameworks; it’s about understanding how to solve problems in logical, systematic ways. In practical terms, 2025 tech roles are still clamoring for people who can code.

The World Economic Forum’s 2025 Future of Jobs report highlights that technology-related roles – from AI specialists to software developers – are among the fastest-growing jobs.

One research forecast even projected 540,000 new software engineering jobs in 2025 (a lot of them tied to AI and automation tasks). These are not jobs for robots – they’re roles to build and manage robots.

So, from first principles: if you want skills that remain valuable when machines are doing more, learn the skills that talk to those machines.

Coding is the ultimate multipurpose skill – it’s durable, flexible, and it compounds. Each concept you learn makes the next one easier, and soon you’re not just coding one project, you’re designing systems, working with data, maybe even teaching others how to code.

Real Stories: My Journey and What Worked

I know some of you might be thinking, “This sounds great, but I’m not a techie. Where do I even start?” Let me share how I did it – and some real-world lessons I learned along the way.

Hustle Start: My own journey began far away from Silicon Valley. In college, I actually studied accounting and finance.

But early on, I discovered entrepreneurship by flipping used graphing calculators. I would buy them cheaply and resell to students.

It sounds small, but that side-hustle taught me supply and demand, profit margin, and market fit.

I was earning more than a typical salary at 17, just by solving a real problem (people needed calculators affordably). That was my first lesson: you can create value from scratch, even as a student.

The Pivot to Tech: After college, I worked in a corporate finance role – but I felt out of place.

The startup world I’d seen at university fascinated me. I realized that tech was at the heart of the biggest companies (think Apple, Google, Amazon).

I wanted to build rather than crunch spreadsheets. So in 2019 I quit my job to learn coding full-time .

That transition wasn’t easy. I couldn’t afford a fancy bootcamp or second degree, so I self-taught using free resources.

Within just six months of intense learning, something interesting happened: my friends who did have IT degrees were coming to me for help.

I was tutoring them on technical interviews and even freelancing on software projects . (Lesson: pure degree doesn’t guarantee skill – what matters is what you do with what you learn.)

To push myself, in 2020 I started what I called The Hacker Collective – basically weekly meetup groups where all of us would learn by doing.

I gave myself a challenge: build one project every month. It forced me to turn theory into reality.

I remember working on little web apps, scripts, anything. We used free resources like Codecademy together, but more importantly we held each other accountable.

That community effort was eye-opening. A lot of people joined – soon dozens of students were showing up.

Many dropped off (as always happens) because coding is tough to sustain alone . But the ones who stuck it out started building real things.

I even started teaching some of the sessions. Teaching others forced me to master the gaps in my own knowledge.

By explaining a concept to someone else, I understood it better myself. That’s a trick I still use: if I get stuck, I say “Imagine I have to teach this to someone who knows nothing.” It clarifies the logic.

Seeing Results: From this effort, something clear emerged: a practical, project-based approach works.

We had students from all walks of life: some had CS degrees, many did not. One time, a colleague named Ming (who was way ahead of me) demoed a project he built.

It lit a fire in me – it proved there was no secret degree needed, just start building.

Over time, I tried many things: running paid workshops, doing freelance work to fund our learning, even taking equity in startups. (Side note: most of those startups failed, but I learned a ton.)

We tracked our students. It turned out, at least 10 people from those early cohorts landed tech jobs – some in Malaysia, some in Singapore, even some remote roles in US companies.

One really cool story: I had hired a very junior developer and coached them intensively – three months later, they snagged a job in Singapore at $4.5K/month. That was proof to me: we were giving people real skills, and it paid off for them.

These experiences showed me that learn-by-doing and peer support is powerful. People who came from finance, marketing, even construction backgrounds were suddenly coding – because they had proof of work.

They had built projects in their portfolio and could talk about them in interviews . In fact, that became our philosophy: you don’t need a CS degree or 10 years of experience. What you need is real skills and examples of using them.

Fast forward to 2022, Ming and I formalized all these experiments into Sigma School, a 3-month coding bootcamp. We tied our incentives to our students’ outcomes: get a job or it’s free.

We started small with just a community course, and the response blew us away – a small marketing spend of RM3,000 turned into RM30,000 in the first month, even before our official launch. Clearly, we weren’t the only ones seeing the gap in the market.

Since then, Sigma School has grown to partner with 50+ companies in Singapore, Malaysia, Australia.

We’ve had graduates from all kinds of backgrounds – there was even a doctor in the UK who pivoted careers, and gig workers who finally broke into tech.

One of our star students got hired just 1.5 months into the program. These stories underline one thing: with the right guidance and real projects, normal people can learn to code quickly and land jobs they never thought possible.

So, from flipping calculators to launching a coding school, my own journey has shown me: small consistent efforts (one project a month, output-based learning) beat cramming theory.

Focus on solving actual problems, build a community around it, and keep pushing. That philosophy has stayed with me, and it’s why I’m passionate about helping others do the same.

Misconceptions: Why AI Doesn’t Replace Coders — It Multiplies Them

Now, let’s tackle the elephant in the room: AI. I get it – everyone’s talking about AI these days. Tools like ChatGPT or Copilot can spit out code snippets instantly.

So it’s tempting to think: “Maybe I don’t need to code anymore, AI will do it.” But here’s the straight talk: AI will never replace the need for human programmers – it will only make skilled coders more valuable.

Tech leaders and researchers keep saying this. For example, AI pioneers like Andrew Ng and Bill Gates explicitly argue now is the best time to learn coding.

Andrew Ng put it succinctly: “As these tools continue to make coding easier, this is the best time yet to learn to code, to learn the language of software, and learn to make computers do exactly what you want them to do.”

He’s saying – as AI lowers the barrier, now more people can and should become code-savvy.

Here’s why: AI can generate code, but it doesn’t understand your project goals, your company’s needs, or how to design a system from scratch.

It’s like asking a calculator to solve a problem without telling it the right numbers and context. Coders do that work.

Even the biggest AI models need humans to fine-tune, fix, and apply them. Bill Gates noted on his podcast that even if AI changes industries, “three jobs will remain the same,” and he listed coding first.

Why? Because coders will always be needed to debug, refine, and improve AI systems.

Think of it this way: AI is a tool – a very powerful assistant for writing routine code or generating ideas.

But like any tool, its output needs review and direction. A well-known quote from a Microsoft engineer, Will Kencel, sums it up: trying to replace engineers with AI “is the most over-hyped trend in 2025”.

He points out that you can automate tasks like bartending or toll booths, but not the creative, problem-solving work of software engineering. In fact, he says jobs will open up where we need people to maintain AI systems.

Data backs this up too. The World Economic Forum’s Future of Jobs 2025 report finds technology roles – software developers, AI and machine learning specialists, big data experts – to be among the fastest-growing fields.

One forecast I saw predicted that AI advancements would create about 540,000 additional software engineering jobs in 2025. So rather than shrinking, coding jobs are growing, just changing shape (more AI collaboration, more data integration).

In practice, what I see is that AI multiplies a coder’s output. When I was building Codeo.ai (a mobile app to teach coding), we used AI tools to prototype faster.

Our students in Sigma School are encouraged to use Copilot or ChatGPT as part of their workflow – but only as helpers.

They still have to understand the core logic and debug the results. It’s like having a digital apprentice: it writes rough drafts of code, but you supervise and refine it. The result? A faster feedback loop, not a replacement of skills.

Another way to see it: coding is 90% thinking, testing, and designing, and 10% typing. (This idea was mentioned by software engineers in the industry.)

AI can assist with the typing – suggesting code snippets, catching syntax errors. But it can’t replace the creative and critical parts – only you can do that.

That’s why learning how to think like a programmer (first principles, breaking down problems, writing algorithms) is more important than ever. AI tools can’t do that for you.

So if you’re worried AI will take the jobs, think again: AI is more like a partner. It automates the boring stuff, freeing you to do the interesting stuff (architecture, design, strategy).

As one expert put it, it will open up new jobs where people are needed to maintain AI, and essentially make good engineers better. Embrace AI as a force multiplier for your coding, not an opponent.

Strategic Perspective: Opportunity Landscape in SEA and Globally

Let’s zoom out and look at the bigger picture. You might be thinking, “Okay, coders are needed, but where are the jobs and opportunities?” Good question.

The reality is, the world is moving online faster than ever, and regions like Southeast Asia are right in the middle of this transformation.

In Southeast Asia (SEA), digital adoption is booming. The World Economic Forum highlights that digital skills are seen as key for companies across South-East Asia, yet significant gaps exist.

In simpler terms: companies in SEA desperately want more people who can build software, but they can’t find enough. This is exactly the supply-demand gap we talked about.

Singapore, in particular, is pushing hard into tech. Reports show Singapore’s tech sector still pays competitively – software engineers there saw salaries rise 3.3% in 2025 after a slight dip.

The city-state’s government programs and startup culture (Smart Nation initiatives, fintech boom, etc.) mean more tech companies and investments are pouring in.

In fact, NodeFlair’s 2025 report notes Singapore is “at the forefront, especially in fast-growing sectors like AI, cybersecurity, and fintech”. That means jobs.

But it’s not just Singapore. Around Southeast Asia, the service economy and tech startups are expanding (think e-commerce, digital finance, edtech).

SEA’s share of global tech startup funding is climbing – just recently SEA startups raised over $2 billion in H1 2025 alone (a new high).

These startups need developers: full-stack engineers, mobile devs, data experts, AI engineers, and more.

And with remote work now normalized, you could even work for a company in Singapore, Australia, or beyond without leaving your home country. The whole world is in play.

This global view applies outside SEA too. In the US and Europe, tech continues to hire: even if some big companies had headcount freezes, many others (including emerging ones) are desperate for coding talent – often more so in niche fields like machine learning, cloud computing, and cybersecurity.

And there’s a notable trend: many SEA startups are building products for a global market. In that scenario, they pay and recruit internationally. For you, that means learning to code can connect you to global markets and salaries that far exceed local averages.

Going back to the data: tech leaders globally see that 37% of work will be done by technology by 2030, up from 27% today (per WEF estimates).

In SEA, 92% of employers expect growth in skills like networking and cybersecurity, and 83% foresee rising soft skills needs (resilience, agility). The takeaway? If you’re good at digital stuff and flexible, you’ll be in demand in any economy.

In short, the tech opportunity landscape is huge. In emerging economies like ours, it’s still early days compared to say Silicon Valley – that means less competition among job seekers, more mentorship chances, and still enough room to catch up and grow.

I see this every day at Sigma School: our graduates are getting hired in Malaysia, Singapore, and even remotely by US and European firms. The sky’s the limit once you have the core skill.

What to Focus On: Tools, Mindset, and Outcomes

Now that we’ve convinced you why to learn coding, let’s talk how. Here are the practical things I’d focus on – delivered with that first-principles, outcome-based mindset that works:

Learn Fundamentals, Not Just Buzzwords

Start with the basics. Choose a solid programming language (like Python or JavaScript) and really understand it.

Don’t rush to frameworks. Learn how data structures and algorithms work at a conceptual level.

In 2025 and beyond, fundamentals still matter most – frameworks change, but core concepts carry over. (Think: if you know the logic, you can adapt to whatever new tools come along).

Build Real Projects (Outcome-Based Learning)

The fastest way to learn is by doing. For example, challenge yourself to build one small project every week or month.

It could be a simple to-do app, a website, or a data analysis script. The key is it solves a real problem – something you care about. This way you learn not just syntax, but problem-solving, debugging, and actual software craftsmanship.

Remember, employers care about what you’ve built, not just what courses you’ve watched. In my bootcamp, we always say: proof of work beats proof of degree. So focus on outcomes: your code, your apps, your GitHub portfolio.

Use AI as a Tool 

Embrace AI assistants like ChatGPT or Copilot to boost your productivity – they’re part of the new toolbox.

They’re great for boilerplate or brainstorming solutions, but use them wisely. Always review and understand the code they give you; it’s not a magic bullet.

As one engineer noted, AI code can be “not maintainable or reusable,” so treat it like a collaborator, not a teacher. Leveraging AI well will actually set you apart, because you’ll learn how to integrate it intelligently.

Iterate in Small Steps (Small Bets)

Don’t attempt a huge, unachievable project on day one. Break it down. If you want to build a full e-commerce site, start by making a simple product listing page.

Then add a cart, then payments, step by step. Each small success builds your confidence and skills.

This “small bets” approach – keep shipping tiny features and learning from them – compounds over time. I lived this when I scaled my calculator side business and later startups: bite-sized progress leads to big wins.

Learn with Others (Community & Feedback)

Coding can be lonely, so join a community. That could be an online forum, a local meetup, or a coding cohort like a bootcamp or a study group.

Having peers means you can ask questions, share knowledge, and stay motivated. I ran meetups and even charged a token fee to make it formal – many CS grads actually paid me to coach them!

So don’t skip the human element. You can also “build in public”: tweet about your projects, share progress on LinkedIn, blog about your bugs and fixes. It keeps you accountable and sometimes even attracts mentors or employers.

Outcome over Hours (Outcome-Based Execution)

Especially if you’re self-learning, don’t fall into the trap of just collecting course certificates. Focus on what you can deliver.

For example, set milestones like “I will build a portfolio website by X date” or “I will deploy an app to the cloud.”

Measure success by those achievements, not hours logged. Even better, find someone (a coach, mentor, or friend) to give feedback on what you build. This turns learning into a real feedback loop, much like a job would.

Stay Flexible and Keep Learning (Compounding Skills)

The tech field changes, so keep your mindset of a learner. But be strategic – you don’t have to learn everything.

First-principles thinking helps: if you know one language and one framework well, you can usually pick up another quickly by applying the same core logic.

Build related skills that compound: for example, learning JavaScript well also means you can understand many front-end and back-end frameworks (React, Node.js).

Learning how to work with Git and the command line will help no matter what language you use. These skills keep compounding – six months from now you’ll see how far you’ve come.

Career Savvy (Building In Public and Networking)

In our Sigma School career coaching, we stress portfolio and network. Even as you learn, put your projects on GitHub or a personal website so you have evidence of your skills.

Connect with professionals on LinkedIn, attend tech webinars or hackathons. Building in public (sharing your journey) not only keeps you honest but can open doors.

One of our alumni got hired because a recruiter saw her coding bootcamp projects on GitHub – she had tangible proof of her abilities.

Stay Resilient and Iterate (First-Principles Again)

Lastly, keep a cool head. Coding is challenging. You will hit bugs and frustrations. Approach each roadblock analytically: break down why your code failed, test assumptions, and try again.

Use first principles – if something breaks, isolate it, simplify the problem, fix, and then rebuild. Remember, every expert was once a beginner who spent a thousand hours debugging code they wrote poorly. It’s all part of the process.

Encouragement: A Grounded Nudge to Take Action

Look, I won’t sugarcoat it: learning to code takes effort. It’s not some get-rich-quick scheme or a gimmick. But if you’re reading this, you’re already curious, which means you have the mindset part down. You just need to start acting on it.

Think of coding as an investment in yourself. In five years, I guarantee the skills you build today will pay off.

You might end up in a higher-paying job, or maybe even start something that solves a problem you care about.

Or you might use coding to level up any career – even in marketing or operations, coding could let you automate your work or launch a side project.

It’s a compound interest game. Spend an hour a day on it, and one year from now you’ll be hundreds of hours in – enough to be comfortably proficient in something.

Remember my story: if a guy who once flipped calculators and freelanced with no tech background could do this, so can you.

I launched a business, pivoted it into a coding school, and I’m still learning (I just finished integrating AI into our full-stack course). If I can learn a new tech tool in my late 20s, surely you can too.

So here’s the nudge: pick one thing right now. Maybe it’s signing up for a free coding bootcamp trial, or opening that Codeacademy tutorial you bookmarked.

Maybe it’s just writing “Hello World” in Python after watching a 5-minute YouTube video. Tiny steps matter. Do something today. Tomorrow, do a little more – fix a bug, tweak a line of code. Stay consistent.

You might be thinking, “Yeah, but what about real life responsibilities?” I get it. We all have jobs, bills, families.

That’s why learning efficiently matters. Do the things that give the most return: hands-on practice and project-building. If you can dedicate a few hours each week, that’s enough to get you going.

And remember, you’re not alone. Thousands of people your age are in exactly your shoes, learning to code alongside AI. Join them. Communities, bootcamps, online forums are all there. You’ll find mentors (like me and my team) and peers willing to help.

2025 is a crossroads year. AI is here, but it’s empowering us. Companies are still hiring, especially for people who get tech. If you start now, six months from now you’ll have made huge strides. A year from now, you could be interviewing for your first dev job.

Learning to code isn’t just about code – it’s about thinking clearly, solving problems, and creating value. Do that consistently, and you’ll ride this wave rather than get swept under it.

So do it. Make the small bets, stack those skills, and build something real. Your future self will thank you. The best time to start was yesterday, the next best time is today.