With the rapid growth of Large Language Models (LLMs) like OpenAI’s GPT, I’ve started to question the future of our field. As a 21-year-old, I realize I don’t have a clue what I will be working on in just a couple years. Every year, AI capabilities grow, proving our old theories wrong. Just a few years back, we thought computers were very efficient in doing calculations but couldn’t understand complex things like human language. But look where we are now.
Since the 90s, more people have been using the internet, with faster speeds and more affordable devices, even in undeveloped countries. This has led to more and more parts of our lives going online, creating a new digital economy. This economy is a big money-making machine. Take Amazon, for example, which started by replacing bookstores and then moved on to most retail. Or Google, which now dominates the ads market. This shift affects all businesses, big and small. From restaurants joining delivery apps to companies managing all their documents online, everyone is moving online.
This change has made tech companies some of the most valuable in the world. Capitalism is really pushing this move online because it makes things more productive and efficient.
As a result, there’s a huge demand for jobs to help with this move online. These jobs usually fall into a few categories.
Engineers have made it easier to build common programs that businesses need, like user interfaces (web and mobile applications) and backends (the set of technologies like APIs, databases and servers that manage business logic). Most of the time, being a software engineer in a company means turning business needs into framework code. This job is relatively easy to learn and many people are learning to do it. There’s so much demand for this skill that it still pays well.
A big part of software engineering today is writing code to connect different software together. This means making sure various online services, databases, and even AI models work well together.
These are the people doing the really tough work. Hackers create things that have never been made before. They develop AI models, find new ways to use technologies, and figure out how others’ solutions work.
But in my experience, these jobs are rare. Most companies are just tweaking solutions that already exist to fit their needs. You don’t need to know all the details about computers or the internet when you’re making a simple app for scheduling gym classes. You usually just use or combine things other engineers have made.
As AI starts doing more routine tasks, there’s a chance for a teamwork kind of relationship where AI helps us do things better. This could change job roles to focus more on managing and solving creative problems. Humans are still better than AI at being creative and understanding feelings. Working together with AI could lead to more meaningful jobs, where AI handles the boring stuff, and humans handle the creative and impactful work.
AI models are starting to slowly replace the manual work of programmers. The speed of this evolution is so fast that it’s not hard to imagine that eventually, “Translators” and “Glue Coders” will be replaced by AI. Even today, AI can build user interfaces and backends, but not yet reliably reflecting complex business requirements. But it’s already better at many tasks than beginner programmers and is much faster at repetitive and less intellectually demanding tasks (like designing HTML, parsing/processing data, and quick calculations).
The main point is that we didn’t have these tools just a few years ago. Extrapolating even 3, 5, or let’s say 10 years into the future, I’m mostly sure translators and glue coders will not be as prevalent, accounting today for 90% of job positions.
Present and Future
Just like we moved online in the 90s, now we’re starting to move into an AI-driven world. Big companies and investors are putting a lot of money into AI. Looking at the most valuable companies in the world, we see:
- Microsoft, which owns a big part of OpenAI and is making a huge bet on AI, integrating it into its whole suite of software and search engine.
- Nvidia, which makes the hardware for training LLMs.
The other top companies, even if AI isn’t their main thing, are spending a lot on it, like Alphabet (Google), Meta, and Tesla.
Also, seeing portfolios of Venture Capital firms, we see that most startups getting funded are AI-related.
Seeing money shift from just moving the economy online to “let AI move things online for us,” I think ignoring AI in software engineering is like ignoring the internet in 1995.
The focus will be more and more on:
- Making AI models better (probably centralized by a few big companies like OpenAI, Meta, and Google).
- Building tools that use these models to replace human jobs (specially engineering as they are very expensive) and getting better at understanding business needs.
- Making better interfaces to interact with AI models.
Thanks to AI, some companies are now run by just one person, using AI for coding, customer support, marketing, sales, and more.
I think this change won’t happen overnight and won’t be straight-forward, but with more focus on using AI to automate computer work, more jobs will shift to linking AI with real-life businesses. This could mean less need for repetitive computer jobs, which are a big part of professional work.
Those who adapt quickly will make a lot of money. The big question is whether the pace of change will let most people adjust to the new economy and find new jobs. History says that tech advancements usually create more and better jobs than they destroy, but we’ll have to see if it’s different this time.