Was 2024 a good year for software engineers?
At the beginning of 2024, I attempted to make some predictions on how the year will turn out for software engineers. The year is now almost over. Let’s see how many of these predictions came true.
If you don’t want to read the original article, the predictions I made were as follows:
Tech companies will start hiring again
Rust will continue its growth
There will be an ever-increasing demand for custom ML/AI tools
Developers who use AI tools will outcompete developers who don’t
The salaries of CRUD developers will stagnate
There will be more demand for tools that check for AI plagiarism
TypeScript + React will be the de-facto standard way of building front-end
Front-end developers will decrease their reliance on manual testing
We will see more and more broken web pages, even on reputable websites
There will be a surge in popularity of the blockchain technology towards the end of the year
So, let’s see if any of these predictions came true.
1. Have the tech companies started hiring again?
2023 was one of the worst years for layoffs in tech. It’s been estimated that over 262,000 employees have been laid off across more than 1,180 companies. There were macroeconomic reasons for it, such as historically high inflation rates across the developed countries, high base interest rates, etc.
However, these issues have started to subside towards the end of 2023 and I predicted that the tech companies will start hiring again. Has this happened?
Well, it’s been inconclusive. Layoffs have definitely subsided. In 2024, there were 150,000 layoffs across 525 companies, which was almost half of what we had in 2023. Moreover, some of these layoffs were hidden, such as the decision of Amazon and Tesla to force their employees back into the office as a strategy to make many of them quit.
But have the companies started hiring again? Well, this depends on the region, industry, and technology stack. It was hit-and-miss. Personally, I got a job at Microsoft in 2024 and many of my friends who have been unemployed for several months got jobs too. I have also started receiving noticeably more messages from recruiters. But I have also seen people who have been struggling to find a job.
So, I would say that this prediction partially came true. We aren’t out of the woods yet, but things are moving in the right direction by the look of it.
2. What happened to Rust?
At the beginning of 2023, everyone seemed to have been talking about Rust programming language. It was pitched as the language that would replace C and C++. A lot of programmers started experimenting with it. It was going to go big. So did it go big?
Well, yes. Various indexes, such as TIOBE, clearly show its growth in popularity:
Source: https://www.tiobe.com/tiobe-index/rust/
Now, even those developers who work on Linux kernel are suggesting that it should be gradually replaced with Rust! Using anything other than C or C++ in the Linux kernel was traditionally a blasphemous idea. Completely unthinkable. But now there are conversations about it happening in the open.
And there is no surprise there. Rust is a language that is capable of doing everything C and C++ are capable of but is much safer to use. And as the infamous CrowdStrike incident taught us, doing unsafe operations with memory can lead to very dire consequences.
3. What about the demand for ML/AI tools?
I think it goes without saying that 2024 saw the demand for LM/AI tools skyrocket. This is especially applicable to large language models (LLMs).
If you’ve been following the tech news, there was a feeling that a new “revolutionary” LLM was being released on a daily basis. Some names were completely forgotten by now, while others are still with us.
As well as these, the LLMs have started being adopted more frequently, even by the most conservative skeptics who have previously refused to touch ChatGPT. So yes, this prediction definitely came true.
4. Did developers who use AI tools start to outcompete those who don’t?
I am on the fence on this one. Yes, AI tools boost the developer’s productivity. There’s no doubt about that. However, some of the best developers I know are still skeptical about AI and still don’t use any of these tools. Yet, they still have their jobs and they still produce great software.
However, the trend that I couldn’t help but notice is that the market became pretty tough for those developers (mainly junior developers) who lacked the ability to do anything other than write code according to the requirements. This is because any good LLM, such as the latest model of ChatGPT can do exactly that: output code according to the requirements you give it.
LLMs are effective at eliminating the grunt work of software development, of which coding is a major part. Therefore, if all you are good at is the said grunt work, you will have a tough time competing against a machine.
5. What happened to the salaries of CRUD developers?
The vast majority of developers build the so-called CRUD applications, which are the applications that Create, Read, Update, and Delete data. Some of these applications require implementations of complex bespoke algorithms, but many others are relatively simple.
I predicted that the salaries of those developers who build relatively simple CRUD apps would stagnate. This is because, with the above-mentioned AI tools, these applications became significantly easier to build.
Did this prediction come true? Unfortunately, yes. The vast majority of the software engineering vacancies I’ve been seeing weren’t offering great salaries and this even applied to some top companies. I’ve even seen some companies I interviewed with back in 2022 listing a significantly lower salary for the same position!
This is another indicator that being a competent coder isn’t enough anymore. The era of AI tools requires some other skills to stand out.
6. Did the AI plagiarism tools become more popular?
Well, the short answer is yes. 2024 saw a skyrocketing growth of the AI plagiarism industry.
Some of these tools are designed to detect plagiarism. Some are designed to detect AI-generated text. Some are designed to detect the tell-tale signs of AI-generated images and videos.
However, the industry is still immature and there are many of these tools are still not perfect. For example, there was a whole series of memes laughing at the fact that these tools tend to think that Macbeth and other English classics were generated by the AI.
For the time being, the only way to guarantee that your text won’t be flagged by one of those AI-detection tools is to write the said text really badly.
7. Did react become the de facto standard for the front end?
Well, the answer is both yes and no. Yes, because it’s still, by far, the most used front-end technology compared to others. No, because, while it is dominating, its popularity is actually going down, not up, as the following graph shows:
Source: https://trends.stackoverflow.co/?tags=reactjs,vue.js,angular,svelte,angularjs,vuejs3
But it doesn’t look like there’s another library or framework that is going up in popularity. I can only guess why that is, but it could be that more and more developers are starting to use WebAssembly for front-end development and related technologies, such as Blazor and UNO Platform, so the competition comes from outside of the traditional JavaScript ecosystem.
8. Did the UI developers decrease their reliance on manual testing?
Well, I couldn’t find hard data on this, so I can only speak from my subjective experience. And based on my experience, this prediction did come true.
I interact with UI developers on a regular basis and it’s been a while since I’ve met a UI developer who didn’t know how to test the UI by using an automation-testing framework.
I’ve also seen more and more UI developers talking about ditching manual testing entirely and relying solely on automated tools for UI testing. I’ve even seen this being done on a project I’ve been involved with. Therefore, I would conclude that there is a noticeable trend in the UI development niche to minimize the reliance on manual ytesting.
9. Did we see more and more broken web pages?
This is a personal pet peeve of mine and I don’t understand how did this suddenly become an acceptable norm.
You can probably all relate to the following situation. You open a web page and there is a button on it that you want to click. However, you end up clicking on something else. Some other button just appeared out of nowhere and you clicked on it instead because the original button moved.
Well, I don’t know if the prevalence of this problem became worse in 2024, but the problem certainly didn’t go anywhere. There are still plently of websites where this happens. So, I would say that this prediction was at least partially true.
10. Did the blockchain technology surge in popularity?
Finally, I predicted that, because Bitcoin was due to surge in value in 2024, as it does every four years, the blockchain technology would become popular once again. This is the thing that I definitely got wrong.
Yes, Bitcoin did surge in value. At the time of writing, it is still at the historic high of almost $100k per coin, although it has now gone down somewhat. This surge in value is caused by a natural event that is inbuilt into Bitcoin’s blockchain algorithm, known as halving. When it happens, the reward for mining new Bitcoins gets reduced by half. There is now much less incentive to mine it and the creation of new coins slows down. This is what drives the value of the coins up.
However, the part that I didn’t get right was that this would lead to an increased interest in the blockchain technology. This happened during pretty much every halving before. Howeevr, this time, it didn’t. It’s hard to say why, but my guess is that the NFTs that were severely overhyped shortly after the previous halving of 2020 left a bitter taste in people’s mouth that still didn’t go away.
Wrapping up
So, let’s see how many predictions did I get right. I’ll use the following scoring criteria:
Give 10% to a prediction that definitely came true
Give 5% to a prediction that came true partially or that I’m not 100% sure about
Give 0% to a prediction that definitely didn’t come true
So, the scores are:
5 + 10 + 10 + 5 + 10 + 10 + 5 + 5 + 5 + 0 = 65
So, my predictions appear to be 65% accurate. Not amazing, but definitely better than the random guess.
Would you agree with my assessment of the tech trends? And if you were to make predictions for 2025, what would those be? I would love to read your opinion in the comments.
P.S. If you want me to help you improve your software development skills, you can check out my courses and my books. You can also book me for one-on-one mentorship.