The State of Software Development in 2025
49k+ engineers have shared their preferred technologies, how they use AI, work environment and many more interesting things!
This newsletter is sponsored by DevStats.
Shipping late? DevStats shows you why.
Still pretending your delivery issues are a mystery? They’re not. You’re just not looking in the right place.
DevStats gives engineering leaders brutal clarity on where delivery breaks down, so you can fix the process instead of pointing fingers.
✅ Track DORA and flow metrics like a grown-up
✅ Spot stuck work, burnout risks, and aging issues
✅ Cut cycle time without cutting corners
✅ Ship faster. With fewer surprises.
More AI tools won’t fix your delivery. More Clarity will.
Thanks to DevStats for sponsoring this newsletter. Let’s get back to this week’s thought.
Intro
84% of engineers use or plan to use AI tools (up from 76% in 2024).
This has been an interesting insight from the Stack Overflow 2025 Developer Survey that I have closely looked at.
Some other interesting insights:
Trust in AI accuracy worsened -> 46% of engineers now distrust AI outputs (versus 31% in 2024)
Experienced engineers are the most skeptical -> only ~2.5% highly trust AI, and 20.7% highly distrust it (versus 8.3% in 2024)
AI-generated code lacks context or project-specific nuance → 45% of engineers reported that (versus 39% in 2024)
I have reviewed both the 2024 and 2025 Stack Overflow Developer Surveys in detail, and I am sharing my thoughts on the most interesting parts in this article.
Let’s get straight into it.
Over 49k respondents from 177 countries
The first interesting insight was the deviation from the number of people responding to the survey in 2024 vs 2025.
In 2024, there were 65k+ and in 2025, 49k+, which is interesting, because we have been speculating a decrease in usage of Stack Overflow for a while and this is another sign that it’s the case.
I posted this post on Friday and I am curious how many people are actually still using Stack Overflow. I personally hardly used it this year. If you do, make sure to let me know in the comments.
1. Profile of Respondents
1.1 Coding
76% are professional engineers, while 15% are adjacent or upcoming engineers. This is important info, so we know how we can interpret the results of the survey.
1.2 Country
No surprise here, the United States being the most prominent, it also went up by 1.5% from 2024, while the UK, France and Germany decreased a bit.
1.3 Age
60% of engineers are between the ages of 25 and 44. And also interesting insight is the decrease in 18-24 by 3~% and increase of 55-65 by 1.3% (from 2024 to 2025).
My gut feeling is that engineers in the age group of 18-24 don’t use Stack Overflow as much anymore and are primarily using AI when they are stuck.
1.4 Education
The interesting insight is an increase (from 2024 to 2025) of Bachelor’s degrees by 1.1% and also 0.6% of Master’s degrees. Seems like finishing traditional education is still growing, but I’ll be curious to see if that’s going to be the case in upcoming years, or we’ll see more self-taught engineers.
1.5 Industry
Software Development being prevalent (no surprise) and then Internet, Telecomm + it’s great to see Fintech high up there as well.
You can check the full profile of the respondents here.
2. Technology
1.1 Programming, scripting, and markup languages
The most interesting is the increase of Python (up by 7% from 2024) and yes, no doubt the reason is the increase of popularity of AI, as Python is the go-to for that and also for Data Science, Data Engineering and a lot of people use it as a go-to for Backend development as well.
SQL is continuing to gain popularity (increasing 7.5% from 2024). And of course, JavaScript, together with TypeScript, are very popular languages.
So, based on this data and also in my opinion, the most leveraged programming languages to learn in 2025 are JavaScript/TypeScript for frontend development and then Python for backend/AI, ML/Data related and of course, SQL is an all-around important language to learn.
Full data available here.
With that, you have a great all-around knowledge that you can use in so many different situations.
Honorable mentions are Rust (up by 2% from 2024) and Go (up by 3%) gaining more traction, PHP gaining a bit more traction, surprisingly (up by 0.7% from 2024) and Bash/Shell (up by 15% from 2024) and PowerShell (up by 10% from 2024) are all gaining a lot more importance.
I believe such a big upside for Bash/Shell is an increase in demand for automation.
On the downside, Java is down by 1%, which I believe is just because Python is becoming more popular and then Kotlin is down by 1% and that’s in my opinion because cross-platform development is gaining more traction over the years.
1.2 Databases
No surprise that PostgreSQL is up by 7% from 2024 and is the most popular database to use. You can use it in combination with so many things. And yes, that would be my recommendation if you don’t know which particular DB to pick to learn.
Also, Redis is increasingly becoming more and more popular (up by 8% from 2024), which is also not a surprise to me as it’s very easy to use and it can provide a lot of performance upside quickly.
Full data available here.
The interesting data is the increase of Supabase by 2% from 2024 and yes, it’s what most vibe coding tools pick as a default database. Some also offer the integration with it before you start/scaffold your project.
1.3 Web frameworks and technologies
Node.js (up 8%) and React (up 5%) are continuously gaining popularity, which is no surprise; both are a good combination for a web app and you can use JS/TS for both.
Interesting is the increase of jQuery by 2% and it's amazing how popular it still is. I’ve used it substantially for building websites and I can see how it still helps if you don’t wish to use a full-blown framework and just want to do some DOM manipulations.
It’s a lot easier than doing it with plain JS (less typing) → $("#element")
, whereas with plain JS, you need to do document.getElementById("element").
Full data available here.
And then FastAPI is up by 5% from 2024, which definitely makes sense as Python is gaining traction. FastAPI has now also become the most popular Python framework, followed by Flask.
What amuses me from the results is also the increase of AngularJS by 0.4%, which is ironic as Angular 2 (now Angular) was released in 2016 and even though it’s been out for 9 years, AngularJS is still used by a lot of companies.
AngularJS is no longer supported as of January 2022, therefore, there are no new releases and no new security patches. Which tells me that if you are an Angular dev, you may find a good niche to be good at migrating AngularJS apps to Angular (I know it’s hard though → from personal experience :))
1.4 Dev IDEs
Visual Studio Code usage is up by 2% and is the most popular IDE by far. The increase is a surprise to me as I have personally switched to Cursor and I know a lot of devs doing that as well.
Also, a very interesting insight is that Notepad++ is quite popular with 27.1%, which is quite interesting as I personally don’t know anyone using it for coding :)
Full data available here.
We can see the rise of Cursor, Claude Code and also Windsurf is in the mix as well. I am assuming that in 2026 the percentage is going to increase in all 3 of them. We’ll see about VS Code → I am curious to see the results in 2026.
3. AI
3.1 Large language models
OpenAI GPT, in the first place, doesn’t surprise me as ChatGPT is the most popular AI product. Following Claude Sonnet → I personally think it’s great for coding. And then Gemini in the third place.
Meta’s Llama is falling behind with only 17.8%, which definitely makes sense as to why Meta is looking to get top-tier talent and pay up to $100M total compensation.
3.2 AI tools in the development process
We can see that 84% of engineers use or plan to use AI tools (up from 76% in 2024). Not surprisingly, but still, I know there are a lot of companies out there that have a fear regarding how secure their data is when using AI.
An interesting insight is that in 2024 -> 13.8% mentioned they "plan to use". In 2025, only 5.3% mentioned. I think a lot of these have converted in 2024/2025 to using, so that’s why the number is lower.
3.2 AI tool sentiment
We can see an overall decrease in a favorable outlook on AI. This is what I have noticed with engineering leaders already, where 51% of Engineering Leaders believe AI is impacting the industry negatively.
Overall positive sentiment decreased from 70% (2024) to 60% (2025). We saw an increase of Unfavorable by 5% and Very unfavorable by 8%.
This makes sense, as I have already noted down that maintaining motivation in teams is becoming a bigger challenge in 2025. And the reason is unrealistic expectations regarding AI and also enforcing the usage due to FOMO by company leaders.
To help you manage unrealistic expectations regarding AI, this guide will help you with that.
3.3 Accuracy of AI tools
The interesting insight, when I dug deeper into this data, is that experienced engineers are the most skeptical -> only 2.5% highly trust AI, and 20.7% highly distrust it (versus 8.3% in 2024). You can check out the data here.
So, overall, we can see that the trust is becoming a bigger issue in 2025 (up 12% from 2024).
3.4 AI tools' ability to handle complex tasks
The more complex a task is, and also the more complex the codebase is, the more overall time it takes for the AI agent to handle it and also the token usage can skyrocket.
So, my opinion on this is that unless you are able to dissect a complex task into clear, smaller ones and give clear instructions, it can be pretty problematic and also expensive.
3.5 AI in the development workflow
Searching for answers is definitely a good case (that’s also the reason why Stack Overflow has become a less popular way of finding answers, imo).
And then the most resistance to using AI is in "deployment and monitoring" and "committing and reviewing code". Both definitely make sense, as human engineers should be accountable for the outcome of the code, not AI.
So, blindly trusting and “accepting” a PR, because AI has reviewed it and accepted it → is definitely not the way to go.
3.6 Vibe coding
Most of the engineers don’t wish to do any vibe coding which makes a lot of sense. At the end of the day, you are accountable of the outcomes and and trusting AI to make the right decisions is not going to do well in most of the organizations.
Any engineering team that does code reviews and has multiple people working on the same project can’t sustain vibe coding practices.
3.7 Is AI a threat to your job?
Love seeing that most people believe it’s not. And that is the right way to look at it. I’ve also mentioned why I believe that the smartest companies will actually hire more engineers in the age of AI here.
4. Work
4.1 Employment status
69.8% employed, which is 0.8% more than in 2024, which is great to see.
It’s also great to see that most of the people who have filled out the survey are either working as engineers or are students.
4.2 Work environment
The majority of people work in a hybrid or remote/very flexible environment and only 17.9% are working in an office daily (2.1% down from 2024).
Even though there have been attempts from companies to remove hybrid/remote work in 2025, based on the data, we can see that there are fewer on-site only roles in 2025 in comparison to 2024.
4.2 Individual Contributor or People Manager
The interesting insight here is that the number of people managers is up by 1.6% in comparison to 2024. Which means more managers in 2025.
We saw this trend here, where 22% of engineering leaders have reported the increase of managerial roles, while 28% reported a decrease.
Last words
I've really enjoyed diving into the data, comparing it to last year's results and sharing my observations and insights.
I'd love to hear if this kind of article resonates with you. Let me know!
Liked this article? Make sure to 💙 click the like button.
Feedback or addition? Make sure to 💬 comment.
Know someone that would find this helpful? Make sure to 🔁 share this post.
Whenever you are ready, here is how I can help you further
Join the Cohort course Senior Engineer to Lead: Grow and thrive in the role here.
Interested in sponsoring this newsletter? Check the sponsorship options here.
Take a look at the cool swag in the Engineering Leadership Store here.
Want to work with me? You can see all the options here.
Get in touch
You can find me on LinkedIn, X, YouTube, Bluesky, Instagram or Threads.
If you wish to make a request on particular topic you would like to read, you can send me an email to info@gregorojstersek.com.
This newsletter is funded by paid subscriptions from readers like yourself.
If you aren’t already, consider becoming a paid subscriber to receive the full experience!
You are more than welcome to find whatever interests you here and try it out in your particular case. Let me know how it went! Topics are normally about all things engineering related, leadership, management, developing scalable products, building teams etc.