Python is still the best language.
In the recently released IEEE Spectrum 2025 programming language ranking, Python has once again won the most popular programming language.
**Not only is it the champion, but it is also a cliff like lead - the gap between second place Java and Python is quite significant.
What’s even more frightening is that Python is still maintaining a rapid growth rate, and the gap between first and second place is widening.
##Python has won ten consecutive championships
After winning the top spot in IEEE Spectrum for nine consecutive years, Python has once again made efforts this year and won ten consecutive titles in the comprehensive ranking of programming languages.
Moreover, this is the first time that Python has topped all three categories including comprehensive ranking, growth rate, and employment orientation, and it is also the first triple crown on the IEEE list in 12 years. **
Back in the day, Python was renowned for its simplicity and gradually replaced Java as the teaching language in universities.
Subsequently, libraries such as NumPy, SciPy, matplotlib, pandas, and others were introduced, rapidly making Python a ace in the fields of scientific research, finance, and data analysis.
During this process, network effects played a significant role.
More and more programmers are choosing Python and developing various tools based on it, contributing massive amounts of code and solutions to the open source community, gradually building an ecological moat belonging to Python.
AI further amplifies this advantage.
Compared to other programming languages, Python has richer training data, which allows large models such as Claude to perform better when using Python. Conversely, this has also attracted more AI users with zero foundation to choose Python as their preferred language.
**I never expected that Python would reap the benefits of artificial intelligence before programmers. **
Some people are happy, while others are worried. Python has become the “biological son” of AI, but JavaScript has fallen into a big mold.
This year, JavaScript has experienced the greatest fluctuations in its overall ranking, dropping directly from the top three last year to sixth place. **
The advantage of JavaScript lies in web page writing, and this peak is being submerged by the water level of Vibe Coding, with fewer and fewer people discussing JavaScript.
The throne of SQL, the king of data analysis, has also been challenged.
You should know that SQL has been the most popular skill among employers for many years and a must-have skill on programmer resumes.
But this long-standing defense line has also been breached by Python.
However, SQL is still much safer than JavaScript. Even if it is no longer dazzling, it is still a highly valuable employment skill, and the gap with Python is not significant.
This is because SQL is positioned more vertically and is the standard language for all enterprise databases.
Despite Python’s expertise in general programming, accessing databases still relies heavily on SQL.
##Behind the Ranking: The Stagnation Moment of Programming Languages
As the head effect of programming languages becomes increasingly prominent, the way programmers work is also undergoing subtle changes.
The most obvious is that the once popular ‘community culture’ is declining.
Programmers’ activity on question and answer websites has sharply declined. When encountering problems, they no longer immediately post in the community, but are accustomed to directly asking big models or relying on intelligent agents such as Cursor for automatic completion.
The data also confirms this point:
In March 2023, Stack Overflow added about 87000 new questions, but by March 2024, this number had dropped to about 58800, a decrease of 32.5% in just one year.
**By December 2024, the situation will be even more bleak, with a year-on-year decline of 40%. **
Many programmers bluntly state that the community atmosphere can sometimes be uncomfortable, and they are now more willing to consult AI when encountering problems:
The people there are really mean. They will blame you for not searching in advance, making you feel foolish, and most of the time, they just glance away and won’t help you at all.
After the emergence of LLM, things have improved a lot. No artificial intelligence would bully you just because they think you’re stupid and leave.
Not only the learning methods, but also the work logic of programmers is quietly changing.
Whether it’s experienced professionals using AI to handle tedious tasks or beginners trying to write complete web applications, the intervention of AI gradually frees programmers from their obsession with programming details.
Firstly, there are grammar details, followed by process control and functions… These contents that used to require constant effort are increasingly being handed over to AI for processing.
Soon, the details of the source code will no longer be important, and no one will fight over whether to use tabs or spaces for indentation.
**Even programmers may no longer have to worry about which language to choose, as there will only be one answer left. **
The reason why programming languages flourished in the past was because there was always a certain language that had an advantage when faced with specific tasks - you wouldn’t use R to control a washing machine, nor would you use C to run large-scale statistical analysis.
But from a technical perspective, this is not impossible, it’s just that it’s as difficult as climbing to the sky for humans, and we don’t have much energy to do this thankless task.
However, AI does not have such troubles. As long as there is enough training data, it can solve any problem using a specified language.
This means that the Matthew effect will further intensify and the diversity of programming languages will significantly decrease.
On the one hand, mainstream universal languages will experience exponential growth in positive user feedback, solidifying the pattern of ‘the strong always remain strong’.
On the other hand, non mainstream languages are further marginalized due to the lack of training data, making it difficult for AI to support them, and the threshold for programmers to get started is high.
Just as today’s developers almost no longer care about CPU instruction sets and underlying assembly languages, whether programming should be done in Python or Java in the future may become irrelevant details.
##Is AI Ending Programming Languages?
Currently, programming is undergoing its biggest transformation since the emergence of compilers in the early 1950s.
The initial mission of high-level languages like Python was actually quite simple:
-
Create an abstraction layer that eliminates the need for programmers to manually manipulate registers, memory, and instruction sets.
-
Provide error prompts when there are code errors, instead of crashing directly, to prevent programmers from “lifting a stone and hitting their own feet”.
From a computer perspective, high-level languages have never been a necessity, it’s just that humans need a more convenient way of interaction.
Unfortunately, the emergence of high-level languages has not made programming much simpler. For most people, natural language is the most ideal.
Nowadays, AI is still only helping us improve the efficiency of coding, but in the future, can we skip this non essential step and directly communicate with the compiler through Prompt? **
This mode has a drawback, which is that it makes programming a black box that cannot be explained like a large model.
However, although we can no longer directly read the source code, we can break down the program into modular functional units and ensure the reliability of the program by supervising each unit.
At that time, programmers no longer need to maintain thousands of lines of source code, as long as they adjust the prompt, they can optimize the software with one click through AI.
So in the future without source code, what is the role of programmers?
There is no definite answer, but what we can be certain of is that the underlying architecture design and algorithm selection are still core competencies. **
Programming languages may retire, but the underlying logic of code will not.
Reference link:
[1] https://spectrum.ieee.org/top-programming-languages-2025?utm _source=homepage&utm_medium=hero&utm_campaign=2025-09-25&utm_content=hero1
[2] https://hodgef.com/blog/stack-overflows-decline-how-ai-is-changing-community-engagement-forever-a1ab2/?utm _source=chatgpt.com
[3] https://news.ycombinator.com/item?id=41364798&utm _source=chatgpt.com