You want to know what’s actually wild? Python started as a bored developer’s holiday side project. Guido van Rossum, stuck at home over Christmas in the late 1980s, started building a language to keep himself occupied. That language now runs AI, powers Mars rovers, and just hit a popularity rating that no programming language has matched since Java in 2001.
A Christmas project. Ruling the world. Let that sink in.
Quick Facts
| Detail | Info |
| Created by | Guido van Rossum |
| Origin | CWI, Amsterdam, Netherlands (~1989) |
| First public release | 1991 |
| TIOBE Rating (Jul 2025) | 26.98% (all-time record) |
| PYPL Rating (Aug 2025) | 30.5% (first place) |
| TIOBE Lead over #2 | 15+ percentage points over C++ |
| Python 3.14 Released | October 7, 2025 |
| Python 3.15 Target | October 2026 |
| Python 3.10 End of Life | October 2026 |
| Creator’s current employer | Microsoft |
| Governing body | Python Software Foundation (PSF) + Steering Council |
| Documentary | Released August 29, 2025 on YouTube |
The Origin Story People Keep Getting Wrong
Everyone acts like Python’s success was inevitable. It was not. Not even close.
Guido van Rossum admitted on camera — in a 90-minute documentary released on YouTube in August 2025 — that he flat-out rage quit as Python’s “Benevolent Dictator for Life” in 2018. Those are his words. Rage quit. The man who invented the language, walked away from it in frustration.
The documentary, produced by Cult.Repo (formerly Honeypot) and directed by Ida Bechtle, premiered at PyCon Greece 2025 and pulled 198,000 views in four days. It’s free to watch. It covers the CWI origins, the community that built Python up, the Python 2-to-3 disaster that nearly killed everything, and the slow rebuild afterward. Guido is in it. Travis Oliphant is in it. Barry Warsaw is in it. All names you’ll care about if you’re deep in this world.
The Python 2 to 3 transition deserves its own moment of silence. Years of fragmentation. Libraries refusing to update. Companies holding on to Python 2 well past its death date. It nearly fractured the entire community permanently. Instead — somehow — the language came out stronger on the other side. That’s not a given. That’s a miracle.
See also “AllysPlay.com Review: Is This Amsterdam-Based Game Store Worth Your Money?“
The Dominance Is Actually Unprecedented
Let’s stop treating Python’s popularity as background noise. This is statistically unusual.
In May 2025, Python hit 25.35% on the TIOBE index. To put that in context: the only time a language has had a higher score in TIOBE history was Java in June 2001 — and TIOBE was only tracking around 20 languages back then versus 282 today. Python did this with the full competition running.
By July 2025, that number climbed further to 26.98%. A record. TIOBE CEO Paul Jansen literally titled his commentary “Python rules the world.” That’s not breathless blogger hype. That’s the CEO of the index company.
The lead over second place? Over 15 percentage points ahead of C++. First time any language has ever held that kind of gap in the index’s history. Nobody was debating Python’s throne before — but now it’s not even a competition.
AI is part of this story. TIOBE noted that AI coding assistants generate more Python than any other language, and that creates a feedback loop. More Python tutorials get written. More Python content gets indexed. More Python shows up everywhere. A language that was already winning started winning harder because the machines chose it first.
The GIL Was Python’s Original Sin — 3.14 Finally Addressed It
Here’s the thing about Python that its fans sometimes tiptoe around: it had a fundamental concurrency problem baked in for decades. The Global Interpreter Lock — the GIL — meant that even on a 16-core machine, Python code was mostly running on one core at a time. Threads existed but they weren’t really running in parallel. It was a limitation that made performance-critical developers groan for years.
Python 3.13 introduced free-threaded execution as an experiment. Python 3.14, released October 7, 2025, made it officially supported.
There’s a catch. And it’s a big one if you’re not paying attention. The free-threaded version doesn’t install by default. On Windows you have to specifically run py install 3.14t. On Mac it’s a custom install option. And on Linux? You have to compile it yourself. This is “official support” in the most technically-accurate-but-practically-frustrating way possible.
There’s also the JIT problem. Free-threaded builds don’t support the JIT compiler. So you can have parallelism, or you can have just-in-time compilation. Not both. Not yet.
But here’s the actual good news: the performance penalty for running free-threaded code compared to standard builds went from roughly 40% slower in 3.13 down to around 5-10% in 3.14. That’s a massive improvement. And on CPU-bound parallel workloads, the gains can hit 10x or more. That’s not nothing. That’s the GIL cage opening.
Python core developer Thomas Wouters — who deployed Python internally at Google for 17 years before moving to Meta, and is now on the 2026 Python Steering Council — put it bluntly: the GIL is going away. His colleagues on the Steering Council might not have formally decided it yet, but he said it like a fact.
The PyCharm Drama Nobody Wanted
JetBrains dropped a bomb on the Python community in April 2025: PyCharm Community Edition was being discontinued.
Well. Sort of. The split between free Community and paid Professional editions got collapsed into one unified product. JetBrains says the core functionality remains free. The community says some features that were free in Community are now locked behind a license. The GitHub source code stays public. But you can’t redistribute the new binary in third-party products the way you used to.
For people who’d been using PyCharm Community for years — at school, at their first job, on side projects — this landed wrong. Not because the price jump was huge, but because the principle changed. PyCharm 2025.2 was the last standalone Community Edition release. After that, everything merged.
The counterargument is that the unified version actually includes more features accessible for free than before. JetBrains says Jupyter notebook support, for example, is now available to everyone. Maybe they’re right. But trust is hard to rebuild once people feel a rug pull, and a chunk of the community absolutely felt that here.
The PSF Hit Its Budget Cap. In August.
The Python Software Foundation — the nonprofit that funds Python development, supports the community, and runs PyCon — paused its Grants Program on August 1, 2025. Not because of mismanagement. Because they ran out of their annual grant budget months early.
The PSF funds community events, educational initiatives, regional Python meetups, and developer tools across the globe. The fact that demand ran past the budget by August tells you something about how fast the community is growing. It also shows the gap between Python’s commercial dominance and the nonprofit infrastructure trying to support it.
The language is worth billions to the companies that use it. The foundation funding it hits a cap before the year is halfway done. Worth thinking about.
The PEP Process Is Broken and People Are Finally Saying It Out Loud
Python Enhancement Proposals — PEPs — are how the language changes. Someone proposes a feature. It gets debated. The Steering Council approves or rejects it. Simple in theory. Hellish in practice.
Barry Warsaw invented the PEP system. He has been a Python core developer for over 30 years. And in late 2025, he went on the Talk Python to Me podcast and said clearly: authoring a PEP is emotionally draining, the discussions on discuss.python.org can be toxic, and the whole process has become a gauntlet that core developers themselves try to avoid.
Let that sit. The people who maintain the language try to avoid the official process for changing it.
Warsaw’s concern isn’t just workflow friction. He pointed out that Python has millions of users worldwide who would be directly affected by decisions about lazy imports, free-threading, and other major changes — but those people have no voice. They don’t know they have a voice. The PEP discussion boards are, as Warsaw put it, just “the tip of the iceberg” of actual user sentiment.
His goal for 2026: figure out a better way. More inclusive, less toxic, more representative. Whether anything actually changes remains to be seen. But the fact that someone with Warsaw’s tenure is saying this publicly matters.
Python 3.15 Is Already Looking Interesting
The ink is barely dry on 3.14 and 3.15 is already in beta. Beta 3 dropped June 23, 2026.
Two big features are the story here.
Lazy imports (PEP 810): approved by the Steering Council in November 2025, shipping in 3.15 as October 2026. The idea is simple and powerful. When Python starts up, it loads every imported module even if you never use most of them. Lazy imports defer that loading until the module is actually needed. Meta reported a 70% reduction in startup time after converting libraries to lazy imports internally. Hudson River Trading migrated their whole firm to lazy-by-default Python and cut multi-minute startup times down to seconds. The pypistats CLI went from 104ms cold start to 36ms. That’s a 3x improvement from one keyword change.
frozendict: Python is finally getting a built-in immutable dictionary type. PEP 814, accepted by the Steering Council in February 2026, adds frozendict as a native built-in in 3.15. No import required. It’s hashable, supports type annotations via frozendict[str, int], works with union operators, and compares equal to regular dicts with the same contents. People have been asking for this for years. It’s here.
Also coming in 3.15 is Tachyon, a new statistical profiling sampler via PEP 799, alongside a new packaging governance structure — a Python Packaging Council was announced in May 2026 to help manage the increasingly complex packaging ecosystem.
Python 3.10 reaches end of life in October 2026. If you’re running that in production, you need to have a migration plan to 3.12 or 3.14. This isn’t optional.
The Type Checking War Nobody Expected
2025 was the year that type checking for Python got genuinely competitive. And the weapons were written in Rust.
Mypy has been around forever. But 2025 brought a pile of new entrants: Meta’s Pyrefly, Astral’s ty, and Zuban. All implemented in Rust. All significantly faster than the old Python-based checkers. We’re talking “things that took hundreds of seconds now take under a second” faster.
Gregory Kapfhammer, a CS professor at Allegheny College who’s on the 2026 Python Steering Council, noted that these Rust-based tools are changing developer behavior. When a type check runs in milliseconds instead of minutes, you actually run it constantly instead of occasionally. That changes how you write code.
There’s also a type server protocol (TSP) in the works — Brett Cannon mentioned it specifically. The idea is that type checkers like Pyrefly could feed type information into a shared protocol that IDEs use, instead of every editor integrating each type checker separately. More standardization. Less fragmentation. If it works.
uv from Astral also deserves its own mention. It’s a package and environment manager that’s eating the Python tooling ecosystem. CS professors are using it instead of teaching Docker. Developers are putting it directly in shebang lines. Barry Warsaw on the Steering Council noted that uv is made by a company — Astral — and flagged the dependency risk. “There’s always the risk they might disappear.” A valid concern about critical infrastructure resting on a startup.
Final Words
Here’s the thing about massive popularity. It creates pressure. The GIL problem festered for years because fixing it was risky and hard. The PEP process got toxic because more people are paying attention now and some of them are not kind. The packaging story is still a mess — uv helps, but it’s a startup running critical infrastructure.
Python’s success is also partly a lock-in story. The AI ecosystem chose Python. TensorFlow is Python. PyTorch is Python. Scikit-learn is Python. If you’re building AI products, you’re building in Python whether you love it or not. That’s commercial gravity, not always pure merit.
But the 3.14 and 3.15 work is genuinely impressive. Free-threading after decades. Lazy imports that Meta proved work at scale. frozendict. A type checking ecosystem that’s suddenly fast. These are structural improvements, not cosmetic ones.
Python isn’t just coasting on AI hype. The core team is actually fixing the hard stuff. And Guido, for what it’s worth, now works at Microsoft working on CPython performance. The man who rage-quit came back and is pushing the language forward from inside one of its biggest corporate backers.
Whatever you think about Python’s warts, you can’t look at where this language was in 2018 — its creator burned out and gone, the Python 2/3 split still festering, the GIL still locked in place — and not be a little impressed by where it is now.
FAQs
1. Is Python still the most popular programming language in 2025?
Yes, by a significant margin. Python hit a record 26.98% on the TIOBE index in July 2025, making it the most dominant any language has been since Java in 2001. The PYPL index had Python at 30.5% in August 2025. Both measures put it firmly in first place.
2. What was new in Python 3.14?
Released October 7, 2025, Python 3.14’s headline feature was officially supported free-threaded execution — the removal of the Global Interpreter Lock (GIL) as an opt-in capability. It also added template strings (t-strings), deferred annotation evaluation, concurrent interpreters, and improved debugging tools. The free-threaded build (python3.14t) isn’t installed by default and doesn’t support the JIT compiler.
3. What is the GIL and why does removing it matter?
The Global Interpreter Lock is a mechanism in CPython that prevents multiple threads from running Python bytecode at the same time. It effectively limited Python to single-core execution for most workloads. Removing it (in opt-in free-threaded mode) lets Python programs use multiple CPU cores simultaneously, which can produce 10x or greater speedups on CPU-bound parallel tasks.
4. What is Python 3.15 adding?
Lazy imports via PEP 810 (allowing deferred module loading for massive startup time improvements), frozendict as a built-in immutable dictionary type via PEP 814, and the Tachyon statistical profiler. Python 3.15 is scheduled for release in October 2026.
5. Why did Guido van Rossum resign as Python’s BDFL in 2018?
Van Rossum has described it himself as a “rage quit.” The community debate around PEP 572 (the walrus operator) was particularly intense, and the cumulative stress of running all major language decisions through one person became unsustainable. He stepped down in July 2018. Python moved to a Steering Council model instead.
6. What happened to PyCharm Community Edition?
JetBrains announced in April 2025 that PyCharm Community and Professional editions were being merged into a single unified product. PyCharm Community 2025.2 was the last standalone Community release. The source code stays public on GitHub, but the new unified binary has different redistribution rules, and some features previously free are now locked to paid licenses.
7. Is the PSF in financial trouble?
Not exactly, but the Python Software Foundation hit its 2025 grant budget cap and paused its Grants Program in August 2025 — earlier than expected due to higher-than-anticipated demand. This isn’t collapse, but it signals that community funding hasn’t kept pace with community growth.
8. What is uv and why is everyone talking about it?
uv is a Python package and project management tool built in Rust by Astral. It’s dramatically faster than pip for dependency resolution and installation. It can be embedded directly in script shebang lines, handles virtual environments automatically, and has simplified Python workflows enough that university professors are using it instead of teaching Docker. The concern is that it’s made by a startup with commercial investors.
9. What is the Python documentary and where can I watch it?
“Python: The Documentary | An origin story” is a 90-minute film directed by Ida Bechtle and produced by Cult.Repo. It premiered on YouTube on August 29, 2025, for free. It features Guido van Rossum, Travis Oliphant, Barry Warsaw, Brett Cannon, and many others, covering Python’s history from its Amsterdam origins through its global rise, the Python 2/3 crisis, and its current dominance in AI.
10. What is Python 3.10’s end-of-life date?
Python 3.10 reaches end of life in October 2026. If you’re running Python 3.10 in production, you need to plan migration to 3.12 or 3.14. After EOL, no security patches will be issued.
11. Why are new Python type checkers written in Rust?
Speed. Python-based type checkers like Mypy could take tens of seconds or even minutes on large codebases. Rust-based tools like Astral’s ty, Meta’s Pyrefly, and Zuban run in milliseconds, making it practical to run type checks continuously during development rather than only in CI pipelines.
12. What is a PEP and who approves them?
PEP stands for Python Enhancement Proposal. It’s the formal process by which changes to the Python language, standard library, or processes are proposed and debated. PEPs are approved (or rejected) by the Python Steering Council, a five-person body elected annually by Python core developers. Barry Warsaw, who created the PEP system, has publicly stated it needs significant reform.
Discover stories that stand at the top with The Pinnacle Magazine.
