Is Python the future of AI programming?

Python founder, Guido van Rossum, didn’t fully appreciate what he was building back in 1989. In a recent interview with The Economist, he said, “I certainly didn’t set out to create a language that was intended for mass consumption.” It’s like Zuckerberg saying he created Facebook to make friends.

Well, he may not have meant to build a juggernaut, but that’s what Python has become. Nearly three decades since its inception, the code language has squeezed the life out of rivals like Ada, Fortran and Lisp. Search queries for python have trebled since 2010 (in in the last year, more Americans Googled ‘Python’ than ‘Kim Kardashian’). The only other platform keeping pace is Javascript. Universities, including RMIT, are developing AI-targeted Python short courses.

So why has python stuck around? And why are we embracing an 80s-era codebase (albeit one that’s undergone several updates) on the cusp of an AI revolution? Will we all be speaking Python in 10 years’ time?

So, what is Python?

Python is a programming language that trades on simplicity and flexibility. The syntax is pretty easy for new coders to pick-up, but robust enough to handle almost any function, including AI. According to Wikileaks, the CIA uses Python as a hacking platform. Google relies on Python to crawl webpages for keywords. Pixar uses it to make animated films. Python’s users (known as ‘Pythonistas’) have uploaded over 145,000 custom-built software packages, many of which use ‘machine learning’ to crunch patterns in big data.

How will Python affect AI development?

In short? Python makes building neural networks simpler, faster and more accessible. There’s a famous example from 2015 where Milo Spencer-Harper built a basic neural network using just nine lines of Python – the program couldn’t exactly drive a car, but still, pretty impressive. Python also has good documentation, it’s platform agnostic, it’s robust, portable, scalable, and (best of all) open source. Even Peter Norvig (director of research at Google, and a long time Lisp fan) has picked it up.

Why should you learn Python?

Going by user trends, it’s likely that Python (with its easy readability and code-friendly syntax) will become the most universal AI programming language over the next twenty years. That’s a pretty safe bet. Research by Accenture has also shown that, not only is there an AI skills gap in Australia, but machine learning will double economic growth by 2035. Mckinsey has even singled out 19 different industries that are particularly susceptible to disruption. This there’s an opportunity: many industries will soon be in flux, and learning to code AI in Python is a smart way to not only future-proof your career, but leapfrog your competition.

Want to learn the Python basics? Check out RMIT’s new online course, AI Programming With Python (developed in partnership with Udacity).