Generative AI has been the major disruptive technology of the last few years. It’s changed how people search the internet, interact with organisations, and create new content. AI can also generate code — instructions that tell software and hardware what to do.
With the digital skills gap in the UK still very much prevalent, coding has been at the forefront of what young people “should” learn to give them the best life chances in the future. But could AI be rendering coders redundant?
A Brief History of Coding
Coding began in the 1840s with Ada Lovelace, arguably the first-ever woman in tech and a computing pioneer. Interestingly, she also considered the possibility of the computer as a generative device. She specifically wrote about the ability of a computer to produce “scientific pieces of music,” something that AI can now do.
John Backus developed the first programming language, FORTRAN, in the 1950s for IBM. Moving into the 1960s, structured programming helped break code into modules for easier error checking. By the ‘80s, object-oriented programming (OOP) has been developed. Creating code as objects means it’s possible to create reusable code rather than crafting everything manually.
The advent of the internet prompted language designed for the web: HyperText Markup language (HTML) and Cascading Style Sheets (CSS). Today, developers can use container-based platforms that help create software that works across multiple devices and operating systems.
The Emergence of AI Coding Tools
The growth of device-agnostic, reusable code has made development faster and simpler. It’s also allowed AI-coding tools to become a major part of the development cycle. Automating the addition of reusable operations makes development faster. However, AI has now gone beyond filling in the gaps with pre-created code and started developing from scratch.
GitHub Copilot is a hugely popular example of these tools. Copilot combines chat and coding to help developers tweak their code or create operations from scratch.
Devin AI is a recent development from Cognition. Billed as a “fully autonomous software engineer,” Devin writes code from prompts, fixes bugs, and claims to be able to handle coding across the entire development lifecycle. Some media outlets have dismissed these boasts as hype. But if AI can manage coding tasks from software conception to deployment, are coders becoming obsolete?
The Shifting Focus of Tech Proficiencies
Some industry leaders suggest so. Jensen Huang, CEO of tech giant NVIDIA, recently spoke at the World Government Summit in Dubai. He announced that AI could make coding obsolete, and that AI developers’ job is to, “…create computing technology such that nobody has to program. And that the programming language is human, everybody in the world is now a programmer.”
He’s referring to the combination of natural language processing/generation (NLP/NLG) and no-code/low-code platforms or tools. These allow users to interact with AI with queries in their own language rather than coded instructions. GitHub Copilot does this to a certain extent, allowing queries to be entered in the user’s usual language.
No-code development apps and even AI-model creators now exist. However, this doesn’t mean coding is obsolete. For a start, these tools aren’t self-created. Someone needs to develop and program each of these low-code platforms, and many require continuous improvement — upgrades, bug fixes, and changes to match shifting consumer demands.
Huang went on to say that tech professionals can “future-proof” their careers by learning AI programming skills. Exploring machine learning, data science, and AI training or model creation could bolster coding skills for a more certain career trajectory.
Code is at the Foundation of AI
The bottom line is that without coding specialists, there is no AI. AI development requires professionals who know how to build algorithms, manage data, and train models — all talents that require coding skills.
The deployment of AI tools requires a deep understanding of cybersecurity and systems management. Software engineering and network administration are also essential for integrating AI with existing systems. And, no matter how advanced AI coding tools become, there must always be a human at some stage of the process to test, check for errors, and report concerns.
Continuous Learning and Adaptation is Key
Regardless of current and future advancements, the most successful tech professionals and pioneers are those who can adapt to the rapidly shifting landscape with enthusiasm. A diverse skillset is the backbone of a healthy career in tech, which is why learning development skills continuously and from a young age is so vital.
So, should young innovators learn coding or AI skills? The answer has to be both. Coding skills should be supplemented with an understanding of AI and machine learning. AI is a useful tool, and helps programmers work faster and smarter — but it doesn’t replace them. Learning about the potential benefits and risks of AI gives young people the know-how to get ahead in an increasingly digital world. For existing tech professionals and programmers, AI is essentially another spanner in their toolbox, and could help provide skills that are transferrable across many industries.
Girls Into Coding provides FREE workshops and events that get girls exploring coding, engineering, and emerging technologies like AI. We’re transforming the girls in tech of today into the women in tech of tomorrow. Find out more here.
Sign up here to our Girls Into Coding mailing list to be among the first to receive our articles and to know about future events!
