AI and the downfall of the programmer
In the increasingly AI-dominated world, a paradox of sorts is unfolding. It's a scenario that has both tragic and poetic elements, particularly for those who have spent their lives in the fields of programming and developing.
Once lauded as the architects of the future, these skilled professionals are now, ironically, threatened by the very creations they have painstakingly developed. We're on the cusp of an era where artificial intelligence will displace a huge swathe of programmers, and the reverberations of this shift are destined to be felt far and wide.
High priests
Since the dawn of computing, programming has held an almost mythic status. The developers, with their unique ability to converse with machines, have often been compared to high priests, mediators between humans and the seemingly impenetrable world of binary logic. They have been the ones to unravel the Gordian knots of code, to orchestrate symphonies of software that power everything from our national grids to the tiniest microprocessors embedded in our household appliances.
But a spectre is haunting this technocratic priesthood – the spectre of AI. A new breed of AI algorithms have proven to be incredibly adept at generating, debugging, and optimising code. Armed with the ability to learn from vast repositories of existing software, these AI systems have become capable of carrying out tasks that until recently were the exclusive domain of human developers.
Code generation
OpenAI's GPT4, for example, has demonstrated a staggering capacity to produce human-like code, simply by being given a natural language prompt. Meanwhile, DeepMind's AlphaCode is making strides in its own right, aspiring to tackle complex coding challenges that have historically required human ingenuity and years of experience.
Human programmers typically generate 20 lines of code per day. From a plain-English description of what is required, AI can generate 20 lines of code in seconds, and can be paired with an independent AI to run the necessary tests on it, warn of weaknesses, and offer solutions.
Dr Aiden, Dentaljuce's AI tutor, was designed, built and embedded in the Dentaljuce web site almost entirely from code generated and tested by AI.
No room at the top
One might argue that programmers will merely move up the value chain, focusing on more complex, creative, or high-level tasks. This is a familiar narrative in the face of automation. Yet the pace of progress in AI development suggests this refuge may be ephemeral. As AI evolves, it is inevitable that it will be able to handle increasingly sophisticated programming tasks.
AI-generated Node JS code designed for hacking
Additionally, the proliferation of AI-driven programming could exacerbate existing issues in the tech world. Machine-generated code could further obfuscate the already complex digital landscapes, making it harder to identify bugs or malicious code. It could also lead to a homogenisation of coding styles and approaches, stifling the creative diversity that often leads to innovative solutions.
What happens when the high priests of technology find themselves outflanked by their own creations?
The downfall of the programmer, while not yet certain, offers a sobering glimpse into a future where AI's capabilities eclipse our own across an ever-widening range of tasks. Analysts at Goldman Sachs have said that overall, AI could ultimately replace 300 million jobs – a quarter of the global workforce. They identified administrative jobs as those most at risk, followed by those in law, architecture and engineering.
As we stand on the precipice of this brave new world, we must confront these challenges head-on and navigate a path that safeguards both our digital and our societal futures.