In the world of software development, the arrival of artificial intelligence has been met with both excitement and apprehension. On one hand, AI has shown remarkable aptitude in generating code, completing tasks faster and more accurately than human programmers ever could. On the other hand, this has led to fears that AI could replace human coders altogether, rendering their skills and expertise obsolete.
The impact of AI on the future of coding cannot be overstated. While many have long considered programming to be a safe haven from the relentless march of technology, recent advances in AI have challenged this assumption. Researchers at OpenAI have found that 19% of workers in the US hold jobs in which at least half of their tasks could be completed by AI. Surprisingly, the most vulnerable jobs were those that required more education and offered higher salaries.
One skill set that is particularly at risk of being automated is computer programming. Large language models, such as the one powering ChatGPT, have been trained on vast repositories of code, making them exceptionally good at generating code themselves. Researchers at Microsoft and its subsidiary GitHub found that software developers assisted by AI were able to complete tasks 56% faster than those without assistance.
Tech companies have already begun to embrace generative AI, recognizing its potential to revolutionize programming. Amazon, for instance, has built its own AI coding assistant, CodeWhisperer, which is being used by its engineers. Google has also developed Bard, a competitor to ChatGPT that offers new coding features. It’s not hard to imagine a future in which AI will make half or even 90% of engineers redundant.
But before we give in to this dystopian vision, it’s important to consider the potential for a more positive outcome. One possibility is that the demand for software will grow so rapidly that we will need both human and AI coders to keep up. As Zachary Tatlock, a professor of computer science at the University of Washington, puts it, “It’s unclear if there’s any cap on the amount of software that humanity wants or needs. We haven’t been meeting software demand.” AI could help us produce software faster, but we’ll still need human coders to build it.
Another potential outcome is that human coders will find new ways to make themselves useful by focusing on tasks that AI cannot yet perform. This is similar to what happened to bank tellers after the widespread adoption of ATMs. Rather than being rendered obsolete, bank tellers evolved into salespeople, building relationships with customers and selling them on additional products and services.
The rise of AI is undoubtedly changing the game for coders. While it is possible that AI could render human programmers redundant, there is also the potential for a more positive outcome in which the demand for software grows so rapidly that we need both human and AI coders to keep up. Regardless of what happens, the future of coding is going to be very different from what we know today.