Most of us were in awe when ChatGPT was released in November due to its uncanny capability of writing rap lyrics, cover letters, and high-school English essays. Nonetheless, software developer Adam Hughes was captivated by how AI could write computer code.
ChatGPT was signed up for by a man who wanted to customize a tic-tac-toe game while testing the bot with coding questions he typically asks in job interviews. The changed rules would prevent the bot from copying code already written by humans.
Hughes didn’t expect the result he got when he threw ChatGPT at any task – amazingly good code. This made him question what it meant for his beloved career, with its financial stability and the personal identity it provided him with.
“I never thought I would be replaced in my job, ever, until ChatGPT,”
“I had an existential crisis right then and there. A lot of the knowledge that I thought was special to me, that I had put seven years into, just became obsolete.”
Coding has historically been seen as a refuge from the ever-increasing developments of technology due to universities rapidly expanding their programmes related to computer science. Thus, it was perceived that those who created the commands for machines would be immune to jobs being replaced by novel tools.
Policymakers’ unanimous suggestion for futureproofing the workforce is that everyone should learn to code. However, in recent times, numerous coders have expressed apprehensions about the introduction of generative AI.
Those who have been automating worry that their jobs may soon be automated. If even programmers are not immune, who else is there to trust?
AI is often discussed as a threat to white-collar employment. OpenAI recently sought to understand this better by researching the employability of large language models to complete the 19,000 tasks that make up the 1,000 occupations in America.
Around 19% of careers involve duties and tasks which artificial intelligence (AI) could easily perform, according to the researchers. They found that those roles requiring more education and which offer significantly higher salaries tend to be at the highest risk of being replaced by AI.
Ethan Mollick, a professor of management at the renowned Wharton School and a reputable expert in innovation, focuses his studies on this field.
Ethan Mollick says:
“We didn’t think that would be the case,”
“AI was always supposed to automate dangerous, dirty tasks — not the things we want to do.”
With the possibility that white-collar skills could be automated, computer programming is especially vulnerable due to large language models like ChatGPT being trained on code repositories. In a study conducted by Microsoft and its subsidiary GitHub, two groups of software developers with differing access to an AI coding assistant were used to explore this concept further.
The automatization AI provided enabled tasks to be completed- 56% faster than not having it, a much more significant increase in efficiency than with the introduction of the steam engine in the mid-1800s, which only improved productivity by 15%.
Mollick continues to say:
“That’s a big number,”
Google and Amazon are at the forefront of utilizing generative AI for programming, with Amazon’s CodeWhisper and Google’s ChatGPT being employed as coding assistants. Both firms have urged developers to experiment with various features for faster coding results.
With the tech industry prioritizing AI, there is a high likelihood of engineering jobs being considerably reduced in the future, perhaps even down to one-tenth or one-hundredth of their present numbers (Emad Mostaque, CEO at Stability AI, predicted that “there’s no programmers in five years.”
Let’s pause and consider both sides before diving into this doomsday scenario. Optimists suggest the industry’s rosier view is true, and there is enough coding demand to support robots and humans.
Undoubtedly, the advent of the tractor put many farmers out of work. However, Zachary Tatlock – a professor of computer science at the University of Washington – states that coding cannot be compared to farming.
Zachary Tatlock says:
“There’s only so much food that 7 billion people can eat,”
“But it’s unclear if there’s any cap on the amount of software that humanity wants or needs. One way to think about it is that for the past 50 years, we have been massively underproducing. We haven’t been meeting software demand.”
Coding will remain in high demand in the future because AI can only help humans create software faster; as a result, this productivity boost from AI calls for even more coders than before. In the best-case scenario, those who specialize in coding should expect plenty of opportunities to come their way.
Even though AI has assumed many coding responsibilities, optimists argue there is still a place for human coders to be useful. It is believed that, like bank tellers, human coders can focus on what AI cannot do and find other ways to be productive after ATMs are introduced.
Despite the belief that ATMs would put them out of work, bank tellers experienced an increase between 1980 and 2010. It was due to their transformation from checkout clerks into salespeople who could build customer relationships and promote additional bank products such as credit cards and loans.
Software engineers of the future might be spending less time writing code and more time verifying it. According to Tatlock, this is necessary due to how much low-cost and possibly hazardous code machines will likely produce.
Tatlock continues to say:
“You probably don’t need to formally verify a widget on your website,”
“but you probably do want to formally verify code that goes into your driving assistant in your car or manages your insulin pump.”
As technology advances, it is believed that the future of programming will involve writers, editors, and fact-checkers. Thus, those who currently code today may eventually have to take on different roles in the field.
The optimists allow that, while human coders may survive long term in some new role yet to be determined, the transition will be difficult. Regardless, it still looks like automation will replace many of their current responsibilities.
Tatlock adds to say:
“It is going to be the case that some people’s lives are upended by this,”
“This happens with every technological change.”
Coders who fail to keep up with the ever-changing technological landscape may find themselves without a job. In contrast, those that successfully adapt to this AI-driven future will be performing unfamiliar activities.
Given the competition in today’s world, who has the most potential to develop and adapt? Conversely, who might struggle? There is a finite amount of food that can serve 7 billion people. But when it comes to software, it is unclear whether a limit exists to how much humanity desires or requires.
Those with more experience in the field – spending their time on conceptual, high-level thinking rather than technical coding – would be less prone to being displaced by AI than those freshly graduated and simply writing code. But in the GitHub study, the less experienced engineers benefited more from using AI.
The emergence of new technology has made the experience less important, meaning the gap between newcomers and veterans has been diminished. Senior engineers might be disadvantaged, as they can no longer justify their large salaries.
Optimists believe that AI will allow us to assign tedious and monotonous tasks to bots, thus allowing us to focus on engaging work. However, what if the opposite occurs and AI takes on all the interesting jobs? This brings up questions about job quality.
No disrespect to my colleagues in the research department, whose work is invaluable, but I’m a writer because I cherish writing; I don’t want my job transformed into one of verifying the eccentric and faulty results of ChatGPT. What’s disconcerting about generative AI is its aptitude to accomplish the skilled tasks people most enjoy.
Hughes continues to say:
“I really love programming,”
“I feel like I’m one of the few people who can say for sure that I’m in the career I want to be in. That’s why it’s scary to see it at risk.”
Predicated though it may be on the assumption that generative AI will serve as a complement to, rather than a replacement of, human labour, optimism concerning AI’s impact is still not without its glitches. The greatest being that it perhaps won’t be OK after all.
Another technology emerged a few decades ago that completely replaced a common job for young women – the mechanical switching of telephones – and this was when ATMs came along. Bank tellers had to adapt because they could do certain tasks better than the machines.
Using the rotary-dial phone, people can make their calls more quickly and conveniently than when going through an operator. Consequently, many of these operators left the workforce altogether or were pushed into lower-paying jobs.
AI technology has become so advanced that the question arises – when will AI be competent enough to render human programmers obsolete?
Why are our best minds researching AI to replace human labour instead of developing technology that empowers and helps them do something entirely new? This question highlights one of the foremost issues with AI research. Far too much focus is being put on replacing human labour when it could be used to enable something different.
Katya Klinova says:
“It’s a sad use of innovation,”
Klinova, the head of AI, labour, and the economy at the nonprofit Partnership on AI, knows that huge issues, such as the pressing demand for more clean energy sources, need resolution.
A more pertinent inquiry we should be making about AI is not its capability to duplicate human activities or the financial gains it might bring to companies but if it is acting in a way that conforms with our societal ideals.
In the interim, coders should focus on improving their skills in areas that AI cannot yet conquer and keep up to date with the latest technology.
“I really think everybody needs to be doing their work with ChatGPT as much as they can, so they can learn what it does and what it doesn’t,”
“The key is thinking about how you work with the system. It’s a centaur model: How do I get more work out of being half person, half horse? The best advice I have is to consider the bundle of tasks that you’re facing and ask: How do I get good at the tasks that are less likely to be replaced by a machine?”
Mollick believes that people’s dismissal of ChatGPT after giving it only a brief try is misguided since AI technology is developing rapidly. He states that although some may initially feel comforted by their superiority over technology, progress will soon challenge this sentiment.
ChatGPT, powered by GPT 3.5, scored in the 10th percentile when it took the bar exam. However, GPT 4 improved drastically and achieved a score in the 90th percentile after only one year of study.
Mollick continues to say:
“Assuming that this is as good as it gets strikes me as a risky assumption,”
“The best advice I have is to consider the bundle of tasks that you’re facing and ask: How do I get good at the tasks that are less likely to be replaced by a machine?”
Hughes experienced a sense of dread when ChatGPT won his tic-tac-toe challenge; he also noticed the same response in other programmers. To cope, he wrote an illuminating post on Medium, outlining how, over the following decade, AI could become sophisticated enough to replace humans in programming roles.
The flood of criticism from developers in the comments section was overwhelming, some of it so harsh and vitriolic that Hughes was compelled to delete them. Post after post, devs argued why they still viewed themselves as superior coders compared to ChatGPT.
Hughes continues to say:
“You are a really bad software developer if you don’t understand the number of AI limitations,”
AI has been making strides in the workplace, but experts are confident that it won’t be able to replace what humans bring to their jobs shortly.
ChatGPT may not spell the end of coding, but it can potentially revolutionize how developers work and accelerate technological innovation. As with any new technology, we must harness its potential and use it to create a better future for all.
Source: Business Insider