Doomsday Predictions for Coders

First there’s reality and then there’s hype. When ChatGPT4 came calling last November, pretty much everyone described it as the next best thing after sliced bread. However, slowly the hype gave way to reality as business verticals shifted away from treating generative AI as the ultimate job destroyer known to humanity to one that could assist its growth. 

While the big tech companies have openly or surreptitiously decided to slash some jobs – one of them being Microsoft’s Satya Nadella – others seem to have taken a more rigid approach. Take the example of Stability AI CEO Emad Mostaque who believes Indian coders will be jobless within the next two years. 

He has shared his views across media platforms that India’s outsourced coders of up to level three programmers will be out of work while those in France may survive – not because they’re better at coding, but due to better worker protection measures in the country. In fact, Mostaque had claimed in a June 29 podcast that AI would replace most programmers in five years. Maybe the enormity of this statement forced him to retract later. 

Of course, there are equally strong opponents to this theory. Thomas Dohnke of GitHub thinks AI could be actually assisting developers to become more productive and efficient. He holds the view that having an AI assistant for handling the more mundane tasks could free up coders to spend more time expanding their knowledge and becoming better at complicated coding. 

So, what could be a more realistic scenario?

Amidst these divergent views, let’s take a look at a more realistic scenario. That software appears to be eating into the software industry is quite obvious. Computers doing work that was done by humans isn’t new and stretches from word processing to factory floors. In fact, some economists call it the “skill-biased technological change” and believe that such instances make workers more productive and capable of taking on more complex tasks. 

The skill bias has been a constant through industrialization and is now moving into the digital world. Coding just happens to be an area that captured our immediate attention, given some of the experiments carried out by coders on ChatGPT. In fact, Microsoft created GitHub Copilot that uses OpenAI tech. Which is why Dohnke’s words make sense and Mostaque doesn’t. 

Most tech layoffs in 2022 and till date were driven by the need to cut costs and if coders were rendered jobless, the blame had to be taken by the macroeconomic trends where fresh recruits were at the receiving end. Experts argue now that AI can actually make programmers more productive and also result in cost savings where one person can do more work. 

Who says coders haven’t been using AI?

Stack Overflow, a company maintaining user-generated data for questions and answers for programmers, did a survey where 70% of the 90,000 respondents admitted to using AI tools in their daily work. And a third of them said its use made them more productive. Which is why companies like Apple, Nvidia and Meta have promised to grow despite layoffs. 

Another aspect that stands out amidst some of these doomsday predictions is that companies are hiring developers, AI or no AI, and the payouts are better than ever before. Then there is also Braintrust, a talent network representing 360K freelancers who are used by customers such as Nike, Google and Meta among others. Maybe, the gig economy will get a leg up too.

In fact, what’s intriguing is that Braintrust itself uses AI that is trained to handle all its job postings, resumes and successful matches between them. What’s more,the company hasn’t really let go of people – it’s merely organized them better and are currently skilling them for handling more complicated tasks. 

Where does the artificial get taken out?

In fact, we spoke to a few senior developers who were quite skeptical about the level to which AI can help coding. They believe that designing customized solutions to handle complex problems would still remain a human-centric work. Also, it may take AI much longer to understand the nuances of existing code at enterprises. 

However, there would be some tasks that would become redundant, these coders admit. Off-the-shelf issues and solutions can be built faster and easier with AI. Earlier this year, IBM’s head honcho Arvind Krishna mentioned that customer service, human resources and some positions within finance and healthcare could be lost due to automation. 

He felt that AI could solve the labor shortage issues at a global level, especially in the light of demographic challenges that the world is facing. Some of the AI tools can definitely replace portions of the labor challenges, which would result in a shift in the education system itself as is being seen currently in Germany. 

For example, the manufacturing sector’s automation resulted in a dual academic system where people do both apprenticeship and university study within their four years at college. Now that the same scenario is developing in the knowledge economy, a similar rethink around education for early white-collar workers may arrive sooner than we imagine. 

Till such time, it’s better to be a fence sitter than climb on to either of the two bandwagons – one that spells doom for coders and the other that believes everything is hunky dory.