Commentary: Should we still be learning to code in 2023?

SINGAPORE: It took just seconds for GPT-4 to write code for a basic but functioning website, off a photograph of a crude sketch in a notebook, in a March demo by OpenAI. 

Tech enthusiasts and non-techies alike have been using its predecessor ChatGPT to generate code for websites and games, with no more than simple text prompts. Microsoft has developed code-writing software that runs on OpenAI’s tech.

For many years, learning to code seemed indispensable - a way to ready students for the future and keep workers relevant and employable as technology changed jobs. 

But now it seems that coding skills are no longer needed to write programs. Amid the massive layoffs in the tech sector, software engineers and developers were reportedly not spared.

The emergence of “Low Code No Code” (LCNC) platforms has sparked debates about the future of coding in a world where software development can now be achieved with minimal coding knowledge. 

The question arises: Should we still be learning to code in 2023? 

ALLURE OF LOW CODE NO CODE TECHNOLOGY

LCNC platforms offer compelling reasons for organisations to adopt them.

First, they allow agility, faster time to market and innovation. These platforms provide pre-built components and visual interfaces that streamline the development process. 

Users can quickly prototype and experiment with new ideas which can result in more creative and adaptive solutions. I have seen finance users develop dashboards that extract data in real time, significantly reducing report preparation time.

Second, they democratise software development and allow business users who might not have extensive programming knowledge to become citizen developers - empowered to create solutions for their use, fostering ownership. Coding was necessary to build sophisticated HR platforms a decade ago, but today a HR scheduler can build and customise their own application.

Third, LCNC technology has the potential to bridge the shortfall of tech talent by leveraging non-tech resources. This can reduce development costs since the costs of skilled tech professionals have risen significantly over the years, especially for smaller projects.

Employers can allocate resources more effectively, reducing the dependency on scarce programming resources. Developers can shift their attention to complex system programming and providing guidance to citizen developers.

LEARNING TO CODE CAN MEAN DIFFERENT THINGS

The short answer to whether we should still learn to code is, unsurprisingly, yes. Not everything can be developed using LCNC platforms.

Custom coding is still required to develop complex solutions that require bespoke logic, say for a tax processing system, or when integrating LCNC platforms with complex interfaces, legacy systems and other custom-built applications.

Non-standardised code used in LCNC platforms can also make it difficult to identify and resolve issues, demanding extra effort to effectively maintain applications. Custom coding also plays a crucial role in creating proprietary competitive advantages for organisations.

The increased involvement of citizen developers through LCNC platforms raises the organisation’s cyber security risks and vulnerabilities. With more individuals having access to development code, there is a need for robust software deployment processes and an enhanced cyber security framework for the organisation.

The longer answer is that learning to code can mean different things to different individuals and roles. 

For some, it will mean learning to use LCNC platforms and working out the basics of how to use the right prompts or assemble the right components into a viable solution. For others, it will mean acquiring a deeper understanding of programming concepts, architecture and frameworks that undergird development work, to get the most out of LCNC technology or venture into custom coding.

NEW OPPORTUNITIES AND CAREERS

LCNC technology presents new opportunities and career paths, in both tech and non-traditionally tech sectors that are increasingly incorporating more complex technology into their work. 

Individuals can acquire “tech-lite” digital skills, upgrade and reskill without the need for extensive programming knowledge. This is supported by various initiatives such as micro-credentialing programmes - qualifications in specific topics - by SkillsFuture Singapore and several institutes of higher learning in Singapore.

However, it is also essential to consider the potential increase in stress on non-tech employees taking on development roles on LCNC. 

The expectation of multitasking and meeting higher development standards using LCNC tools can lead to added pressure and potential burnout.  

Companies need to be mindful when assigning additional LCNC roles to employees, as there is high probability that these added LCNC roles might adversely impact the mental health of employees, if the employees are not able to cope.  

This can also raise questions about performance assessment practices and how employees are evaluated on the expectations or outcomes of the LCNC development activities. The line between some IT and business roles within an organisation may blur, which can lead to friction. 

Who should be the first point of contact when a production issue arises? Who is responsible for the maintenance and deployment of LCNC solutions? Organisations might need to relook into their organisational structure, and the roles and responsibilities of employees.

While the rise of LCNC platforms has transformed the nature of tech jobs, coding expertise remains a valuable asset for individuals to tackle more complex projects and challenges that require deeper technical knowledge.

Sam Liew is President of the Singapore Computer Society and Managing Partner of NCS Gov+.