Is Automation eating Programming?

Writing software code includes many tasks that are repetitive and predictable. The past decade has seen significant advancement of Artificial Intelligence and Robotic Process Automation methods. Are we heading to a future where software applications are mostly designed with AI & Automation Tools, rather than fully written by human programmers?

If the above is true, how does this impact skills that private and public enterprises, and their service providers, need in the future? Extending that theme, will human programming become more creative than it is now, as a result of growing automation of logic?

The chart above the headline is from OECD, the international organisation that tracks and analyses economic development and jobs. It shows that the greatest risk of job replacement is, in general, seen in countries that have lower per capita GDP. However, this is a very broad generalisation across industries, and we must go several layers deeper to understand the implications on software as we know it today.

Artificial Intelligence and Software Automation have been evolving for many decades. They have been increasingly capable of augmenting human programmers in writing application code, in technologies generally known as “low code/no code development”.?Low code tools and technologies essentially use blocks of commonly used code in a predefined way to achieve typical compute outcomes. Other Digital advancements are improving our ability to build, deliver and maintain software. Superlative user experience which was once a secondary outcome, is now a prime objective for application development. Similarly, we now see different kinds of software testing being replaced by agile-friendly methods that ensure better first time build and easier routes to testing. Automation has multiplied the speed and accuracy of software testing.?

Beyond large enterprises, Low Code technologies have also empowered crowd sourcing of software. This method utilises gig workers who will be assigned a specific coding objective, and paid only for the deliverable. This enables the commercial leverage of members of the general public who have skills and inclination, to contribute to predefined application deliverables. Such a model is gaining traction where skills are not available at scale, or where a low cost is a primary requirement, or a distributed development model provides specific business advantages. Obviously this approach has its own problems.

All of this leads us to the question: Can human effort in programming therefore eventually become more focussed on the creative elements, while automation helps with fast and cheap logic-oriented coding? Interestingly, creative programming is not new; Even in the 1960s, visual creative was being produced through computer programming, though not at scale. More recently, Google has been investing in DevArt, which they describe as “Use of technology as canvas and code as raw material to create innovative Digital art”.

If coding can create Art, it begs the question of whether coding can also help drive Innovation. After all, Innovation has creativity as its foundation even when it is in the realm of technology. The often used phrase ‘Digital Engineering’ implies that a solution is being engineered using software and digital technologies, but in a creative manner. Digital Engineering includes ‘Art of the Possible’, which implies two things: Thinking out of the box, and creatively bringing together emerging technologies to multiply value. For example, if AI helps discern a pattern of learning disabilities in children based on nutritional status indicators, that knowledge can help improve educational outcomes in a non-traditional way. That too is ‘Art of the Possible’ using Digital Engineering.

With the advent of Agile Scrum methods in software development, we saw improved agility around changing customer requirements and market needs. With Low Code methods, we additionally obtain faster development and releases. Also there is space for developers to use their innate human creativity to solve problems more effectively, because some of their grunt work is handled automatically.

From a practical perspective, we must understand and avoid common causes of failure in Digital Engineering programmes, so we can fully exploit the benefits. Statistics show that upto 80% of Digital Engineering programmes fail mainly because of inadequate Change Management practices (Arthur D Little, 2020).?Any creative change entails some disruption, and the impact of disruption needs to be managed proactively. Organisational Change Management is an important functional capability that must be deployed on any Digital Engineering programmes in the new world. That way we ensure that the benefits accrued do not get diluted by the negative impact of disruptive change.

Digital Engineering using AI, Low Code and other emerging technologies will help beyond organisational effectiveness and speed. They empower Innovation and create a culture of problem solving, because they open up application building to a broader talent pool. They allow quick creation of virtual environments where new concepts or solutions can be safely tested without danger to people or property. They hide the complexity of computer code, logic and APIs to attract participation by those who better understand the practical significance of products, tools, devices, and services. And thereby, expand the need for higher value human input.

#automation #digitalchange #digitalacceleration #cloudadoption #codeless #lowcodenocode #artificial intelligence #ai #transformation #economy #jobs #applicationdevelopment #india