It seems programmers refuse to admit that they are digging their own grave. Although the “grave” sounds too radical, the automation of the software development reflects the trend capable of forcing many junior and middle developers into unemployment.
Deep learning is coming
The deep learning (machine learning) technologies based on artificial neural networks stepped outside the limits traditionally inherent in the pre-programmed machines. Many AI systems demonstrate the capabilities that were only recently considered “sheer human”. The AI-enabled image and text recognition systems have outpaced humans with their efficiency and accuracy. A huge leap in voice recognition makes virtual AI-type assistants from Google, Amazon, and Apple equal to humans. Since machines are so good in processing human languages, the acquisition of more primitive programming languages is to be the next logical step of the AI evolution.
Automation is reaching coders
It is no secret that the biggest part of coder assignments belongs to compilation, recompilation, and repurposing different pieces of code available in repositories and libraries. Thus, routine tasks rather than creative ones prevail. The experience of the industrial robotization suggests that any routine task is potentially automatable. Moreover, one of the main postulates of the digital transformation insists on the mandatory automation of everything that can be automated. And it would be weird for software developers not to create a tool or a technology capable of freeing them from the coding routine.
The “best code” competitions can share the fate of Go game
The DeepCoder AI project from Microsoft explicitly hints at a grim future many entry-level programmers can shortly face. Shuffling code cuttings in order to create simple apps seems a perfect use of deep learning systems. Even at the current stage of development, DeepCoder is able to compete with human developers successfully. As is known, progress does not stand still. Having no human cognitive limitations, DeepCoder and the like are not lazy to study. The deep learning systems do never get tired of searching for the “fresh” pieces of code appearing continuously on GitHub and StackOverflow in the public domain. Sharing their experience free of charge, enthusiastic coders feed AI systems with knowledge without even knowing it.
The double-edged sword of code sharing
And here an ambivalent aspect of the sharing economy arises. On one hand, the society (in this case the coders’ community) benefits a lot from the free experience exchange. On the other hand, using this free experience the digital leaders are developing commercial products capable of shifting many of those coding volunteers from the labor market. The sharing economy is obviously a double-edged sword. The professional ambitions pushing human coders to share their “best code” publicly can create the shovel capable of digging their unemployment grave. Nonetheless, coders may hardly be accused of any “suicidal” activity. The corporate greed is the defendant to whom the issue is to be addressed.
Developing without coding?
While Microsoft represents the pure “substitute” of human programmers, there are some other indirect technologies aimed at the similar consequences. The AI-enabled website designing tool Grid invites everybody to become a professional web designer instantly online. Appearing out of nowhere (as its creators claim in the description), Grid offers profound web designing capabilities accessible online through its website. The advanced AI-powered image processing along with some other tools and functions of Grid allows even digitally inept users to create rather simple but attractive and professionally looking websites without coding at all. Of course, a huge corporate website with complex architecture is out of Grid’s league. However, the vast majority of the Internet users can be satisfied with such a web design editor making the designing process both simple and amazing. If things keep going this way, many web designers might appear out of business.
Who else can bring grist to the mill of AI?
Yet another threatening trend comes from online training courses inviting the “ordinary” programmers to retrain for AI specs. Moreover, even persons without any IT background can become AI engineers as it is announced by the Artificial Intelligence Nanodegree program from notorious Udacity. At first sight, the program focuses on the important social mission of adapting people to nowadays’ labor challenges. Nonetheless, in terms of the delayed effects, a small number of the AI engineers can indirectly support machines in making a huge programmers’ community unemployed.
For those who are still hesitant regarding the daunting unemployment prospect of programmers, we suggest this great post describing 10 IT occupations at risk along with the technologies that make them vulnerable.
In addition, several powerful groups of AI researchers reported progress in developing learning software capable of developing software. They include the nonprofit research institute OpenAI, MIT, University of California, Berkeley, and Google’s DeepMind.
Vinod Khosla, former Sun Microsystems founder and the billionaire founder of Khosla Ventures is confident about AI-type systems that will replace 80% of IT department staff rather soon. And he calls it “amazing”.
So what do we do?
The above-mentioned employment researchers Brynjolfsson and McAfee suggest people focusing on the activities and specialties that are either too embarrassing for computers or economically unfeasible for automation. Thus, the ideation (the creation of new ideas), complex pattern recognition, and motivating communication compose the meals inedible for AI in the nearest term.
Although re-skilling can hardly protect the whole number of the employees at risk, the timely performed retraining is worth considering. The vast majority of the population worked in agriculture in the recent past. Now only 2 percent of Americans are engaged in farming. Were the rest 98% left unemployed? No, they weren’t. They were reskilled.
Being aware of the threat posed to programmers by AI nowadays, Indeema hopes on the unprecedented adaptability of humans to any kind of environmental change. Since people have been able to survive over millions of years of evolution, they would find out the way to how to cope with AI too. Particularly when it comes to coders.