Advances in AI mean some tedious programming tasks will be replaced by programs

Advances in AI mean some tedious programming tasks will be replaced by programs
Software engineers need not fear redundancy, using frameworks and libraries was a big advance from writing machine code and I am sure there were some machine code programmers who thought it was the slippery slope to their obsolescence.

When writing a feature in machine code it would take a tremendous amount of effort to to achieve the same result that would now be trivial to do in a modern language. The same was true of the early compiled languages, javascript without libraries and frameworks requires much more hand rolling that is just plain time consuming.

There are benefits from pure simple coding, the efficiency and sparsity in environments where resources are scant and where memory saving is essential, reducing the number of lines of code and pairing functionality to the minimum set does lead to some beautiful lean programs. But they do take a lot of effort and a comparatively long time to write and get right, they generally do not do very much and are difficult to read and therefore hard to maintain. To refine them takes acres of refactoring, bench testing, metrics and, yes years of experience. The same brute force that mere mortals needed to find solutions can now be done by machines that aid program synthesis and inductive programming.

Luddites take note I for one will not miss the work that machine learning replaces and have mused on ‘How Machine Learning Builds Your Applications For You’ elsewhere in this publication. Beyond generating code and hyperparameter tuning, machine learning will improve software by understanding the users, improving the reliability and performance of applications and testing procedures. Taking the leg work out of programing will allow software engineers to hang up their mountain boots and polish up their dancing shoes, oh and make better software.

Washing machines did not remove the drudgery of domestic work from many women's life, scullery maids are maybe from another era and the wealth that maintained them in those days has in fact trickled down, and to an extent democratised many of the distinctions derived from rich and poor. It is true it has not equalised society totally but the difference between rich and poor has dramatically narrowed due in large part to labour saving technologies which brought about the emergence of the middle classes and the decline of the monied aristocracy. Machine learning is but another one of these societal technologies and the benefits it will bring, to healthcare for instance, are worth losing a few low paid driving jobs along the way.