Programmers were thinking that their jobs are the most secure jobs in upcoming AI future, but new research shows that they aren’t secure. Programming is one of the toughest jobs out there it takes thousands of hours of practice to be a professional programmer. A new research shows that by 2040 AIs will be writing code for new software, Thanks to machine learning and natural language processing.
A research published by Oak Ridge National Laboratory, titled “Will humans even write code in 2040 and what would that mean for extreme heterogeneity in computing?” suggests that in just 2 decades AI will be doing most of the programming work for us.
Well, this isn’t surprising at all, Google’s AutoML and Microsoft’s DeepCoder already writing executable code with help of Machine Learning and natural language processing. And not just companies like Google and Microsoft have these technologies but The Defense Advanced Project Agency’s (DARPA) Probabilistic Programming for Advancing Machine Learning (PPAML) is also writhing Machine Generated Codes (MGC).
So why do we need MGC? and the answer is TIME, there is no doubt that programming a lot of time to learn and same applies building projects after learning. In the future, if a human does need to write to code, they will be using autocomplete and recommendation features. Means that there will be no for traditional programming.
The paper says that the main challenge of MGC will the understanding the problem domain of scientific research, not just the library itself. According to researchers writing code for scientific research would be more technological than sociological. And this going to be the last piece of the puzzle.
This could be a bad news for programmers but the irony is programmers developing machines to kill their own jobs. So what do think about one of the biggest change in programming by MGC?