Program designers may need new careers in another 20 years, and a paper published by the US Department of Energy's Oak Ridge National Laboratory points out that Machine Generated Code (MGC) may look like 2040 Is nowadays artificial intelligence, or a few years after the self-driving technology as usual, both in the academic institutions or markets have been able to drive the development of MGC's main technology.
Researchers also mentioned several papers in the paper and pointed out that from these papers one can see the axis of technological development, and now it is possible to write new code by grabbing code, High-speed machine learning technologies like Defense Advanced Project Agency's (DARPA) and Probabilistic
Programming for Advancing Machine Learning (PPAML), it is to enhance machine learning to solve some unspecified problems, and Both DeepCoder and AutoML have been able to produce executable code, the ontology generation tool DOG4DAG can produce human semi-automatic due to time constraints Unable to complete the knowledge.
The paper mentions that in the future people will spend most of their time relying on autocomplete or feature recommendations even if they still need to write programs.When machine learning is advanced enough with natural language processing techniques, the quality of the code produced by the machine will be better than that of a human software engineer Prepared even better.
However, MGC still has its own challenges. Due to the heterogeneity of the programming platform, MGC will face the same dilemma as High Productivity Computing Systems and Domain Specific Languages. Scaling scientific computing from scratch, and the high cost of this major project hindered the advancement of technology
The paper concludes with a statement that a highly heterogeneous computing environment requires changes in usability and productivity on demand, and it is even easier to write and execute code in natural language and even machines that can eventually interact with each other without human intervention Division of work to complete the task.