Lots of folks saying not to bother learning coding now, it will be done by machine learning soon, so here is how to monetize your academic expertise...
Optimizing aspects of coding with Machine learning means that coders get more productive because some tasks they do can be automatized. I don’t think it’s a reason against learning coding.
“Computer program fixes old code faster than expert engineers:
One unexpected byproduct of the work is that it lets researchers see the different tricks that programmers used on the old code, such as archaeologists combing through computational fossils.
“We can see the ‘bit hacks’ that engineers use to optimize their algorithms,” says Amarasinghe, “as well as better understand the larger context of how programmers approach different coding challenges.”
“Computer program fixes old code faster than expert engineers:
Fixing old code isn’t a job that most computer programmers want to do. In this particular example it’s simply code that works as a better compiler. It’s similar to how using Amazons S3 is easier then configuring your own server. Computer programmers get empowered by tools.
The problem of programming isn’t to translate natural language into computer code. It’s to think about both the structure of the problem and the structure of the solution.
In practice programmers also want to write safe code. The architecture you link to looks like it doesn’t make clear gurantees about the data. It uses a neural network that works a bit like a black box.
There are likely usecases where such a tool is useful but I don’t think it will bring replace the need for programmers in a meaningful way.
Optimizing aspects of coding with Machine learning means that coders get more productive because some tasks they do can be automatized. I don’t think it’s a reason against learning coding.
Is good to learn some python at least but..
“Computer program fixes old code faster than expert engineers:
One unexpected byproduct of the work is that it lets researchers see the different tricks that programmers used on the old code, such as archaeologists combing through computational fossils.
“We can see the ‘bit hacks’ that engineers use to optimize their algorithms,” says Amarasinghe, “as well as better understand the larger context of how programmers approach different coding challenges.”
http://news.mit.edu/2015/computer-program-fixes-old-code-faster-than-expert-engineers-0609
Fixing old code isn’t a job that most computer programmers want to do. In this particular example it’s simply code that works as a better compiler. It’s similar to how using Amazons S3 is easier then configuring your own server. Computer programmers get empowered by tools.
Here is a new tool architecture
Neural Generation of Regular Expressions from Natural Language with Minimal Domain Knowledge
http://arxiv.org/abs/1608.03000
The problem of programming isn’t to translate natural language into computer code. It’s to think about both the structure of the problem and the structure of the solution.
In practice programmers also want to write safe code. The architecture you link to looks like it doesn’t make clear gurantees about the data. It uses a neural network that works a bit like a black box. There are likely usecases where such a tool is useful but I don’t think it will bring replace the need for programmers in a meaningful way.