27.10.21 at 09.00 am
Most of the products that millions of people rely on every day, from apps to smart home devices, is relied upon to create. It seems that we are on the verge of being able to program only from our natural languages without writing code.
This is because IBM announced several months ago a project called CodeNet. It is the same approach as the ImageNet project, which provides access to a huge database of images and their custom descriptions based on the deep learning of computers.
CodeNet comes as a database too, but for code – that is, the code – as it contains more than 14 million usable code samples. Which allows to solve more than 4,000 programming problems.
This project is being relied upon, as IBM explained, in various operations. It will also allow the identification of code for legacy systems automatically and based on artificial intelligence.
In addition, the CodeNet project relies on natural languages in the programming process. In recent years, companies have increased their efforts in developing Natural Language Processing (NLP) technologies, led by Google and OpenAI.
Programming in natural languages
CodeNet is able to understand the natural languages we speak and then convert them into carefully crafted code. Natural languages are defined as the languages spoken by humans, such as Arabic, English, French, and others.
This technology greatly facilitates the programming process, which is a difficult process that requires great skill, of course. This is where the user will be able to program only in their native language, such as to write in English: “Make a web page with a red background with a picture of a blue plane, my company logo and my email.”
Then the AI turns your requests into usable code. As you can expect the quality of the work may not be the same as that produced by humans, but it is still a very useful feature.
It is expected that Microsoft will participate in similar projects, especially after its acquisition of the famous programming platform GitHub several years ago, which includes the largest possible amount of open source code.
NLC and NLP technologies need significant development in the coming years, and are in themselves an excellent investment for leading technology companies. It is also expected that future projects from IBM, Google, and Microsoft will appear in this field.
CEO of OTTO Magazine