Assisting You to Write Python Is Just the First Step for AI to Replace Programmers

By Super Neuro

According to the current development trend of AI, it should be planned to replace deliverymen first, then drivers, and then receptionists, security guards, ticket sellers, etc. Finally, one day it will extend its claws to the engineers who created them, but before that day comes, AI plans to get along well with programmers and act as an assistant.

Kite is this "ambitious-looking" product. It is no longer a basic auto-completion, but a software that provides real-time code snippets.

Not long ago, Kite also received $17 million in Series A funding from Trith Ventures, and the CEOs of GitHub and Dropbox both participated in the investment.

Make coding easy with AI and ML

At first, Kite did not intend to replace programmers, but felt that spending time reading documents and searching for various code examples was too much effort, and Kite wanted to better solve this problem.

"Our mission is to use the latest achievements in AI and machine learning (ML) to make the process of writing code fun and easy."Adam Smith, founder and CEO of Kite, said in an interview:“Using Kite, developers can save time searching for code or information on the Internet and focus on the next technical challenge.”

With this original intention in mind, Kite developed this product, which is a local application plug-in that supports Windows and macOS systems. It was first launched in 2016. Now its functions are becoming more and more complete and it is loved by more people.

Simply put, it is to use AI technology to help programmers get a better experience when writing code, and beginners can also use this tool to start their coding journey faster.

Currently, Kite has been integrated with Atom, VS Code, Sublime Text, PyCharm, IntelliJ and Vim, and more than 30,000 Python developers worldwide use Kite.

By the way, currently Kite only supports Python.

 Kite is more than just auto-completion

Kite allows users to complete code supplements or function calls with a click, greatly reducing the need to jump out to find information.

This feature is similar to auto-completion, Kite will provide relevant code snippets as you type real-time suggestions. Unlike auto-completion, Kite's most important feature is called "full-line code completion".

Traditional code completion only helps you one code word at a time, but Kite takes this auxiliary function to the top level in the industry. It can predict the code statement that will be entered next, and sometimes even provide a complete line of code.

This is due to their own research and developmentType inference engineKite uses thousands of public code sources from high-star programmers on GitHub to train its machine learning model, using static analysis and machine learning techniques to achieve powerful intelligent recommendation capabilities.

When in use, Kite retrieves the user's code, the code of the entire project, and code resources publicly available on the Internet (such as GitHub and StackOverflow), and then makes intelligent recommendations based on context and frequency of use. The order of display is also based on the user's habits to list the best recommendations.

Its use is also very simple. It provides packages containing these letters when you enter them. The types of these packages include methods, usage documents, and sample codes.

 Kite’s ultimate goal

When Kite was first released, they considered putting it in the cloud because they believed that cloud processing had obvious advantages. However, they did not expect that privacy and security issues would be of great concern to users.

After a few years of seeing how users interacted with Kite, they finally made the switch, and their CEO summarized the advantages as follows:

1. Low latency.First, low latency completion is critical to the Kite experience. Whether running online or locally, Kite runs at blazingly fast speeds regardless of the internet connection.

2. Implement local high-performance tasks.With some optimizations to the Python analysis engine and AI models, Kite can already be used in a desktop setting.

3. The importance of security and privacy.Improving security and privacy is always right and of utmost importance. Many users are concerned about putting their code in the cloud. Kite has now made improvements and taken the best preventive measures, now allowing users to keep their own code base.

Although the number of engineers worldwide has increased significantly, it still cannot meet the growing demand. However, there are not many intelligent assistance tools like Kite, and there is still a lot of room for improvement in functionality.

Kite, which has completed the new financing, still has a long way to go. For example, this year, they hope to support the Linus version as soon as possible and develop versions that support more languages.

Although the ultimate goal of completely replacing programmers seems still far away.

But it is conceivable that Kite, which has been learning the world's best code writing methods, will one day have the opportunity to surpass those programmers who are inefficient and whose code is not neat enough.

Share this with the first batch of programmers you think will be replaced, and let them tremble in this New Year.