Lately, Big Tech seems to have taken it upon themselves to revolutionize software development. Microsoft has taken the first step with GitHub and Copilot, a snippet suggestion feature that already generates 40% of the code for developers who use it. D Google wants to take it one step further, hand in hand with a secret project that seeks create code that can write, repair, and update itself. Of course, all this goes hand in hand with artificial intelligence.

As reported business insiderThe initiative was born under the name pitchforkand currently renamed Help for AI developers (AI help for developers). It’s part of Google’s “generative artificial intelligence” bet, and all indications are that it has secured the backing of Mountain View’s top executives.

The above carrier indicates that the idea was born in X, the research arm of Alphabet. However, last summer he moved to Google Labs to continue development. Of course, the details of how this potentially revolutionary tool works remain a mystery. However, some of the details that have come to light paint a very interesting picture of what to expect from this project.

Pitchfork, or helping AI developers, is itself a tool uses machine learning to teach code to write and rewrite itself. How will you receive it? Learn the styles associated with programming languages ​​and apply that knowledge to write new lines of code.

Google wants to revolutionize software development with generative artificial intelligence

However, the potential of the program turned out to be much greater than originally thought. Now the intention is to give life to a general purpose system that can maintain a standard of quality in the code, but does not depend on human intervention for development and update tasks.

It is not yet known if Google wants this AI-based tool to be used internally only; or if you plan to open it to the public. The project is at a very early stage at the moment and is not expected to go live immediately.

In addition, Mountain View residents must resolve a few issues before showing it publicly. Both legally and ethically. Let’s not forget that the Californian firm came into the spotlight this year in the case of an engineer fired for claiming that LaMDAhis artificial intelligence model for natural language conversations became known.

Although, from the outside looking in, the type of code that can write and rewrite itself seems to create more technical problems than otherwise. After all, a mistake in one line of code can lead to serious bugs. We’ll have to see how this story develops.

Source: Hiper Textual

Previous articleThe iPhone 15 will again be more expensive, and this time Apple will not be completely blamed for it
Next articleSamsung has become an independent supplier of OLED displays for the iPhone 14.

LEAVE A REPLY

Please enter your comment!
Please enter your name here