[ad_1]
A new AI system can read instructions written in conversational language and transform them into working computer code.
Why is this important: The model is the latest example of advancement in natural language processing (NLP), the ability of AIs to read and write text.
- But it also points to a future where coders can offload some of their work to AIs, and ordinary people can code without actually learning to code.
Driving the news: Today, OpenAI is releasing an enhanced version of its Codex AI model and releasing it for developers for private developers through its API.
- Codex is a descendant of OpenAI’s GPT-3 massive text generation model, which was released last summer.
- But while GPT-3 was trained on a huge amount of linguistic data pulled from the Internet – allowing it to read and then complete text prompts submitted by a human user – Codex was trained on both language and billions of lines of computer code available to the public.
- As a result, users can issue commands in written English, and Codex will produce computer code capable of executing those instructions, making it essentially English to Python (or JavaScript or Ruby or one of more than a dozen. of programming languages in the training of Codex data) computer code translator.
What they say : “We believe this is a tool that can remove barriers to entry to allow more people to get into computer coding,” said Greg Brockman, co-founder and chief technology officer at OpenAI. “It really is the beginning of being able to talk to your computer and have it do what you ask in a competent and reliable manner.”
How it works: In an online demo for Axios, Brockman showed how Codex can be tasked with printing text, adding an emotional feel, and even sending that text as an SMS.
And after: OpenAI beta developers will be able to build applications on top of the Codex API, allowing them to “push the boundaries of what’s out there,” says Brockman.
[ad_2]
Source link