Using OpenAI Codex to solve programing challenges
OpenAI’s Codex has opened the world to the general use of AI to write code, but just how good is it at coding?
What is OpenAPI Codex ?
OpenAI’s Codex is an AI system that translates natural language to code.
This means that we can supply it with a description of the intended problem to solve and the AI will genereate the corresponding code.
It is proficient in more than a dozen programming languages and powers the GitHub Copilot funcionality.
OpenAI release the public beta access to Codex with their API, you can register at: https://openai.com/join
How are we going to test it ?
I’ve recently obtain access to the public beta and have been testing Codex and it’s code generation capabilities.
According to the documentation, Codex, is more proficient on Python but can easily generate code in any programming language.
One of the ways to test the code generation capabilities is to solve coding challenges.
In this article I will go trough a series of programming challenges, ask Codex to solve them and then run and check how good is the generated code.
