The AI Code Generation Battles Heat Up with Google Codey Debut

GitHub Copilot is not the only game in town

In early March, a Synthedia post identified 5 product categories already disrupted by generative AI. Topping that list is coding. Almost every software developer I know has tried at least one AI coding assistant. Some agree with the numbers from Github that the tools can reduce job effort by about 50% of the current level of effort for similar or common tasks.

Google has entered the segment with its own coding assistant, which uses the PaLM AI foundation model.

Colab will soon add AI coding features like code completions, natural language to code generation and even a code-assisting chatbot. Colab will use Codey, a family of code models built on PaLM 2, which was just announced at I/O last week. Codey was fine-tuned on a large dataset of high quality, permissively licensed code from external sources to improve performance on coding tasks.

…

We’re also bringing the helpfulness of a chatbot directly into Colab. Soon, you'll be able to ask questions directly in Colab like, "How do I import data from Google Sheets?" or "How do I filter a Pandas DataFrame?"

Codey Availability

Colab is free to use. However, it is tied to Google Cloud. Using computing resources requires a pay-as-you-go purchase of computing credits or a monthly subscription. Google said that Codey will begin rolling out in the U.S. to paying subscribers and then expand later to free users. It will roll out to other geographies in the future, but the company did not indicate a time frame.

Big Benefits

With Google’s official entrance into the now very competitive segment, the AI coding assistant market is suddenly flush with developer options. This is beneficial to users because the competition is driving innovation, and the scenario has kept them low cost or free.

Colab is Google’s IDE and is designed for Android developers. The software “is used by over seven million people for Python programming, machine learning, and data analysis, Colab coding features, including code completions and natural language to code generation.”

A Growing Field

Most industry professionals agree that Codey has a lot of catching up to do with rivals such as GitHub Copilot and Amazon’s Code Whisperer. However, Google has been deft at this catch-up process in the past. We may not know today what product will win, but conflict around the foundation model announcements seems like another arms race.

What we do hear a lot about are the search wars and generative writing solutions. Stories about AI coding assistants are far less common. However, if you want to bet about what is likely to generate the biggest long-term economic returns, code generation certainly has a good chance of becoming the leading use case.