Insider’s Edit: GitHub unveils Copilot X with GPT-4, plus the $600 AI chatbot

This week’s Insider’s Edit features the recent top reads on AI Business. We start with GitHub’s new supercharged Copilot X, followed by an introduction to Alpaca, the open source chatbot made for less than $600. Finally, we unveil all the insights behind OpenAI’s GPT-4!

Stay up-to-date with ChatGPT and generative AI news. Bookmark our ChatGPT page to get breaking news and insights in real time.

1. GitHub unveils Copilot X with GPT-4

GitHub has unveiled the latest version of its generative AI coding tool, Copilot X, now with GPT-4 integration and new features.

This new version adds chat and voice functionality, all built on top of OpenAI’s new GPT-4 large language model. Copilot X capabilities expand beyond code completion, with developers able to use the tool across the entire code creation lifecycle.

It includes a chat interface to GitHub’s code-editing section. The chat function recognizes what code a developer has typed in and can interact with the user on error messages. Copilot X also offers Copilot Voice, a voice-to-code extension enabling users to speak prompts in natural language.

Read in detail: GitHub Unveils Copilot X with GPT-4, New Features

2. Meet Alpaca, the $600 AI chatbot

ChatGPT could be looking at serious competition ? not from Meta, Google or even Baidu, but Stanford University.

Stanford researchers said they have created a generative AI chatbot called Alpaca that “behaves qualitatively similarly to OpenAI’s text-davinci-003 (GPT-3.5)” while being “surprisingly small and easy/cheap to reproduce.”

It cost less than $600 to develop, according to the research team.

While Stanford has pulled the demo after it began hallucinating, both the dataset used to train the model and the code used to fine-tune it remain available on GitHub.

Read on: Meet Alpaca: The Open Source ChatGPT Made for Less Than $600 and That Was Fast: Stanford Yanks Alpaca Demo for Hallucinating

3. OpenAI Unveils GPT-4

After much anticipation, OpenAI has finally unveiled GPT-4, its latest and potentially largest language model to date.

GPT-4 can accept images and text inputs to generate text outputs. According to OpenAI, GPT-4 “exhibits human-level performance,” with its creators claiming that GPT-4 passed a simulated bar exam with a score around the top 10% of test takers.

But OpenAI CEO Sam Altman cautioned that GPT-4 is “still flawed, still limited, and it still seems more impressive on first use than it does after you spend more time with it.”

Read in detail: OpenAI Unveils GPT-4 and Plans to Add it to ChatGPT+