This repository archives a curated list of claims about GPT-3's capabilities, serving as a reference for researchers and developers exploring its potential applications and limitations. It compiles unverified use cases, critical analyses, and discussions from various online sources, providing a broad overview of the early impact and perception of GPT-3.

How It Works

The project functions as a static, community-driven collection of links and brief descriptions. It aggregates claims about GPT-3's abilities across diverse domains, from code generation and content creation to finance and therapy, alongside critical perspectives and technical explanations. The primary value lies in its comprehensive indexing of early discussions and demonstrations.

Quick Start & Requirements

This is an archived repository and does not require installation or execution. It serves as a reference document.

Highlighted Details

• Comprehensive collection of GPT-3 applications across coding, content, finance, and more.
• Includes critical perspectives on GPT-3's limitations, biases, and potential societal risks.
• Features links to foundational papers, visualizations, and expert opinions from figures like Yann LeCun and Andrew Ng.
• Documents early attempts at using GPT-3 for tasks like generating code, writing essays, and even emulating famous personalities.

Maintenance & Community

The repository is archived, indicating no active development or maintenance. The original author, swyxio, has directed users to a new repository (sw-yx/brain) for updated information.

Licensing & Compatibility

The repository itself does not contain code and is not subject to software licensing. The content is a collection of links and text, intended for informational purposes.

Limitations & Caveats

The content is explicitly stated as unverified and is now archived, meaning it reflects early-stage claims and discussions about GPT-3 that may be outdated or superseded by later developments.