The Wait is Over: OpenAI Reveals GPT-4 — it’s Most Powerful AI Language Model Yet
Published in
2 min readMar 14, 2023
TL;DR
- GPT-4 is a large multimodal model capable of accepting both image and text inputs and emitting text outputs, exhibiting human-level performance on various professional and academic benchmarks.
- Through 6 months of meticulous alignment and lessons learned from adversarial testing and ChatGPT, GPT-4 has better results on factuality, steerability, and staying within guardrails.
- OpenAI has rebuilt its entire deep learning stack and co-designed a supercomputer from the ground up to support GPT-4, making it their first large model whose training performance they were able to accurately predict ahead of time.
- GPT-4’s text input capability is now available via ChatGPT and the API, with image input capability set to expand in the future through collaboration with a single partner.
- To facilitate further improvements, OpenAI has open-sourced OpenAI Evals, their framework for automated evaluation of AI model performance, allowing anyone to report shortcomings in their models.
In a live demo, Greg Brockman, the president and co-founder of OpenAI, demonstrates the various capabilities of GPT-4 and how to use it.
Here’s one more demo you might wanna check out to understand how GPT 4 can be used across a wide variety of applications.
To read more:
