What’s going on:
On Tuesday, OpenAI unveiled its much-anticipated GPT-4. This innovative new version of ChatGPT is capable of not only responding to text, but also processing images, making it an important tool in the escalating AI race.
This new AI boasts a 90th percentile BAR score as well as 40% increased accuracy.
Why it matters:
With its improved accuracy, knowledge base, and the ability to “see” and “reason” images, GPT-4 has the potential to take AI to the next level.
Interestingly, its release came just hours after Google announced the inclusion of generative AI features in its Workplace apps. However, GPT-4 is far from perfect and is unable to generate new ideas or accurately predict the future.
How it’ll impact the future:
GPT-4’s multi-modal capability revolutionizes the AI experience; it’s API can accept images as input, enabling it to generate captions, analyses, and reviews with 25,000 words of contextual knowledge, which is enough to comfortably fit full documents into a single prompt.
Businesses such as Stripe, Morgan Stanley, and many others are already exploring the potential of GPT-4 and are integrating it into their operations.