GPT-4 is getting turbocharged (but does it have a Hemi?)

At its first-ever developer conference yesterday, OpenAI unveiled GPT-4 Turbo, an improved version of its flagship chatbot, GPT-4.

The updated version will be rolled out to the masses over the coming weeks. Here’s what’s new:

• More up-to-date information. GPT-4 Turbo provides answers with context up to April 2023. Prior versions were cut off at September 2021.
• Longer prompts. Earlier versions limited prompts to about 3,000 words. GPT-4 Turbo accepts inputs up to 300 pages long, an exciting update for those still writing book reports (SparkNotes has been realllll quiet since this dropped).
• Cheaper prices and more options for developers. GPT-4 Turbo is slated to be 3x cheaper than the previous version. And OpenAI now offers access to APIs for GPT-4 Turbo with vision (which can analyze images and use them in conversations) and DALL-E 3 (which can generate images), as well as its in-house text-to-speech model.

*Billy Mays voice* But wait – there’s more! OpenAI also announced Copyright Shield, which basically means it’ll cover all the legal costs if any of its paying customers are sued for copyright infringement. And, on a separate note, the company is rolling out the ability to make custom versions of its AI chatbot, called GPTs, and will create a marketplace for people to sell them.

🔢 By the numbers… ChatGPT has about 100 million weekly active users, OpenAI said Monday. More than 92% of Fortune 500 companies are also using the platform, up from 80% in August.