You can now call 1-800-CHATGPT

In celebration of the 10th day of “ship-mas”, OpenAI has made it possible to call ChatGPT for free for up to 15 minutes at 1-800-CHATGPT.

This feature was created just a few weeks ago, Kevin Weil, OpenAI’s chief product officer, said on the livestream. Users can now call ChatGPT in the US and send messages via WhatsApp worldwide at 1-800-242-8478. The 15-minute limit applies to one phone number per month, so you can really bump up a few Google Voice numbers to get as much time as you want.

The phone number is created using OpenAI’s Realtime API and WhatsApp functionality is supported by GPT-4o mini through WhatsApp API integration.

OpenAI sees the feature as an significant step for AI newcomers because the service is a simplified version of ChatGPT compared to its web-based counterpart and offers “a low-cost way to try it out through familiar channels.” The company notes that existing users looking for more comprehensive features, higher usage limits, and personalization options should continue to operate their regular ChatGPT accounts through established channels.

Entertaining thing is, Google launched a similar tool in 2007 called GOOG-411, which offered free voice directory assistance. The service was discontinued in 2010 without any official explanation from Google, but some speculate that it was closed because it had already achieved its primary goal: collecting a sufficient database of voice samples to improve speech recognition technology.

At that time, Google vice president Marissa Mayer he said it straight: “The speech recognition experts we employ say: If we want to build a really hearty speech model, we need a lot of phonemes, that is, a syllable spoken by a specific voice with a specific intonation. That’s why we need a lot of people talking, saying different things, so that we can finally learn it. … So 1-800-GOOG-411 is about exactly that: getting a few different speech samples so that we can do it with a high degree of accuracy when you’re calling or when we’re trying to get voice from video.

OpenAI spokeswoman Taya Christianson said the company would not operate these connections to train vast language models.