Looking for a brand new technique to work together with AI? Call or textual content 1-800-CHATGPT. No, actually. OpenAI introduced on Wednesday that customers within the US can now discuss with ChatGPT by way of telephone or ship messages by way of WhatsApp, utilizing the quantity 1-800-242-8478. The service is free, however customers solely obtain quarter-hour of name time per 30 days. Standard provider charges might apply.
The telephone service is out there to anybody within the US with a legitimate US telephone quantity. Meanwhile, ChatGPT by way of WhatsApp is accessible wherever ChatGPT is out there.
Read extra: Here’s Everything You Need to Know About Using ChatGPT
The firm introduced the newest replace throughout a livestream tied to its 12-day vacation launch spree, which beforehand included the roll out of its much-anticipated Sora video technology mannequin. On the livestream, an worker mentioned the decision and texting instrument was solely constructed a number of weeks in the past.
In an e-mail despatched to CNET, OpenAI mentioned 1-800-CHATGPT can be a part of its broader effort to make the service extra accessible.
“Consistent entry to high-speed knowledge could be tough and dear for a lot of,” OpenAI mentioned in a press release. “For these new to AI, that is a straightforward, handy, and low-cost technique to attempt it out by way of acquainted channels.”
We tried it out
CNET referred to as the quantity, and first requested concerning the climate in San Francisco. The female-sounding voice responded promptly however mentioned it lacked entry to real-time climate knowledge and really helpful utilizing an app with up-to-date info. Virtual assistants reminiscent of Apple’s Siri and Amazon’s Alexa can reply this query with ease.
A CNET author discovered blended outcomes when calling the quantity. 1-800-ChatGPT could not report on the climate, and it repeated questions it had already requested the caller.
When requested what yr the Eiffel Tower was accomplished, the voice precisely replied, “1889.” It went on to supply extra particulars concerning the construction, together with recommendation on whether or not tickets are wanted upfront to go to the highest. But the service had limitations, asking if the caller was planning a visit to Paris, receiving a constructive reply, after which afterward, repeating the identical query as if it had by no means requested.
Users must also understand that AI has been recognized to supply inaccurate info, generally known as hallucinations.
In a weblog publish, OpenAI referred to as the characteristic “experimental” with some limitations. For instance, it really works greatest in quiet environments, and the corporate recommends enabling noise isolation options in your gadget.
OpenAI mentioned it doesn’t retailer conversations however might evaluation calls, transcripts and WhatsApp messages for security and abuse prevention functions.
