OpenAI and Common Sense Media partner to protect teens from AI harms and misuse

0
115


On Monday, OpenAI announced a partnership with the nonprofit Frequent Sense Media to create AI pointers and academic supplies focused at mother and father, educators, and teenagers. It contains the curation of family-friendly GPTs in OpenAI’s GPT store. The collaboration goals to handle issues in regards to the impacts of AI on youngsters and youngsters.

Recognized for its opinions of movies and TV exhibits aimed toward mother and father searching for applicable media for his or her children to observe, Frequent Sense Media lately branched out into AI and has been reviewing AI assistants on its web site.

“AI isn’t going anyplace, so it’s vital that we assist children perceive how you can use it responsibly,” Frequent Sense Media wrote on X. “That’s why we’ve partnered with @OpenAI to assist teenagers and households safely harness the potential of AI.”

OpenAI CEO Sam Altman and Frequent Sense Media CEO James Steyer introduced the partnership onstage in San Francisco on the Common Sense Summit for America’s Kids and Families, an occasion that was well-covered by media members on the social media web site X.

For his half, Altman supplied a canned assertion within the press launch, saying, “AI affords unimaginable advantages for households and teenagers, and our partnership with Frequent Sense will additional strengthen our security work, guaranteeing that households and teenagers can use our instruments with confidence.”

The announcement feels barely non-specific within the official information launch, with Steyer providing, “Our guides and curation will likely be designed to coach households and educators about protected, accountable use of ChatGPT, in order that we are able to collectively keep away from any unintended penalties of this rising know-how.”

The partnership appears aimed largely at bringing a patina of family-friendliness to OpenAI’s GPT retailer, with essentially the most strong reveal being the aforementioned proven fact that Frequent Sense media will assist with the “curation of family-friendly GPTs within the GPT Retailer primarily based on Frequent Sense scores and requirements.”

Frequent Sense AI opinions

As talked about above, Frequent Sense Media started reviewing AI assistants on its web site late last year. This places Frequent Sense Media in an fascinating place with potential conflicts of curiosity concerning the brand new partnership with OpenAI. Nonetheless, it would not appear to be providing any favoritism to OpenAI thus far.

For instance, Frequent Sense Media’s review of ChatGPT calls the AI assistant “A strong, at occasions dangerous chatbot for folks 13+ that’s greatest used for creativity, not details.” It labels ChatGPT as being appropriate for ages 13 and up (which is in OpenAI’s Phrases of Service) and provides the OpenAI assistant three out of 5 stars. ChatGPT additionally scores a 48 p.c privateness ranking (which is oddly proven as 55 p.c on another page that goes into privateness particulars). The evaluation we cited was final up to date on October 13, 2023, as of this writing.

For reference, Google Bard gets a three-star overall rating and a 75 p.c privateness ranking in its Frequent Sense Media evaluation. Stable Diffusion, the picture synthesis mannequin, nets a one-star rating with the outline, “Highly effective picture generator can unleash creativity, however is wildly unsafe and perpetuates hurt.” OpenAI’s DALL-E will get two stars and a 48 p.c privateness ranking.

The data that Frequent Sense Media contains about every AI mannequin seems comparatively correct and detailed (and the group cited an Ars Technica article as a reference in a single clarification), so that they really feel honest, even within the face of the OpenAI partnership. Given the low scores, it appears that evidently most AI fashions aren’t off to an awesome begin, however that will change. It is nonetheless early days in generative AI.





Source link