Sam Altman, CEO of OpenAI, attends the 54th annual assembly of the World Financial Discussion board in Davos, Switzerland, on Jan. 18, 2024.
Denis Balibouse | Reuters
OpenAI mentioned Monday that it is partnering with Frequent Sense Media on an initiative designed to assist teenagers perceive the best way to use synthetic intelligence in a secure method.
“We wish to work out the best way to make this instrument safely and responsibly and broadly accessible to teenagers and people who find themselves going to make use of it as a part of their instructional expertise,” OpenAI CEO Sam Altman mentioned at a Frequent Sense Occasion in San Francisco.
Frequent Sense, a nonprofit targeted on making expertise secure and accessible to youngsters, has been working to develop an AI rankings and evaluation system meant for folks, kids and educators to raised perceive the expertise’s dangers and advantages. Among the questions Frequent Sense desires to reply embody whether or not AI fosters a love of studying amongst youth, if it respects human rights and kids’s rights, and if the expertise can perpetuate the unfold of misinformation.
The objective of the brand new partnership is to assist create AI tips and training supplies for kids, educators and fogeys and to assist curate “family-friendly” GPT-branded massive language fashions (LLMs) that adhere to Frequent Sense’s score and requirements. GPT is the spine of OpenAI’s ChatGPT chatbot, which was launched in late 2022.
Frequent Sense Media CEO Jim Steyer mentioned in an announcement that the supplies developed by means of the partnership will “will probably be designed to teach households and educators about secure, accountable use of ChatGPT, in order that we are able to collectively keep away from any unintended penalties of this rising expertise.”
On the occasion on Monday, Altman briefly spoke concerning the partnership and AI extra broadly, saying that he hopes it’ll “profit youngsters with out entry” to AI. A part of OpenAI’s mission is to “make actually useful AI accessible without spending a dime,” he mentioned.
In September, Craig Newmark Philanthropies, the Craigslist founder’s philanthropic arm, mentioned it contributed $3 million to assist fund a Frequent Sense synthetic intelligence and training initiative. Newmark informed CNBC on the time that a few of his issues about AI embody the likelihood that unhealthy actors can use the expertise to affect the data ecosystem and contribute to societal discontent.
OpenAI and Frequent Sense did not say how LLMs will probably be tweaked to assist assist educators or teenagers. Altman mentioned LLMs personalized for instructional functions may assist teenagers “who wish to study science or study biology.”
“I do not assume we all know but precisely how individuals are going to wish to use it,” Altman mentioned. He added that he envisions a world wherein “each teen or each grownup goes to have a customized AI.”
Relating to the upcoming elections and the potential dangers posed by so-called deepfakes to confuse folks, Altman acknowledged that AI-generated photographs pose issues however mentioned “I feel individuals are rather more refined than we give them credit score for and you do not imagine each picture you see.”
He talked about how OpenAI is getting ready for the potential methods unhealthy actors may use AI.
“We have arrange a giant response effort,” he mentioned. “This will probably be monitored very carefully.”
WATCH: Microsoft-OpenAI created the best hype marketing campaign in tech historical past
