Dec 23, 2023 10:25 PM
(This post was last modified: Dec 23, 2023 10:37 PM by C C.)
Robots versus humans: Which would children trust more when learning new information?
https://www.eurekalert.org/news-releases/1029814
EXCERPTS: In this digital age, children are exposed to overwhelming amounts of information online, some of it unverified and increasingly generated by non-human sources, such as AI-driven language models. As children grow older, the ability to assess a source’s reliability is an important skill in cultivating critical thinking.
Children aged three to five years display selective trust based on the informant’s past accuracy when faced with both humans and robots, according to a study published in the journal Child Development titled, ‘Younger, not older, children trust an inaccurate human informant more than an inaccurate robot informant.’
“Children do not just trust anyone to teach them labels, they trust those who were reliable in the past. We believe that this selectivity in social learning reflects young children’s emerging understanding of what makes a good (reliable) source of information,” explained Li Xiaoqian, a research scholar at Singapore University of Technology and Design (SUTD) who co-authored the study with her PhD supervisor Professor Yow Wei Quin, a psychology professor and head of Humanities Arts and Social Sciences cluster at SUTD. “The question at stake is how young children use their intelligence to decide when to learn and whom to trust.”
[...] The study revealed that children were willing to accept new information from both human and robot informants who had previously given accurate information, but not from a potentially unreliable informant who had made mistakes in the past—especially when the informant was a robot. As for the age effect, the authors reported that younger children were likelier to accept information from an unreliable human than an unreliable robot, but older children were found to distrust or reject information from an unreliable informant, human or robot... (MORE - missing details, no ads)
Will superintelligent AI sneak up on us? New study offers reassurance
https://www.nature.com/articles/d41586-023-04094-z
EXCERPTS: Will an artificial intelligence (AI) superintelligence appear suddenly, or will scientists see it coming, and have a chance to warn the world? That’s a question that has received a lot of attention recently, with the rise of large language models, such as ChatGPT, which have achieved vast new abilities as their size has grown.
Some findings point to “emergence”, a phenomenon in which AI models gain intelligence in a sharp and unpredictable way. But a recent study calls these cases “mirages” — artefacts arising from how the systems are tested — and suggests that innovative abilities instead build more gradually.
“I think they did a good job of saying ‘nothing magical has happened’,” says Deborah Raji, a computer scientist at the Mozilla Foundation who studies the auditing of artificial intelligence. It’s “a really good, solid, measurement-based critique.”
The work was presented last week at the NeurIPS machine-learning conference in New Orleans. [...] The new research tested claims of emergence in several ways.
[...] Study co-author Sanmi Koyejo, a computer scientist at Stanford University in Palo Alto, California, says that it wasn’t unreasonable for people to accept the idea of emergence, given that some systems exhibit abrupt “phase changes”. He also notes that the study can’t completely rule it out in large language models — let alone in future systems — but adds that "scientific study to date strongly suggests most aspects of language models are indeed predictable”... (MORE - missing details)
https://www.eurekalert.org/news-releases/1029814
EXCERPTS: In this digital age, children are exposed to overwhelming amounts of information online, some of it unverified and increasingly generated by non-human sources, such as AI-driven language models. As children grow older, the ability to assess a source’s reliability is an important skill in cultivating critical thinking.
Children aged three to five years display selective trust based on the informant’s past accuracy when faced with both humans and robots, according to a study published in the journal Child Development titled, ‘Younger, not older, children trust an inaccurate human informant more than an inaccurate robot informant.’
“Children do not just trust anyone to teach them labels, they trust those who were reliable in the past. We believe that this selectivity in social learning reflects young children’s emerging understanding of what makes a good (reliable) source of information,” explained Li Xiaoqian, a research scholar at Singapore University of Technology and Design (SUTD) who co-authored the study with her PhD supervisor Professor Yow Wei Quin, a psychology professor and head of Humanities Arts and Social Sciences cluster at SUTD. “The question at stake is how young children use their intelligence to decide when to learn and whom to trust.”
[...] The study revealed that children were willing to accept new information from both human and robot informants who had previously given accurate information, but not from a potentially unreliable informant who had made mistakes in the past—especially when the informant was a robot. As for the age effect, the authors reported that younger children were likelier to accept information from an unreliable human than an unreliable robot, but older children were found to distrust or reject information from an unreliable informant, human or robot... (MORE - missing details, no ads)
Will superintelligent AI sneak up on us? New study offers reassurance
https://www.nature.com/articles/d41586-023-04094-z
EXCERPTS: Will an artificial intelligence (AI) superintelligence appear suddenly, or will scientists see it coming, and have a chance to warn the world? That’s a question that has received a lot of attention recently, with the rise of large language models, such as ChatGPT, which have achieved vast new abilities as their size has grown.
Some findings point to “emergence”, a phenomenon in which AI models gain intelligence in a sharp and unpredictable way. But a recent study calls these cases “mirages” — artefacts arising from how the systems are tested — and suggests that innovative abilities instead build more gradually.
“I think they did a good job of saying ‘nothing magical has happened’,” says Deborah Raji, a computer scientist at the Mozilla Foundation who studies the auditing of artificial intelligence. It’s “a really good, solid, measurement-based critique.”
The work was presented last week at the NeurIPS machine-learning conference in New Orleans. [...] The new research tested claims of emergence in several ways.
[...] Study co-author Sanmi Koyejo, a computer scientist at Stanford University in Palo Alto, California, says that it wasn’t unreasonable for people to accept the idea of emergence, given that some systems exhibit abrupt “phase changes”. He also notes that the study can’t completely rule it out in large language models — let alone in future systems — but adds that "scientific study to date strongly suggests most aspects of language models are indeed predictable”... (MORE - missing details)