↩ Accueil

Vue lecture

Collective licence to ensure UK authors get paid for works used to train AI

A pioneering alternative to the opt-out system proposed by the government is supported by publishers and writers and is set to be available for use this summer

UK licensing bodies have announced a “pioneering” collective licence that will allow authors to be paid for the use of their works to train generative AI models.

The Copyright Licensing Agency (CLA) – which is directed by the Publishers’ Licensing Services (PLS) and the Authors’ Licensing and Collecting Society (ALCS), representing publishers and authors – will develop the licence, set to be the first of its kind in the UK.

Continue reading...

© Photograph: Vuk Valcic/ZUMA Press Wire/REX/Shutterstock

© Photograph: Vuk Valcic/ZUMA Press Wire/REX/Shutterstock

  •  

Gillian Anderson announces ‘even more daring’ follow-up to bestselling book of sexual fantasies

Sex Education star calls on women to send in their anonymous submissions for a second volume of her 2024 title Want

Gillian Anderson has announced a follow-up to her bestselling anthology of female sexual fantasies, Want, with the hope that it will be “more international, and even more daring”.

The original book “gave thousands of women the freedom to talk about sex without shame or judgment; to see themselves in the words of strangers, and reflect on their own desires – some for the very first time,” Anderson said. “But Want unlocked so much more for so many and felt like just the beginning of a deeper conversation.”

Continue reading...

© Photograph: Richard Phibbs/Bloomsbury

© Photograph: Richard Phibbs/Bloomsbury

  •  

Sarah Palin loses retrial of defamation case against New York Times

Jury finds newspaper not liable for allegedly defaming former Alaska governor in 2017 editorial about gun control

Sarah Palin on Tuesday lost in the retrial of her defamation case against the New York Times – a second defeat in the efforts by the former Republican vice-presidential candidate.

A federal jury in New York deliberated for two hours then found the newspaper not liable for allegedly defaming Palin in a 2017 editorial about gun control. Palin appeared dejected as she left the courthouse in Manhattan.

Continue reading...

© Photograph: Larry Neumeister/AP

© Photograph: Larry Neumeister/AP

  •  

‘I must say, mein Führer, I’m so thankful I came’: Larry David spoofs Bill Maher’s fawning White House visit with Trump

Essay describes a surprise invitation in 1939 to a previously vocal critic for dinner with the Nazi leader, where ‘suddenly he seemed so human’

Larry David has written a long spoof essay in the New York Times in response to Bill Maher’s recent glowing account of his dinner with President Trump in the White House.

The essay, entitled My Dinner With Adolf, purports to be written by someone who was “a vocal critic of his on the radio from the beginning, pretty much predicting everything he was going to do on the road to dictatorship”. But he agrees to dine with the Führer because he “concluded that hate gets us nowhere. I knew I couldn’t change his views, but we need to talk to the other side”.

Continue reading...

© Photograph: Dave Allocca/StarPix/REX/Shutterstock

© Photograph: Dave Allocca/StarPix/REX/Shutterstock

  •  

Allegations of sexual misconduct have immediate impact on perpetrator’s citations, finds study

Scientists who have been publicly accused of sexual misconduct see a significant and immediate decrease in the rate at which their work is cited, according to a study by behavioural scientists in the US. However, researchers who are publicly accused of scientific misconduct are found not to suffer the same drop in citations (PLOS One 20 e0317736). Despite its flaws, citation rates are often seen a marker of impact and quality.

The study was carried by a team led by Giulia Maimone from the University of California, Los Angeles, who collected data from the Web of Science covering 31,941 scientific publications across 18 disciplines. They then analysed the citation rates for 5888 papers authored by 30 researchers accused of either sexual or scientific misconduct, the latter including data fabrication, falsification and plagiarism.

Maimone told Physics World that they used strict selection criteria to ensure that the two groups of academics were comparable and that the accusations against them were public. This meant her team only used scholars whose misconduct allegations have been reported in the media and had “detailed accounts of the allegations online”.

Maimone’s team concluded that papers by scientists accused of sexual misconduct experienced a significant drop in citations in the three years after allegations become public compared with a “control” group of academics of a similar professional standing. Those accused of scientific fraud, meanwhile, saw no statistically significant change in the citation rates of their papers.

Further work

To further explore attitudes towards sexual and scientific misconduct, the researchers surveyed 231 non-academics and 240 academics. The non-academics considered sexual misconduct more reprehensible than scientific misconduct and more deserving of punishment, while academics claimed that they would more likely continue to cite researchers accused of sexual misconduct as compared to scientific misconduct. “Exactly the opposite of what we observe in the real data,” adds Maimone.

According to the researchers, there are two possible explanations for this discrepancy. One is that academics, according to Maimone, “overestimate their ability to disentangle the scientists from the science”. Another is that scientists are aware that they would not cite sexual harassers, but they are unwilling to admit it because they feel they should take a harsher professional approach towards scientific misconduct.

Maimone says they would now like to explore the longer-term consequences of misconduct as well as the psychological mechanisms behind the citation drop for those accused of sexual misconduct. “Do [academics] simply want to distance themselves from these allegations or are they actively trying to punish these scholars?” she asks.

The post Allegations of sexual misconduct have immediate impact on perpetrator’s citations, finds study appeared first on Physics World.

  •  

Seen a paper changed without notification? Study reveals the growing trend of ‘stealth corrections’

The integrity of science could be threatened by publishers changing scientific papers after they have been published – but without making any formal public notification.  That’s the verdict of a new study by an international team of researchers, who coin such changes “stealth corrections”. They want publishers to publicly log all changes that are made to published scientific research (Learned Publishing 38 e1660).

When corrections are made to a paper after publication, it is standard practice for a notice to be added to the article explaining what has been changed and why. This transparent record keeping is designed to retain trust in the scientific record. But last year, René Aquarius, a neurosurgery researcher at Radboud University Medical Center in the Netherlands, noticed this does not always happen.

After spotting an issue with an image in a published paper, he raised concerns with the authors, who acknowledged the concerns and stated that they were “checking the original data to figure out the problem” and would keep him updated. However, Aquarius was surprised to see that the figure had been updated a month later, but without a correction notice stating that the paper had been changed.

Teaming up with colleagues from Belgium, France, the UK and the US, Aquarius began to identify and document similar stealth corrections. They did so by recording instances that they and other “science sleuths” had already found and by searching online for for terms such as “no erratum”, “no corrigendum” and “stealth” on PubPeer – an online platform where users discuss and review scientific publications.

Sustained vigilance

The researchers define a stealth correction as at least one post-publication change being made to a scientific article that does not provide a correction note or any other indicator that the publication has been temporarily or permanently altered. The researchers identified 131 stealth corrections spread across 10 scientific publishers and in different fields of research. In 92 of the cases, the stealth correction involved a change in the content of the article, such as to figures, data or text.

The remaining unrecorded changes covered three categories: “author information” such as the addition of authors or changes in affiliation; “additional information”, including edits to ethics and conflict of interest statements; and “the record of editorial process”, for instance alterations to editor details and publication dates. “For most cases, we think that the issue was big enough to have a correction notice that informs the readers what was happening,” Aquarius says.

After the authors began drawing attention to the stealth corrections, five of the papers received an official correction notice, nine were given expressions of concern, 17 reverted to the original version and 11 were retracted. Aquarius says he believes it is “important” that reader knows what has happened to a paper “so they can make up their own mind whether they want to trust [it] or not”.

The researchers would now like to see publishers implementing online correction logs that make it impossible to change anything in a published article without it being transparently reported, however small the edit. They also say that clearer definitions and guidelines are required concerning what constitutes a correction and needs a correction notice.

“We need to have sustained vigilance in the scientific community to spot these stealth corrections and also register them publicly, for example on PubPeer,” Aquarius says.

The post Seen a paper changed without notification? Study reveals the growing trend of ‘stealth corrections’ appeared first on Physics World.

  •