Brazilian cosmetics company O BoticĂĄrio wants to combat the misuse of AI with an initiative called Code Her. It includes a bot on X and a digital guide on reporting images manipulated by artificial intelligence.
#tech4good #womenintech #responsibleAI
www.meioemensagem.com.br/comunicacao/...
Posts by Lidiany Cerqueira, PhD
Thanks for the feedback @iankingston.com! Happy to see that CERCA is helpful to you.
WIth regard to hallucinated references, I suspect that, for many journals, the big publishers do only cursory copy-editing. I've seen errors in paper titles (not just references) that should never have passed the first stage of the submission process. But they place profit over quality. 4/4
To lighten the load, I use the open source tool Cerca (sourceforge.net/projects/cer...), by @lidianycs.bsky.social, which checks reference lists for validity. I can highly recommend it. 3/4
Hallucinated references now mean that every reference has to be checked â just in case. I have yet to see an example in the journal that I work on (not published by one of the Big 5), probably because they have robust measures in place to screen out any problems before they reach me. 2/4
The question about fake references should NOT be âWhat can be done?â
We KNOW what can be done: human peer reviewers and editors check the damn references individually, by hand.
The question should be, âWhy are so many publishers NOT doing those basic checks?â
www.nature.com/articles/d41...
I'd also reject it. It isn't worth the effort to review the rest of the paper if the references are hallucinated. Specially in a systematic mapping.
I'm reviewing a systematic mapping. Of the first 6 referenced papers, 2 seem to have been invented by an LLM. @irisvanrooij.bsky.social is right. If the authors state that they used an LLM for any purpose, then it's not worth reviewing.
We mentioned this in our EmĂlias Podcast - Giro de NotĂcias episode. This Thursday. Fernanda, my co-host for that episode, had a lot to say.
âTheyâre not required to be subject-matter experts in terms of teaching students, because our AI platform is able to provide that tutoring experience,â Price said.
As far as I know, nothing happened to them and it was treated as controversy. It's just part of the misogyny we have to face everyday while pretending that everything is alright.
Aproveitando a discussĂŁo pra divulgar a ferramenta que lancei para ajudar revisores depois que fui citada em uma referĂȘncia alucinada (nĂŁo vai funcionar em edital de concurso, mas funciona para artigos)
github.com/lidianycs/ce...
The img shows a group of brazilian male students presenting a game on a TV. The tv is showing a picture of Epstein and other people.
Few days after the International Women's Day, a group of male students at an elite engineering institute in Brazil created a game based on the Epstein case, where a 15-year-old girl has to escape from six men on an island. This is what women in tech courses have to face daily.
Caramba. InacreditĂĄvel.
The alternatives require uploading the file or installing more stuff which I wanted to avoid to keep the tool as simple as possible. Sometimes I understand why everyone uses Python.
Yesterday I was reviewing a manuscript and CERCA only found 30 of 70 references in the file. I was so bummed that I spent the whole afternoon trying to implement a new alternative to extract the references. The problem is finding a good library or model to parse a PDF. #Java #OpenSource #Programming
I advise any reviewer to check references before accepting a manuscript. You can even use CERCA to help you.
github.com/lidianycs/ce...
âI was really shocked that an article published in a Springer journal and which presumably had gone through a peer or editorial review could have such a high number of references to hallucinated articles. I can see how one or two may slip through the net, but this was a preposterous number.â
Tive o prazer de participar do podcast Fronteiras da Engenharia de Software com os professores @adolfont.github.io e Maria Claudia Emer. Foi uma Ăłtima oportunidade para falar sobre minha pesquisa de doutorado em empatia na engenharia de software.
www.youtube.com/watch?v=d6Z8...
Thousands of workers are spending hours watching footage from smart glasses. They see naked bodies and many intimate moments are captured without anyone's knowledge. I can see how empathy was ignored during the development process in many ways here.
www.svd.se/a/K8nrV4/met...
Devemos usar IA em tudo?
Comigo e com @renatagames.bsky.social
10/03 19h00
Ative o sininho e coloque na agenda!
youtube.com/live/WoER47q...
Deixo aqui o spoiler: NĂO DEVEMOS đ
Dos porquĂȘs a gente fala na terça!
Morta đ
Foi muito boa. ConteĂșdo de alto nĂvel.
Na moral: a gente Ă© muito foda. MUITO.
Ia mandar a notĂcia pra vc. Mas vc chegou primeiro đ
A guerra fria dos pandas. Imagino eles decidindo de quem vĂŁo tomar os pandas e pra onde vĂŁo enviar đ€
If it was a survey, you probably would get many answers even in English, but for interviews, I think most of them would prefer to talk in Portuguese đ«€
Clearly the dog worked at Barone Sanitation. I hope he's unionized.