Marco A. Lopez-SanchezI couldn't agree more with this answer from recent Nobel laureate Katalin Karikó, and the thing is, it's something we all know in the scientific world.👇
Baybars
- 24 Followers
- 102 Following
- 69 Posts
🙊 Speech tech 🤖 MLops 🛠️ Product
Carl T. BergstromDifficult or not, it is worth adding to your vocabulary if you care about the impacts of information technology, machine learning, artificial intelligence, and so forth. In a recent FAccT ’21 conference paper Measurement and Fairness https://dl.acm.org/doi/pdf/10.1145/3442188.3445901 authors Abigail Jacobs @az_jacobs and Hanna Wallach @hanna observe that AI systems often aim to quantify—and critically, make decisions based upon—unobservable theoretical constructs such as "teacher effectiveness" or "risk of recidivism".
Hervé Bredin"How I reached 1st place at Ego4D 2022, 1st place at Albayzin 2022, and 6th place at VoxSRC 2022 speaker diarization challenges" 😎
Short answer: with 🎹#pyannote.audio
Long answer: https://github.com/pyannote/pyannote-audio/blob/develop/tutorials/adapting_pretrained_pipeline.ipynb
Kathy Baxter@upol OpenAI is not alone in taking advantage of vulnerable populations by paying them pennies to read & label horrific content. Google, Meta, & Amazon have employed ppl in refugee camps to label their data.
#responsibleAI
https://restofworld.org/2021/refugees-machine-learning-big-tech/
#responsibleAI
https://restofworld.org/2021/refugees-machine-learning-big-tech/
Gary MarcusWow! Who could possibly have seen this coming? “CNET Is Reviewing the Accuracy of All Its AI-Written Articles After Multiple Major Corrections”
Franck LeroyI just found Mapstodon : a very nice interactive map of #Mastodon instances and their public timeline, grouped by region and interest.
dear software comrade @arda just started his self hosted writefreely instance at @[email protected]
👏👏👏
are there any other active writefreely blogs out there?
(or actually how does the discoverability work? 🤔)
Mike Masnick ✅Oh wow. I know there have been a ton of debates and discussion about quote tweet functionality and whether it's good or bad (I'm on the "I think it would be good" side), @hildabast has put together a VERY detailed review of all research on the topic. Her take based on that: it is not a major vector of abuse, as some claim. https://absolutelymaybe.plos.org/2023/01/12/quote-tweeting-over-30-studies-dispel-some-myths/
Eileen GuoLast month, I published an investigation into 15 images captured + shared by iRobot test devices from inside the homes of "paid data collectors and employees" who, the co. said, had consented to the collection.
But testers have since shared those consent agreements w/ me, calling into question what exactly they consented to, and the characterization of payment.
"Cheap bastards," as one angry tester put it, even b4 knowing "my naked ass could now be on the Internet."
Roomba testers feel misled after intimate images ended up on Facebook
An MIT Technology Review investigation recently revealed how images of a minor and a tester on the toilet ended up on social media. iRobot said it had consent to collect this kind of data from inside homes—but participants say otherwise.
Erin KissanePlenty of great people have cycled through #Medium at various points, but I would love to see a little bit of memory retention about their organizational history:
"Medium abruptly cancels the membership programs of its 21 remaining subscription publishers" https://www.niemanlab.org/2018/05/medium-abruptly-cancels-the-membership-programs-of-its-21-remaining-publisher-partners/
"‘We had no idea that it was coming’: Medium pulls the rug from under publications" https://www.cjr.org/business_of_news/medium-publication.php
"The long, complicated, and extremely frustrating history of Medium, 2012–present" https://www.niemanlab.org/2019/03/the-long-complicated-and-extremely-frustrating-history-of-medium-2012-present/
