«Se vende el mito de una tecnología milagrosa e increíblemente poderosa, pero esta tecnología está basada en el trabajo en negro, precarizado y explotador de millones de trabajadores manuales sin los que la IA no funcionaría».
Entrevista a @milamiceli socióloga, doctora en Ciencias de la Computación, experta en IA en el Weizenbaum Institute, investigadora en el Dair Institute, y una de las principales expertas en un campo poco conocido: el de los trabajadores de datos.
La experta del Instituto Alemán de Internet advierte de que el auge de herramientas como ChatGPT aumentará la demanda de gente que genere contenidos por poco dinero para enriquecer las bases de datos
La stupidité du monde… Les consignes incendie c'est un truc que tu dois pouvoir lire dans l'urgence, tu dois pas avoir à sortir ton smartphone et scanner un QR code pour les avoir.
Oh et si vous pensiez que ça vous donne un PDF avec les infos pour sortir du bâtiment, non ça ouvre une vidéo de 2 minutes avec des généralités… Totalement inutile.
E-Waste is the fastest growing waste stream in the world. Most of it is not recycled and what is is often done in the Global South with terrible conditions and awful pay. Big Tech wants to keep the truly dirty secret of e-waste as hidden as possible, and keep paying poor workers as little as possible.
If you want to truly understand modern technology, start with e-waste.
Apparently the arXiv* is relocating from servers at Cornell University to Google Cloud. Putting all our eggs into Google's basket seems unwise, especially at this point in history.
This jobs listing gives a number of reasons for moving to the cloud, but to my inexpert eye it looks like they could do everything they want to do while remaining self-hosted. At the relatively modest scale of the arXiv the cloud really is just someone else's computer.
It's worth noting that the arXiv used to have a network of mirrors, but they decided to shut down the last of those in September of last year. The explanation was that they were no longer worthwhile as the Cornell servers + CDN delivered better performance than the regional mirrors: info.arxiv.org/help/mirrors.ht…
The multi-site (and multi-country) backup provided by those mirrors does not appear to have been a consideration, because of course nothing bad would ever happen to an American university.
*The arXiv is a repository providing free access to pre-prints of academic papers in a range of fields across physics, maths, biology, computer science, etc. In some fields, including astronomy, it is the main way that researchers keep up to date with new publications in their area of research.
I really enjoyed c-ares as a serious project to study a few years ago. Back then it looked like it didn't support encrypted DNS, maybe thats ok, maybe I didn't understand it at the time. Any opinions on the state of encrypting DNS going forward in the future? Did it catch on? will it?
@kiyo I don't know and I don't care that much. If people want it added there it will be added. For users such as #curl, we add things like DoH ourselves anyway and it would be hard to use any such provided by c-ares because of the "different layer" it works on.
I think I thought that then too, although without having a full picture. Long live c-ares, its small enough to study entirely, and it's a serious little lib! Thanks for all the hard work!
I have forgotten more than I ever knew about C. Nevertheless I have two ? 1) why is namelen on line 104 not being set to null if it is null on the input? Perhaps malloc initializes memory but I don't recall that or perhaps there is something else of why it isn't used. I get the ptrs here confused especially since I have no clue how c-ares works.
2) really a style question. line 11. Any reason why it's own dot.h isn't last? It shouldn't matter. Perhaps one could argue it should be first
After the Code of Conduct and the Mediation process, the Core Governance document is now ready for wider feedback.
This document describes what JoinJabber does and what we aim to achieve. It also describes what our values are and a way to make proposals to the entire community.
I think completely ruling out "LLMs and other probabilistic tools" (which rules out almost any autocompletion) is ill-informed. LLMs are way overhyped, the environmental concerns are real, but it's about the same as with any other program, it is possible to use them wisely, ie, not take their output at face value. That consent part feels very weird to me. I remember Napster and singing "copying is not theft". (1/2)
Fuck OpenAI and the others, sure. Also, I understand where this stance comes from, as the current LLM bubble and the ridiculous associated claims are insufferable. But this feels like a case of throwing the baby out with the bathwater.
You asked for feedback, this was mine, I hope that does not get me excluded as I am grateful for joinjabber's existence. (2/2)
Your weekly dose of information that keeps you up to date on the latest developments in the field of technology designed to assist people with disabilities and special needs. Special Guest: Cody Mi…
Address common accessibility challenges in mobile web forms for blind users, focusing on issues like invisible fields, keyboard overlap, and focus jumping, with tips for improvement.
Jan Gampe took things to the next level by actually making this cross-stitch out of the pattern I previously posted online. The flowers really gave it an extra level of charm I think.
NVDA 2025.1 Beta 2 is now available for testing. As well as all the amazing updates in NVDA 2025.1 (from Beta 1), this new beta includes updates to some translations, as well as a minor bug fix for SAPI 5 voices using rate boost. Read the full release notes (including all the 2025.1 features & fixes) and download from: nvaccess.org/post/nvda-2025-1b…
Beta2 of NVDA 2025.1 is now available for download and testing. For anyone who is interested in trying out what the next version of NVDA has to offer before it is officially released, we welcome yo…
@JamminJerry There is an "Update channel selector" which lets you do it from within NVDA, otherwise, download the beta and install that and that will move you over to the beta / RC channel. Similiarly, install an alpha or stable version to jump to those channels.
(I see your next post saying you installed the beta, but thought I'd reply to this one for anyone seeing this)
@JamminJerry It might be because you were already on an alpha and thus already using a 2025.x build from the point of view of add-ons - so you would have needed to tell it to show you incompatible add-ons (and then you'd have to overrride that compatibility check AND hope that add-on actually did work........ installing the beta manually was definitely the easiestr for THIS particular situation)
@JamminJerry Alpha builds will keep coming up, but now targeted towards NVDA 2025.2. We're definitely keen to ensure we get 2025.1 right first though :)
Will NVDA run on the Surface 7 ARM-Based machines from Microsoft? I am obtaining one and wanted to check on this configuration. It uses the Elite Snap Dragon chip.
@JeffBishop The last time I tried it, NVDA generally had no issues, but there were some painpoints, I think, with Window's emulation layer which did not at all give it a polished experience, felt like beta at best. Note that my test of that was in the Summer (northern hemisphere) of 2024, so hopefully things have gotten better since then.
@JeffBishop Hi Jeff, as others have said, it should work fine. There were some improvements in Windows 11 24H2 which should improve the experience further. Please do reach out if anything isn't working as it should
@fireborn @kellylford @JeffBishop Aaron, we do get asked about ARM support regularly and I was thinking it would be a great opportunity to write up an article on someone using a newer surface - would you be interested in sharing your experience? I'd like to write up something like nvaccess.org/category/nvda-sto… - if so, please email me, quentin@nvaccess.org
How the TAD Foundation Is Redefining Career Pathways for Blind Students in Corporate America: Applications Now Open: Deadline To Apply: April 30, 2025 drkirkadams.com/tad-fellowship…
Discover how the 2025 TAD Fellowship is transforming career pathways for blind and low-vision students through elite mentorship, technical training, and private-sector job placement. Applications open through April 30, 2025.
I think it's relying on a voice service myself. mosto f the voice seam to come from Amazon Polly, and the Snoop Dogg voice? I dont' think it's relying on elevenlabs, I think it's relying on a different platform but I could be totally wrong.
From the Maryland State Library for the Blind and Print Disabled: TUG Meeting Announcement; Resources for Tech News, Live Assistance; Saturday, May 10, 10:00 AM Eastern Time groups.io/g/tech-vi/message/91…
From the Carroll Center for the Blind: Free Webinar: The Windows Screen Reader Primer, 4th Edition Launch; Tuesday, May 6, 7:00 PM Eastern Time groups.io/g/tech-vi/message/91…
How does this read to you? No comment! :) At the White House Easter Egg Roll, the first lady has risen - The Washington Post apple.news/AzRsE7-9QR8ugFGwi6S…
I had an epiphany last night about something in the nginx config and after sorting that, all the optional XEP listed also pass!
Now to convince wife and kids to use it for our group chat... they really don't want to install any more apps, because I'm "being difficult" because they already use whatsapp 😑
The main reason I'm impressed with it over matrix, is reliable notifications. It just works. So far, anyway.
Thank you to the people that replied to my post last week suggesting I try xmpp after the matrix server we used shut down. I had written it off as an old platform for nerds, but it's... actually really good ❤️
lmao @matrix STILL has the "Matrix 2.0 is here" (mastodon.matrix.org/@matrix/11…) post pinned to the profile even though "here" is STILL a very long stretch
It's official - Matrix 2.0 implementations are here and ready for mainstream use! 🎉🥳🎊 Come watch the keynote from The Matrix Conference and learn about the joy of instant sync ⚡️, next-gen auth 🔐, native VoIP 🎦 and Invisible Crypto 👻: matrix.org/blog/2024/10/29/mat…
I did not miss that but my initial criticism from when the blog post was published still largely holds true. From my perspective of trying to not have three different partially desynced room states on my clients MAS is probably the once thing I care the least about
well, synapse shipped matrix 2.0 sync back in Sept: github.com/element-hq/synapse/… which addresses all known issues which could cause client sync to drift apart. if you’re still having bugs let us know.
Synapse 1.114.0 (2024-09-02)
This release enables support for MSC4186 — Simplified Sliding Sync. This allows using the upcoming releases of the Element X mobile apps without having to run a Sliding...
first of all, that's great, but, there are no clients which are usable for me which implement this.
Looking at the proposal there seem to be some major questions still unaddressed so I don't entirely understand the energy in implementing and the big announcement for something that is in an unfinished state with a lot of ambiguity in it's specification.
Keepass-Tool (Format 4) "Secrets". Einfacher und weniger Möglichkeiten im Vergleich zu KeepassXC & Co., dafür hübsch und übersichtlich. Vielleicht eine Möglichkeit, Familienmitgliedern einen Passwortmanager schmackhaft zu machen ...
Verwalten Sie Ihre Passwörter – Secrets ist eine Passwortverwaltung, die das KeePass-Format in Version 4 nutzt. Die Anwendung integriert sich perfekt in die GNOME-Arbeitsumgebung und stellt eine einfache und übersichtliche Schnitts...
If someone's profile looks blank on here, don't assume it is actually blank. Some accounts, especially newer ones or ones with few followers, may look blank even if they aren't.
There's more info about this, why it happens and how to check what the profile really looks like in the guide:
Archos
in reply to meuge • • •meuge
in reply to Archos • • •Archos
in reply to meuge • • •meuge
in reply to Archos • • •