Тоталитуризм

Как нам тут одно время рекламировали туры по Северной Корее, так и мы теоретически могли бы показать, как "весело" живётся в условиях, когда пол-интернета перебанено. А это вполне себе новые впечатления!

Вдохновлено тредом @a1ba: suya.place/objects/664d04d7-4d…

#словотворчество

¿Ya has podido explorar la base de datos con algoritmos y sistemas de IA en la sanidad pública que acabamos de publicar?

Es la primera vez que esta información está disponible, y además en un único buscador en el que puedes filtrar por comunidad autónoma, especialidad o naturaleza de su creación (pública, privada, mixta o desconocida).

Échale un vistazo para conocer qué algoritmos hay en la sanidad de tu comunidad y en qué fase de implantación están:

🔗 civio.es/sanidad/2025/11/06/al…

El aterrizaje de la IA en la sanidad pública: opaco, desigual y con escasa supervisión

Todas las comunidades autónomas reconocen haber desarrollado algoritmos y sistemas de IA en los servicios de salud, pero la mayoría no da suficiente información para evaluar su despliegue.

^Civio

Thanks to WINE, Proton, and Valve's hardware, a significant milestone has been achieved with 90% of Windows games running on Linux, making nine out of ten Windows games accessible on a new Linux installation. Users are flocking to other platforms during the transition from the Windows 10 to 11 eras. Thank you Microsoft once again for enabling the age of Linux gaming, you are our biggest marketing partner 😉

boilingsteam.com/windows-games…

So, @anildash wrote an interesting reflection on "AI" and Firefox a few days ago: anildash.com/2025/11/14/wantin…

He argues that the number of people who want Mozilla to just stop with the AI development and focus on a more traditional browser is small (probably true) and that people are so used to using AI in their everyday lives that it's Firefox's/Mozilla's job to make that as secure and "less big tech dependy" as possible.

I think that's not an unreasonable argument. I do think that the actual question is more about _what Mozilla is for_ and not "AI" (or other tech hypes).

From my reading the people who don't want AI in Firefox are often AI critical, sure. But it's also about resources and narrative. The idea of Mozilla was to have something that would work for the good of the open web, that would fight for users through participating in standards development but that would also argue based on what is right. Mozilla's sales pitch was a moral one - at least that is how many in the community on Mastodon for example interpreted it.

So when Mozilla cuts down on policy work, cuts work on technologies like Servo or Rust that were supposed to materially improve the security of browsers and people online while setting a lot of developer hours on fire in order to integrate fundamentally insecure (and some would say fundamentally anti-"open web") systems "just because people use them", it feels like an organization having lost their mission or the drive to push their values.

I think "AI" is just the latest (and probably biggest) event that illustrates a sentiment that has been brewing for a while: That Mozilla's mission or goals have shifted in a way that their original supporters no longer feel aligned with.

I know you don’t want them to want AI, but… - Anil Dash

A blog about making culture. Since 1999.

^{Anil Dash}

So if I'm reading this right: microsoft.com/en-us/security/b…

(If one forgives the sampling limitations (one question, 100 positive responses, etc.), which I'm not sure you should...)

The output of a language model is predictable enough that, even with simple ML models, if you

• use an LLM to generate prompts about a topic x and not about a topic x
• record the timestamps of when the LLM responded with the next chunk of tokens for each generated prompt
• train the ML model to classify "known to be/not be about x" from the timestamps

then you can get effectively perfect precision of guessing when something is about x.

Or, if Microsoft's own research is true, then the output of LLMs is not in fact a divinely powerful information oracle, but so stereotyped you can predict the topic by the timing of responses. The oracle only appears divine because we cannot see any of the other requests, and are led to believe our question is unique and personal.

Or, this is a profound self own about the use of LLMs for information, period.

​​Whisper Leak: A novel side-channel attack on remote language models | Microsoft Security Blog

Understand the risks of encrypted AI traffic exposure and explore practical steps users and cloud providers can take to stay secure. Learn more.

^{Geoff McDonald (Microsoft Security Blog)}

A já dnes začínám co? Já dnes začínám dovolenou!

Do kindle jsem stáhl Harry Pottera, v televizi nastavil Hanibala Lectera, v ruce mám kafe a jdu na to.

Slibuju, že celých 12 dnů nebudu řídit žádný autobus!

We’re back with our biggest #BlackFriday deal - save 62% on Tuta Legend!

This one’s too good to miss 👇

tuta.com/black-friday

Just want to point out that none of those important and powerful billionaires stepped up during the government shutdown to make sure people were still being fed or didn’t get evicted from their homes.

Billionaires shouldn’t exist. There is no justification for that kind of hoarding.

Today, my VPS served over 51.5 million requests. Well over 99% of that was AI crawlers and other obnoxious shits.

This is not normal. This is complete and utter bullshit. This is also happening all over the place.

It can be caught, it's not even hard. But we shouldn't need to. This is about three orders of magnitude more requests I'd normally receive, and it's almost entirely useless garbage.

Every single one of you who use GenAI tools, you personally, are complicit in this. You are responsible for these bots hammering the entire internet, you are enabling it.

If you think this price is acceptable, that every single person who hosts anything outside of BigTech walled gardens deserves this relentless assault of thieving robots, then you are a garbage human being.

But it is not too late to change course. You too can look back at the carnage you enabled, and feel remorse. It's okay. We'll forgive you.

You don't need to look at the environment damage LLMs cause - we can have an educated guess (it's very bad). You don't need to look at the unsustainability of it all. All of those are things that we don't directly feel right now.

But look at the damage these things cause to everyone outside of the BigTech walled gardens. That is measurable. These attacks are fact. You can't debate it. You can't justify it.

You, dear enabler of GenAI bullshit, you are responsible for enabling this carnage. Think about that. Feel bad about it, and stop. Today is a great day to do that.

Ein Lebenszeichen von mir. Nach meinem Start mit Ylvi am Bike bei einem kleinen Rennen in Leipzig bin ich bereits mittags zurück gefahren, da mein Bruder einen runden Geburtstag feierte. Weil ich das Rennen schon vor der Siegerehrung verlassen habe, weiß ich noch keine Zeiten, aber das ist am Ende eh egal. Es lief gut für uns am Bike. 1/2

#Zughundesport #Bikejoring #RaceDay #Canicross #Dogscooter #Dogs #EuropeanSleddog #Schlittenhunde #Hunde #Wettkampf #Sports

Here is a transcript of the Access On episode segment in which Jonathan Mosen reads the letter that Mark Riccobono, President of the National Federation of the Blind (NFB), wrote to Vispero’s new Chief Executive Officer.

Dear Ms. Bassett Spears, as you know, the National Federation of the Blind is the nation's transformative organization of blind people, with affiliates in every state as well as the District of Columbia and Puerto Rico. As the voice of the nation's blind, we have a duty to alert those who offer products and services to us when there is a groundswell of concern. This is the commitment I have also made to you in the two one-on-one conversations we have held.

Following feedback from many of our members who are raising new issues almost daily, I feel compelled to formally convey to you a series of pressing concerns regarding Vispero's direction and engagement. As the producer of JAWS, ZoomText and Fusion, Vispero impacts the daily lives and opportunities of millions of blind individuals across the United States and worldwide. With this influence comes a profound responsibility to prioritize the needs, aspirations and leadership of those you serve.

Pricing and Licensing Changes While we appreciate that Vispero must operate in a manner that ensures ongoing commercial viability, we are concerned by the limited and unclear communication around recent changes to JAWS licensing models and pricing structures. These changes have generated significant confusion, anxiety and frustration worldwide. The temporary removal of the Home Annual License option from the website, reversed only after public backlash, created unnecessary alarm. This decision led many to believe that a product they used daily was suddenly out of their financial reach. We urge Vispero to provide a clear explanation of the new pricing options and to show respect for its loyal customers by explaining clearly why these changes are necessary. Erosion of Home Annual License Functionality We are disappointed that JAWS Home Annual License holders have been excluded from the new Page Explorer feature.

Vispero highlighted this feature at our recent National Convention in July, valuing it sufficiently to produce a video demonstrating the new functionality. It was marketed at that time as an addition to FS Companion. At no time was there any mention that Vispero would create different tiers of JAWS users when it came to cutting-edge AI tools.

When the Home Annual License was introduced in 2018, the National Federation of the Blind applauded the move. We believed it worked for everyone. The unemployment rate among blind people is far too high.

The JAWS Home Annual License allows those who do not qualify for funding to use JAWS at a price they can afford. They can become familiar with the product, which in turn improves employment prospects. When an Annual Home License user finds work, Vispero benefits when an appropriate license is purchased for the blind person's workplace.

This program has been an investment in the capacity building of blind people. Depriving Annual Home License customers who have paid in good faith of new core features is a concerning new policy that changes the implicit social contract of this license. We call upon you to reconsider this decision and make Page Explorer along with future AI features available to this category of user.

Vispero Account In the United States, Jaws 2026 requires users to create and sign into a Vispero account. If it eventually becomes possible for customers to return authorizations when upgrading computers and share settings across devices, then users will benefit from this change. However, we are also hearing concerns.

Many of these may be addressed with clear, transparent, respectful communication. Some Jaws users are concerned about the amount of data being collected, such as the name of a person's employer. We seek assurance that all data Vispero is collecting is actually necessary for the operation of Jaws and that users are not being required to provide marketing data, which should be optional, in order to use their screen reader.

Additionally, there are concerns about whether a cloud outage, such as that which affected Amazon Web Services in October, may disrupt Jaws if ongoing use requires verification that users are logged in. We seek further clarity and assurance on these issues. Technical support.

We are hearing from members that Vispero is no longer accepting technical support requests via email. While we appreciate that requiring users to submit a web form may aid Vispero through the mandatory provision of information that can help the technical support process, email is a ubiquitous medium. Even those who have access to older note taker products, whose Jaws may not be working at all, can submit a written technical support request via email. (1/2)

Buenos días desde la Administración Pública.

Ejerciendo mis atribuciones según el principio de dedicación al servicio público absteniéndome no solo de conductas contrarias al mismo, sino también de cualesquiera otras que comprometan la neutralidad en el ejercicio de los servicios públicos.

Y vosotros, ¿qué tal?

Que casi se me olvida lo de la consideración con los administrados.