Nostr notes by Penguin Rebellion

Leaving aside debates about security: If you are using and ...

2026-10-25T16:24:02Z

Leaving aside debates about security:

If you are using and promoting the #Signal #messenger, every day: good.

Now, as a fallback, install the #DeltaChat secure messenger, and send a few test messages. You can then forget about it, and let your devices keep it updated, in the background.

When you need it, it'll likely be available, because its technical basis is different from Signal's. And this kind of redundancy is getting more and more important.

Resist the urge of dismissing fallback solutions. Don't be naive. Ask your important others to have a fallback in place. If they can tolerate a silly game on their smartphones, for years, that has 20+ trackers and was played only once: they can live with another privacy-focused messenger app sitting there, for sure.<li><a href="https://delta.chat/en/"; target="_blank" rel="nofollow noopener" translate="no">https://delta.chat/en/</a></li><li><a href="https://signal.org/#signal"; target="_blank" rel="nofollow noopener" translate="no">https://signal.org/#signal</a></li>

#privacy #security

Wednesday was a sad day, not only for #privacy, but also for the ...

2026-10-23T19:31:04Z

Wednesday was a sad day, not only for #privacy, but also for the #EU. Brussels may be used already to be despised by the far right, but that did not seem to be enough. They are working hard to earn the contempt of as many potential supporters as possible.

»On Wednesday, #EU Member States approved a change to the #GDPR that would put the profit-making interests of #OpenAI, #Anthropic, #Google, #Meta and #SpaceX over European's fundamental right to data protection.

**According to the approved text, processing personal data “in the context of AI” should automatically be lawful. In particular, Big Tech would be able to use all personal data collected over the past decades with virtually no restrictions, as long as this happens “in the context” of AI.**« (emphasis added)

https://noyb.eu/en/ai-eu-member-states-approve-digital-expropriation-europeans-ai-companies

" #AI "

Remember this when somebody "promises" you to ...

2026-10-14T09:30:09Z

Remember this when somebody "promises" you to "evaluate" whether opening a flood gate was a good idea.

»Linux kernel documentation will stop requiring an Assisted-by tag on patches written with help from large language models, following a strong consensus at the 2026 Maintainers Summit that the tag no longer does the job it was kept for. […]

**By the 7.3 cycle, only about 10% of non-merge commits carried the tag even though LLM use is far more common. Some contributors never learned the rule, others omit it deliberately, and few maintainers chase missing tags. Absence therefore says nothing about how a patch was made, and for developers who lean on the tools constantly, presence says little more.**«

https://freenode.net/article/kernel-to-drop-assisted-by-llm-tag-after-summit-consensus

" #AI " #Linux #kernel #slop

The purpose of state #surveillance is not to target you now, ...

2026-10-13T09:46:26Z

The purpose of state #surveillance is not to target you now, specifically. It is to stay capable of targeting you later, maybe.

It is naive to joke about powerful intelligence agencies "for sure" not being interested in an average human being like you.

Because they are. They are interested in establishing the dull baseline, which you may be part of today, which helps to automatically detect the outliers.

For that, it's not exceptional "lawful access" that is required, heavily constrained by regulation. You need an ongoing, massive stream of behavioral data that gets collected and analyzed easily, en passant, because everybody's got "nothing to hide".

You will notice that only when you feel the time has come to resist. Because by then, you will become the outlier, and the patterns extracted from the behaviors of your former, careless self will help to put the spotlight on you, and to hand you over to the butcher.

#privacy

Sadly, an article by @npub1hxa…caxd that does not only contain ...

2026-10-01T08:20:07Z

Sadly, an article by Freedom of the Press (npub1hxa…caxd) that does not only contain *bad* advice for the average person, but also *dangerous* advice for journalists.

An organization claiming it's giving »security tips that keep you, your sources, and your devices safe« should never, ever recommend you use " #AI " agents, only cautioned by "guidelines" like »Don’t be the lab rat« and »Let’s talk about privacy-preserving chatbots.«

Instead:<li>Not only Do not use "AI" agents.</li><li>Also, Never share your credentials with people who do (or with anybody else, which should go without saying.)</li><li>And if you are in control of who gets access to your data: Never allow such people to obtain credentials</li>

https://freedom.press/digisec/blog/metas-muse-ai-surprises-users-but-not-in-a-good-way/

Don't even think about debating "success rates" here, ...

2026-09-29T12:20:28Z

Don't even think about debating "success rates" here, conditionally agreeing to such ubiquitous #surveillance

»Everything depends on the police watchlists; if you’re not on a watchlist, the live FRT cameras do not ‘see’ you«

That's a dangerous lie:<li>The cameras do see you, and the biometrics of your face get processed. Half a million faces were scanned, just like that. Period.</li><li>The software questions neither the watchlists, nor who gets added to them, nor whether a fascist regime has taken over the infrastructure, jumping for joy. It simply processes watchlists. It is only following orders…</li>

https://www.theguardian.com/technology/2026/sep/29/trial-live-facial-recognition-cameras-london-stations-false-positive

#surveillance " #AI "

@npub1fh7…ee83 have just passed, with a considerable majority, ...

2026-07-22T10:24:43Z

Codeberg (npub1fh7…ee83) have just passed, with a considerable majority, [this modification of their Terms of Use](https://codeberg.org/Codeberg/org/pulls/1253/files ):

»7. **You must not share projects that mostly consist of code written by "generative AI"-tools** (including services such as *Claude*, *OpenAI Codex*). Such projects having an unclear copyright status (see requirements § 2 (1) 1 and § 2 (1) 3) and furthermore have little safeguards to ensure that they do not include harmful code (c.f. § 2 (1) 5).« (emphasis added)

We welcome the change. We'd welcome it even more if that tiny word "mostly" wasn't watering it down so badly, leading to foreseeable conflicts.

We're not worried about to which extent it can be enforced. To any extent it's a net positive. People demanding perfection in every step of the combat are usually combatants *on the other side*, following the [Simple Sabotage Field Manual](https://de.wikipedia.org/wiki/Simple_Sabotage_Field_Manual ).

#Codeberg " #AI"

#Hackergal Ambassador Neetya Pujara, a Grade 10 student from ...

2026-05-31T12:51:25Z

#Hackergal Ambassador Neetya Pujara, a Grade 10 student from Aurora, Ontario:

»During one session, presenters displayed images of male tech CEOs, and everyone immediately recognized them. When the slide shifted to female tech CEOs, none of us could name them. That moment struck me deeply.

I didn’t know Melanie Perkins was the founder and CEO of Canva, that Ada Lovelace was the world’s first programmer, or that Hedy Lamarr created the foundation for WiFi. This realization pushed me to consciously learn about the accomplishments of women in #STEAM and inspired me to encourage other girls to recognize the impact women have had in technology throughout history.«

https://hackergal.org/im-a-hackergal-ambassador-heres-how-it-changed-my-life/

Today, the regime of #Iran has ended its longest #internet ...

2026-05-26T19:17:20Z

Today, the regime of #Iran has ended its longest #internet #censorship shutdown.

Not only because they were shooting their own economy in the foot, but also because it encouraged people to familiarize themselves with ways to bypass it:

»Iran’s communications minister, Sattar Hashemi, hailed the government announcement on Tuesday that the blackout was being brought to an end but said restoring the internet would be a gradual process.

He said: “Internet restrictions in recent months have caused significant damage to the digital economy, online businesses and the country’s service industries. The continuation of this situation could, in addition to the economic damage, have led to the weakening of investment, the emigration of elite human resources, and the expansion of communication patterns outside the framework of the country’s official governance.”«

https://www.theguardian.com/world/2026/may/26/iran-internet-blackout

Excellent article by @npub1l76…cve4 over at 404 Media, who ...

2026-05-04T18:43:35Z

Excellent article by Emanuel Maiberg (npub1l76…cve4) over at 404 Media, who looked into why a meta study paper has been retracted that concluded #ChatGPT had beneficial effects on #learning:

»“This meta analysis on #ChatGPT effects on learning appeared only two and a half years after ChatGPT was launched,” Williamson said. “So what we are supposed to believe is that in the intervening period, dozens of high quality studies of the effect of ChatGPT on learning performance took place, were written up, submitted for peer review, and published, which the meta analysis authors then painstakingly synthesized using robust methods.

**What appeared actually to be the case is that the meta analysis aggregated a whole bunch of very low quality research published in disreputable journals**.« (emphasis added)

https://www.404media.co/nature-retracts-paper-on-the-benefits-of-chatgpt-in-education/

"#AI"

»When Peter #Thiel said, “I no longer believe that #freedom ...

2026-04-20T15:39:00Z

»When Peter #Thiel said, “I no longer believe that #freedom and #democracy are compatible,” he wasn’t talking about your freedom. He was talking about his own. You don’t exist.«

https://www.theatlantic.com/magazine/2026/05/billionaire-consequence-free-reality/686588/?gift=7cjjyPZIuS21UKXXq4Z-JNg5WJmVwq9FbLl-9yvqo1Y

A wonderful text by @npub17ux…329m "#AI" #EchoChamber ...

2026-04-14T20:02:35Z

A wonderful text by Esther Payne :bisexual_flag: (npub17ux…329m)

"#AI" #EchoChamber #GreekMythology

»We will not help you steal others work, or enable genocides by using the companies that provide the infrastructure for governments to do harms to their peoples. We won't be polite to the people that allow the sloppy auto correct to cause extra work in an already overloaded technical community.

We will judge you. Which is what you fear really. Being an outcast and not being welcome in our spaces. The spaces we had to build for our safety.

You're not welcome here. You have plenty of other spaces online to peddle this. Leave our spaces alone.«

https://dotart.blog/cobbles/where-is-the-echo

Because it has been shown, and to keep playing the "it'll ...

2026-04-13T12:31:44Z

In reply to nevent1q…dscf
_________________________

Because it has been shown, and to keep playing the "it'll be fine, one day" card won't help.

People ignoring this and giving "guarantees" one cannot trust in are an integral part of the problem.

»LLMs can generate license-protected code without providing the necessary license information, leading to potential intellectual property violations during software production. This paper addresses the critical, yet underexplored, issue of license compliance in LLM-generated code by establishing a benchmark to evaluate the ability of LLMs to provide accurate license information for their generated code.«

https://arxiv.org/html/2408.02487v3#S7

Neither did I agree both terms were equivalent, as you are ...

2026-04-13T12:12:11Z

In reply to nevent1q…gskw
_________________________

Neither did I agree both terms were equivalent, as you are implying with "they"; nor did I make the claim you are falsely attributing to me.

Please stick with what I actually said, else you leave me no choice but to block you.

The heading approved by #Torvalds is »AI Coding Assistants«, ...

2026-04-13T11:29:06Z

In reply to nevent1q…ndl5
_________________________

The heading approved by #Torvalds is »AI Coding Assistants«, the respective document name is "coding-assistants.rst".

https://github.com/torvalds/linux/blob/master/Documentation/process/coding-assistants.rst

If you have information that says otherwise, please let me know.

New #Linux #kernel #policy: when writing code, humans can be ...

2026-04-13T10:16:12Z

New #Linux #kernel #policy: when writing code, humans can be "assisted" by "#AI", but they have to disclose it, and take full responsibility, as contributors.

While this gets celebrated as a "pragmatic stance", it simply delegates responsibilities to individual contributors that no one in good conscience can reasonably take.

Would you be willing to guarantee, legally binding, with all consequences, that your "AI" "assistant" didn't copy-paste code that's under an incompatible license? Or even proprietary, stolen one?

This is a cop out, not a responsible policy. Basically the dirty #subcontractor pattern: Everybody knows that nobody can actually guarantee what they're promising, but hey, *wink* *wink* here's their signature, they "promised" it *wink* *wink*

https://www.tomshardware.com/software/linux/linux-lays-down-the-law-on-ai-generated-code-yes-to-copilot-no-to-ai-slop-and-humans-take-the-fall-for-mistakes-after-months-of-fierce-debate-torvalds-and-maintainers-come-to-an-agreement

"Compress the kill cycle with Red Hat Device Edge" In ...

2026-04-10T07:56:43Z

"Compress the kill cycle with Red Hat Device Edge"

In case you're wondering why you can't find this #whitepaper by #RedHat anymore: they've purged it from their website, and made everybody who hosted a copy do the same.

Apart from the #InternetArchive, that is. (So far.)

While you still can, have a look at how proud one of your favorite #Linux #distro providers is about helping to speed up killing people via warplanes and drones:

»Solutions for emerging challenges
Accelerating the sensor-to-shooter cycle«

https://web.archive.org/web/20260000000000*/https://www.redhat.com/rhdc/managed-files/ve-compress-the-kill-cycle-detail-693397pr-202402-en_3.pdf

It is useful to keep in mind there can never be any kind of ...

2026-04-07T07:43:49Z

It is useful to keep in mind there can never be any kind of actual end-to-end encryption (#E2EE).

Your understanding of "end" is probably deliberately ignorant.

While you understandably so wish to believe the "end" was in the respective minds of the conversation partners, it is factually and technically located in the hardware and software of the devices used.

Which opens up a myriad of surveillance options, from sinister keyloggers; to simply glancing at your screens; to "AI" "assistants" "summarizing" what you're receiving, and allegedly just "fleshing out" what you're sending.

Even if conversation data was transmitted, "end-to-end encrypted", between brain implants, the producers or vendors of such implants would still happily provide backdoors for "law enforcement".

**No, Mozilla is not rolling out a "free #VPN" with ...

2026-03-31T12:03:00Z

**No, Mozilla is not rolling out a "free #VPN" with #Firefox.**

It's merely a #proxy, traffic is encrypted via TLS only, and you need a user account.

Since there is a 50GB per-month limit, it follows some amount of traffic logging and accounting must take place on Mozillas servers, as well.

Which begs the question: What data exactly is collected, and may fall under US jurisdiction, potentially to be handed over to US law enforcement, without you even being notified, due to gag orders?

This may still be better than nothing, but it's no alternative to a real VPN, and also deliberately misleading marketing.

#Media and #journalists should not simply adopt such wording without question.

https://blog.mozilla.org/en/firefox/built-in-vpn/