@burnitdown@beige.party Exactly!

"I would never suggest that LLMs can be used to generate documentation beyond carefully delimited scenarios, such as API docs or code snippets, that is, cases where documentation is produced from clear and concise sources, in almost programmatic ways. Echoing the thoughts in Get the hell out of the LLM as soon as possible, you should limit LLM output to tasks where they truly excel.

On the other hand, when you let LLMs generate entire documentation sets, problems arise. “Oh, you can’t do serious programming, but you can still use LLMs for docs” is a naïve, reckless statement that not only overlooks what goes into crafting excellent documentation but also shows disregard for user experience (because docs are a product, too)."

https://passo.uno/whats-wrong-ai-generated-docs/

@MichaelM@sciences.social Countries shouldn’t act as nanny states. Education is something to be left to the parents and to the school.

@demiguru@fosstodon.org I’m a Facebook user, but I only use the web site’s desktop version and I also have uBlock Origin and Privacy Badger installed. I don’t think that I exactly gave away my online privacy. Even if I gave it away, I could always recover it.

@demiguru Sure, you can go ahead and live on your own secret desert island. Meanwhile, people with deep connections with other living human beings need to rely on Facebook to maintain communication with each other. There’s no way to escape. Social ties are just too strong to be cut out all of a sudden. What we need is #REGULATION, and not individual rebellious gestures.

https://www.youtube.com/watch?v=z6kK7rIbyB0

“AI companies are not pitching a future of AI-enabled centaurs. They’re colluding with bosses to build a world of AI-shackled reverse centaurs. Some people are using AI tools (often standalone tools derived from open models, running on their own computers) to do some fun and exciting centaur stuff. But for the AI companies, these centaurs are a bug, not a feature – and they’re the kind of bug that’s far easier to spot and crush than the bugs that AI code-bots churn out in volumes no human can catalog, let alone understand.”

@WhatGodIsMadeOf@feddit.org I beg do disagree. That’s, as a matter of fact, a very self-fulfilling prophecy.

https://garymarcus.substack.com/p/breaking-news-ai-coding-may-not-be

@NebulaTide@bsd.cafe Yes, but that’s in the U.S., only - at least as of now.

Privacy’s Defender: My Thirty-Year Fight Against Digital Surveillance
by Cindy Cohn

"A personal chronicle of three key legal privacy battles that have defined the digital age and shaped the internet as we know it.

From the very beginning, Cindy Cohn was driven by a fundamental question: can we still have private conversations if we live our lives online? Privacy’s Defender chronicles her thirty-year battle to protect our right to digital privacy and shows just how central this right is to all our other rights, including our ability to organize and make change in the world.

Shattering the hypermasculine myth that our digital reality was solely the work of a handful of charismatic tech founders, the author weaves her own personal story with the history of Crypto Wars, FBI gag orders, and the post-9/11 surveillance state. She describes how she became a seasoned leader in the early digital rights movement, as well as how this work serendipitously helped her discover her birth parents and find her life partner. Along the way, she also details the development of the Electronic Frontier Foundation, which she grew from a ragtag group of lawyers and hackers into one of the most powerful digital rights organizations in the world.

Part memoir and part legal history for the general reader, the book is a compelling testament to just how hard-won the privacy rights we now enjoy as tech users are, but also how crucial these rights are in our efforts to combat authoritarianism, grow democracy, and strengthen human rights."

https://mitpress.mit.edu/9780262051248/privacys-defender/

"On its face, that might sound not altogether different from Google Photos, which similarly might suggest AI tweaks to your images after you opt into Google Gemini. But unlike Google, which explicitly states that it does not train generative AI models with personal data gleaned from Google Photos, Meta’s current AI usage terms, which have been in place since June 23, 2024, do not provide any clarity as to whether unpublished photos accessed through “cloud processing” are exempt from being used as training data — and Meta would not clear that up for us going forward.

And while Daniels and Cubeta tell The Verge that opting in only gives Meta permission to retrieve 30 days worth of your unpublished camera roll at a time, it appears that Meta is retaining some data longer than that. “Camera roll suggestions based on themes, such as pets, weddings and graduations, may include media that is older than 30 days,” Meta writes.

Thankfully, Facebook users do have an option to turn off camera roll cloud processing in their settings, which, once activated, will also start removing unpublished photos from the cloud after 30 days.

The feature suggests a new incursion into our previously private data, one that bypasses the point of friction known as conscientiously deciding to post a photo for public consumption. And according to Reddit posts found by TechCrunch, Meta’s already offering AI restyling suggestions on previously-uploaded photos, even if users hadn’t been aware of the feature: one user reported that Facebook had Studio Ghiblified her wedding photos without her knowledge."

https://www.theverge.com/meta/694685/meta-ai-camera-roll

"Design Patterns for Securing LLM Agents against Prompt Injections (2025) by Luca Beurer-Kellner, Beat Buesser, Ana-Maria Creţu, Edoardo Debenedetti, Daniel Dobos, Daniel Fabian, Marc Fischer, David Froelicher, Kathrin Grosse, Daniel Naeff, Ezinwanne Ozoani, Andrew Paverd, Florian Tramèr, and Václav Volhejn.

I’m so excited to see papers like this starting to appear. I wrote about Google DeepMind’s Defeating Prompt Injections by Design paper (aka the CaMeL paper) back in April, which was the first paper I’d seen that proposed a credible solution to some of the challenges posed by prompt injection against tool-using LLM systems (often referred to as “agents”).

This new paper provides a robust explanation of prompt injection, then proposes six design patterns to help protect against it, including the pattern proposed by the CaMeL paper."

https://simonwillison.net/2025/Jun/13/prompt-injection-design-patterns/

@Schwim@lemmy.zip I don’t think it’s so obvious as that. At least AI companies are giving users the option of deleting their data. And they also allow users to make use of their services very much for free. Copyright companies don’t care about that. They want total control so that every online use of the works they own must be licensed. They want everyone to pay a rent. The ideia that they value art, culture, knowledge or public enlightenment is just bullshit. Even more so for media companies such as The New York Times which are always stating that they’re essential to Democracy.

"Amp is an agentic coding tool built by Sourcegraph. It runs in VS Code (and compatible forks like Cursor, Windsurf, and VSCodium) and as a command-line tool. It’s also multiplayer — you can share threads and collaborate with your team.

Principles

• Amp is unconstrained in token usage (and therefore cost). Our sole incentive is to make it valuable, not to match the cost of a subscription.

• No model selector, always the best models. You don’t pick models, we do. Instead of offering selectors and checkboxes and building for the lowest common denominator, Amp is built to use the full capabilities of the best models.

• We assume that you want to access the raw power these models have to offer.

• Built to change. Products that are overfit on the capabilities of today’s models will be obsolete in a matter of months."

"Israel’s tech world is small, and Israel leads the world in exports of spyware and digital forensics tools. Former Israeli Defense Forces intelligence officials have gone on to found tech companies, especially in the field of spyware and surveillance. Companies like NSO Group, maker of the controversial Pegasus spyware, were founded and staffed by Israeli officials. The company was sanctioned by the Biden Administration. Israeli spyware firms have become known for selling their products to oppressive regimes around the world, where the tech is often used to target human rights advocates, journalists, and dissidents.

TeleMessage was founded in 1999, the same year its founder left his Israeli military intelligence job. The company found early success in the burgeoning world of text message technology. The company grew and eventually pivoted to focusing on extracting messages from common communication platforms for the purposes of archiving, often to meet record keeping regulations.

Tech professionals have moved between companies like TeleMessage and some of the leading Israeli spyware firms. For example, Alon Falah, a technical support manager at TeleMessage until 2021, left the company to join NSO Group, according to his LinkedIn profile. Another employee, Itzhak Demoza, joined Telemessage last year after a stint at Cellebrite, maker of hardware and software widely used by law enforcement to extract data from smartphones."

"Unknown hackers last month targeted leaders of the exiled Uyghur community in a campaign involving Windows spyware, researchers revealed Monday.

Citizen Lab, a digital rights research group based at the University of Toronto, detailed an espionage campaign against members of the World Uyghur Congress (WUC), an organization that represents the Muslim-minority group, which has for years faced repression, discrimination, surveillance, and hacking from China’s government."