Grindr has been caught sharing users' HIV status and precise location with advertisers without obtaining valid consent. They must immediately stop prioritizing profits over users’ safety, and make privacy the default across its platform. https://www.eff.org/deeplinks/2026/06/grindr-put-queer-safety-and-privacy-over-profits
⚠ Several critical #Guix daemon vulnerabilities and a minor vulnerability in ‘pull’ and ‘time-machine’ have been fixed: time to upgrade! ⚠
https://guix.gnu.org/blog/2026/guix-substitute-pull-vulnerabilities/
Today I learned GHC got sloppy https://git-annex.branchable.com/no_llm_code/
A politician investigating Pegasus spyware… had their phone hacked with Pegasus multiple times. The compromises came days ahead of key meetings of the spyware inquiry:
Today the good people of the world commemorate the passing of Google Reader thirteen short years ago on July 1, 2013.
To honor their memory (and in lieu of flowers) please add an RSS, Atom, JSON or equivalent feed to your web site and keep building towards a web that connects people in meaningful and positive ways.
#German Court Says #Google Is Liable For False Claims In Its #AI Overviews Because They Are Its Own Words - https://www.techdirt.com/2026/07/01/german-court-says-google-is-liable-for-false-claims-in-its-ai-overviews-because-they-are-its-own-words/ "More generally, #liability concerns could mean that many current use cases for #agents won’t be commercially viable."
The Interledger Foundation @Interledger has an open funding call for libraries implementing #RFC9421 (HTTP Signature). This is crucial infrastructure for #ActivityPub. Go get that money!
Almost exactly ten years ago, I argued that 'free as in beer' and 'free as in speech' could only coexist in the absence of software liability.
https://exple.tive.org/blarg/2016/08/29/free-as-in-health-care/
Recently a German court established that AI summaries count as the issuing company's voice, meaning companies are directly liable for what their AI summaries or overviews say:
Britain is currently running on an electricity generation mix that is over 90% renewables. The amount generated is in fact 112% of demand and the rest is being exported. That means that, were it not for the need to currently keep gas running to provide frequency and voltage stability, the country could technically be running on wind, solar and a bit of hydro.
(not yet read, but promising)
The Philosophical Prospects of Large Language Models in the Future of Mathematics
Fenner Stanley Tanswell, Ásgeir Berg
In this article, we examine the philosophical implications Large Language Models might have on mathematical practice in the near future. Some prominent researchers argue that Large Language Models will soon have the ability to generate or check proofs, lifting a great burden of human mathematicians. We claim, however, that the implementation of LLM technologies in mathematics is not merely a neutral tool that assists mathematicians to continue on as before, but instead entails a radical change to the practices of mathematics with important philosophical implications. We will argue that we cannot be confident such tools will continue to work as expected, even if they become arbitrarily more reliable than they currently are, and that the kind of justification we get from LLM-generated proofs can never be properly mathematical. We will evaluate solutions to this problem involving either computer verification or human checking and argue that these cannot fix the philosophical gap to give us proper mathematical justification.
https://www.mxphi.com/all-issues/volume-iii-number-2-2025/philosophical-prospects-llm/
RE: https://eupolicy.social/@edri/116849529482171100
As the Digital Markets Act’s shadow rapporteur, I am really glad that now, with the DMA, Europe has a much faster instrument to deal with dominant position abuse than antitrust. Now, for the sake of Europe, the @EUCommission has to use it.
Pretty chuffed to hear from the organisers that my "How not to use AI" preconference workshop helped set the tone of the following days, and that scholarly values and the benefits of slowing down were recurring conversation topics
If your conference or grad school is looking for a workshop format that avoids the tired marketing cliches and productivity narratives around "AI", I'd be happy to do a rerun (time permitting) https://ideophone.org/how-not-to-use-ai-workshop/
Kukly: Story of Putin - Little Zaches
https://redirect.invidious.io/watch?v=eZJx9bgwdv0
h/t https://redirect.invidious.io/watch?v=iqnFFvADRUY
@civodul FWIW, Godot has just adjusted its policies for LLMs being banned:
https://godotengine.org/article/contribution-policy-2026/
Given its a short description it may be interesting regarding wording.
Follow-up to my piece on the Monsanto playbook.
Most common response in the comments: "Nobody forced farmers to use the products. Farmer's choice."
Yes, the contract was signed freely. In the same way, nobody forced anyone to sign a subprime mortgage in 2006. The system that made the alternative impossible was built deliberately.
New piece covers:
The agricultural treadmill (Cochrane, 1958): early adopters gain a temporary cost advantage; commodity prices fall as adoption spreads; the advantage disappears; and survivors consolidate. Rinse, repeat. The consistent winner in this dynamic is the technology vendor.
The risk transfer: when the system fails, the input companies still profit. The banks still collect. The downside concentrates where the power doesn't.
And the third squeeze: the precision ag platform that owns your field data when you leave.
https://gaggl.com/blogs/2026-06-28-the-treadmill-nobody-chose/
#RegenerativeAgriculture #AgTech #FarmDebt #FoodSovereignty #RuralAustralia #CorporatePower #Tech4Good
> According to reporting from the Netherlands, Microsoft allegedly shared the names and internal communications of Dutch officials working on EU platform regulation with the U.S. House of Representatives, including email addresses, meeting minutes, and invitations. Those officials were tied to agencies that enforce the Digital Services Act, making the context especially sensitive because the data belonged to regulators shaping Europe’s platform rules. While the House and Microsoft refuse to comment, the issue highlights the asymmetry of digital power. A European government can think it is operating within its own administrative boundaries while its data still sits in a system accessible from Washington.
The Korean branch of Starbucks used AI to create a marketing campaign, but whoops - this campaign celebrated the May 18th massacre of Korean civilians peacefully protesting a military coup in 1980!
Nobody checked before they launched the ad campaign.
Result: massive protests and boycotts, legal actions against Starbucks, collapse in stock prices, etc.
Linkedin is annoying, but this article is great, and you can read it without registering. Just click the X on the top right of the box to make it go away:
https://www.linkedin.com/pulse/from-national-gift-insult-how-starbucks-korea-destroyed-tz0ve
keep_replies_posts to true in server.json, remote posts that a local user has replied to are retained in the public timeline, preventing them from being purged (contributed by hanchan).keep_replied_me to true in server.json, remote replies to local posts are retained in the public timeline, preventing them from being purged (contributed by hanchan).A Practical Guide to #SSHTunnels: Local and Remote Port Forwarding