-
This amazing double thread by @daphnehk on California’s cloning of @5RightsFound’s Age Appropriate Design Code…
…are some of the AADC observations which we wish we had seen coming out of the House of Lords, except coming out of the US 9th circuit appeal court because the House of Lords contains Baroness Kidron. Alas, there are too many loose ends still. > The easy part was about the law’s requirement to
-
Brussels [European Commission] slaps down Thierry Breton over ‘harmful content’ letter to Elon Musk | …and rightfully throws @ThierryBreton under a metaphorical bus
Breton is a loose cannon: > On Tuesday the European Commission denied that Breton had approval from its president Ursula von der Leyen to send the letter. “The timing and the wording of the letter were neither co-ordinated or agreed with the president nor with the [commissioners],” it said. An EU official, who asked not
-
The former (“focus on the failure”) space is also where those people intent upon “regulating AI for the public good” largely live; their goal is to prevent advancement rather than improve it…
…unless of course all development can all happen on their terms.
-
The Many Reasons Why NCMEC’s Board Is Failing Its Mission, From A NCMEC Insider | Techdirt
If you work in Trust & Safety, OMG this podcast is eye-opening, and it’s a must-listen. I’ve heard gossip before re: a lack of ability to execute, but this explains so much about NCMEC and its haphazard and ill-considered approach to technology regulation. https://www.techdirt.com/2024/08/08/the-many-reasons-why-ncmecs-board-is-failing-its-mission-from-a-ncmec-insider/
-
“What’s the most powerful uncensored LLM?” | …one person’s safety regulation is another person’s bowdlerisation
Quote: I am working on a project that requires the user to provide some of the early traumas of childhood but most commercial llm’s refuse to work on that and only allow surface questions. I was able to make it happen with a Jailbreak but that is not safe since anytime they can update the
-
We received internal Trump documents from ‘Robert.’ The campaign just confirmed it was hacked | POLITICO
:popcorn: https://www.politico.com/news/2024/08/10/trump-campaign-hack-00173503
-
From 1999 | “how can I become a cryptographer?” | Schneier on Security
I wonder what this would look like nowadays? Quote: The short answer to “how can I become a cryptographer” is: “Get a PhD in cryptography.” This is not the only way to become a cryptographer, but it is by far the easiest. https://www.schneier.com/crypto-gram/archives/1999/1015.html#SoYouWanttobeaCryptographer
-
‘A Platform That’s Teetering on the Edge of Becoming a User-Experience Joke Akin to Windows Vista’ | Daring Fireball
If you are a nerd of a certain age, you will remember this advert: Via: https://daringfireball.net/linked/2024/08/08/snell-mac-permission-balance
-
UK ‘desperately exposed’ to cyber-threats and pandemics, says Minister @PeterKyle | The Guardian
Peter Kyle is *this close* to realising that maybe “cyber” would better be dealt with via a “public health” rather than “policing” approach: “National resilience suffered terribly, catastrophically,” he said. “The open warfare of the previous government prevented any kind of progress in these areas and left our country desperately exposed not just to a
-
“Keeping your accounts safe from AI content moderation” | …pacifying or circumventing “trust & safety” will be the SEO of the next decade
RedGIFs Creators: “…the reality you deal with now that AI is running the show with content moderation. Nobody knows EXACTLY how it measures what to take down and who to ban, but it’s pretty safe to assume that the more interaction a post gets, the more it can get away with. So, unless you’re working