There is no 'unedited' version of Anne's diary, as Anne herself intentionally edited and re-edited her work during her time in the Annex. A remarkable young woman. What was published to readers are various versions that have additionally been further somewhat edited by others and in places censored, with the trend being towards gradually less censorship over time.

We've already been here in the 1980s.

The tech industry needs to cultivate people who are interested in the real capabilities and the nuance around that, and eject the set of people who am to turn the tech industry into a "you don't even need a product" warmed-over acolytes of Tony Robbins.

The incentives of different parties / actors are different. 'Everyone' necessarily comprises an extremely broad category, and we should only invoke that category with care.

I could claim "Everyone" wants banks to be secure - and you would be correct to reject that claim. Note that if the actual sense of the term in that sentence is really "almost everyone, but definitely not everyone", then threat landscape is entirely different.

Plenty of people will encounter snippets of this document and/or summaries of it in the process of interacting with Claude's AI models, and encountering it through that experience rather than as a static reference document will likely amplify its intended effect on consumer perceptions. In a way, the answer to your second question answers your first question.

It is not that the document isn't used to train the models, of course it is. Instead the objection is whether the actions of the "AI Safety" crew amount to "expedient marketing strategies" or whether it's instead a "genuine attempt to produce a tool constrained by ethical values and capable of balancing them". The latter would presumably involve extremely detailed work with human experts trained in ethical reasoning, and the result would be documents grappling with emotionally charged and divisive moral issues, and much less concerned with to convincing readers that Claude has "emotions" and is a "moral patient".

1. "thou shalt not destroy the world" communicates that the product is powerful and thus desirable.

2. "do not generate CSAM" indicates a response to the widespread public notoriety around AI and CSAM generation, and an indication that observers of this document should feel reassured with the choice of this particular AI company rather than another.

Why do I owe you any specific "explanation" when the context here is that you are treating Shrier's pseudoscientific book that literally tells parents in the closing chapters that if their kid has a trans friend they should consider moving cities to get their child away from their trans friend as though we are supposed to take transphobic hate literature at face value.

Maybe a better step than me agreeing to do that is that instead you should take the entire corpus of medical literature on the subject, as well as the voices of trans people on the subject of trans people at face value first.

I have no interest in your JAQing off[1].

[1] https://en.wikipedia.org/wiki/Just_Asking_Questions

No, instead she is likely to avoid talking in abstractions and instead talk about personal experiences of getting stalked online by multiple people she has had to show her details to in the past, who may include storekeeps, police, university staff, etc, etc. Eva Galperin is an excellent source on the way many of our procedures are designed in ways that do not at all account for the potential of stalking and harassment, though her focus is on how this continues to unfold in the technology space.

There is considerable overlap between those who subscribe to the "trans people are a contagion" moral panic of writer Abigail Schrier, and the "ban social media" advocates in AU who were instrumental in creating this legislation.

Ask a woman in a liquor store whether her anonymity is maintained by this scenario...?

The current liquor store approach for buying liquor is hazardous for a good chunk of people and we need to acknowledge that - even if acquiring a token somewhat ameliorates the compounded risk from presenting ID multiple times

So many of these internet ban proposals feel like someone creates a single cartoon scenario that captures ~2% of the use cases, and happily charges ahead to a proposed solution as though they've sufficiently thought about the people affected and the harms involved.

I happen to think there are plenty of valid points regarding harmful content on steam and valid arguments about the harms of social media, but I do not believe Collective Shout is a benevolent actor in combatting those harms or steering the solutions, as their proposals nearly always deliver harmful effects on LGBTQ people - and this fits with Reist's previous work[2], eg under Sen. Harradine

[1] https://en.wikipedia.org/wiki/Collective_Shout

[2] https://en.wikipedia.org/wiki/Melinda_Tankard_Reist

Regulatory capture. It's worth noting that an enormous amount of time and energy has already been allocated in this exact direction.

> It boggles the mind that in an open software culture whose central ethos is continuous iteration and improvement made possible by openness, our licensing stack and its ingrained principles are apparently immutable.

If/when the huge Satoshi bitcoin stash gets traded in, we'll see similar outcomes there too.

How is anyone supposed to believe the administration is being at all genuine when it categorizes that sentence as an endorsement of murder and then applies punitive action toward the man who wrote it?

Are we now at the point where (in Soviet Russian style) the government gleefully makes absurd factual claims and administers capricious punishments specifically as a demonstration of the government power to oppress?

Government sanctions (not simply those limited to imprisonment) used as a punishment for speech is most definitely a free speech issue.