How should I feel about this potential mandated block on X AI?
-
How should I feel about this potential mandated block on X AI? On the one hand I am not in favour of a blocking requirement being dumped on eyeball networks. On the other hand, clearly X AI is making problematic and illegal content.
How to navigate this, morally and ethically?
I feel people are often too willing to bend their preciously-held principles when the target is someone they don’t like. And I’m trying to be careful not to be that kind of person.
@bloor Making or posessing CSAM is a criminal offence with strict liability so (1) anyone in the UK using Grok to make or view CSAM ought to be referred for prosecution; (2) publishing CSAM where UK people can see it and making available in the UK a system that generates CSAM ought to be criminal offences too (so request Musk’s extradition to stand trial).
Enforcing those laws might reduce the imperative to treat the symptoms by blocking the content.
-
Fine by me. It shouldn't be necessary in a tolerant and empathetic society but fascists don't get that level of acceptance.
-
How should I feel about this potential mandated block on X AI? On the one hand I am not in favour of a blocking requirement being dumped on eyeball networks. On the other hand, clearly X AI is making problematic and illegal content.
How to navigate this, morally and ethically?
I feel people are often too willing to bend their preciously-held principles when the target is someone they don’t like. And I’m trying to be careful not to be that kind of person.
@bloor I find it interesting that the trigger for this was images of Kate Middleton, yet that is mentioned nowhere in the UK news. Would it have gotten the same attention otherwise?
-
@bloor I have no problem blocking sites known to be pushing CSAM. To be honest, though, I wonder if in this case the solution isn’t to block the site but to block its ability to generate any revenue from the UK? I think that would have a greater impact on the type of content that the site pushes than a block.
I think there are already procedures to block visa/mastercard transactions to certain suppliers?@rachel @bloor there notionally are but visa and mastercrud run their own rules which are tighter. Had it been anyone who wasnt large and American they'd already have acted.
I would rather see OFCON block Grok as that's the piece of the system at issue. If the chief twit refuses then the rest is on his head.
-
@bloor I think the OSA is the wrong tool - the problematic stuff is being generated *by users* and clearly illegal, so they should be getting prosecuted for making CSAM/non consenting images.
Targeting just the platform rather ignores half of the problem
@ahnlak @bloor That's the machine guns don't kill people, people do argument. There are lots of things in UK law where there is a line you cross if you know about bad problems and don't act. Not that you are liable for users actions but that you are liable for enabling a problem, knowing about it and not acting.
True for pubs, rentals and all sorts of things It's no different to allowing your property to be used as a crack den.Prosecute those doing it too sure
-
@bloor I have no problem blocking sites known to be pushing CSAM. To be honest, though, I wonder if in this case the solution isn’t to block the site but to block its ability to generate any revenue from the UK? I think that would have a greater impact on the type of content that the site pushes than a block.
I think there are already procedures to block visa/mastercard transactions to certain suppliers? -
How should I feel about this potential mandated block on X AI? On the one hand I am not in favour of a blocking requirement being dumped on eyeball networks. On the other hand, clearly X AI is making problematic and illegal content.
How to navigate this, morally and ethically?
I feel people are often too willing to bend their preciously-held principles when the target is someone they don’t like. And I’m trying to be careful not to be that kind of person.
@bloor The steps should be:
- What is X AI / Grok specifically (legally) falling foul of?
- What regulatory body has officially informed X that they are in contravention of the above?
- What level of expectation was there on X to not be in contravention (i.e. what type of penalty is applicable that this happened at all) vs is this an appropriate situation that as long as X remediates within a reasonable timescale then no action necessarily needs to be taken?
- What is X's immediate response to being informed of contravention (statement to regulator, not just changing a platform setting and pretending all is sorted)?
- Is the immediate response deemed suitable and appropriate by the regulatory body?
- What is the long-term response provided by X?
- Is the long-term response implemented as agreed/with approval of regulatory body and does it reasonably prevent similar future reoccurrences?
If - after all that - any of the answers/responses are unsatisfactory, then a ban becomes appropriate.
I know many would want something much more decisive and quicker (not to say the above points can't be done quickly if desired), but that's the reasonable steps any business should be taken through. It isn't right for the highest levels of government to jump on a bandwagon of prejudice simply because it's becoming the "in thing" to hate on a particular platform/company (although won't be the first time).
-
@ahnlak @bloor That's the machine guns don't kill people, people do argument. There are lots of things in UK law where there is a line you cross if you know about bad problems and don't act. Not that you are liable for users actions but that you are liable for enabling a problem, knowing about it and not acting.
True for pubs, rentals and all sorts of things It's no different to allowing your property to be used as a crack den.Prosecute those doing it too sure
@etchedpixels @bloor that's why I said it's ignoring half the problem - at the moment *all* the talk it about using the OSA to punish X, and I genuinely don't understand why they're not saying "and we're going to prosecute everyone who's used these tools to break other, existing law too"
-
@bloor I feel similarly, I would not care if X was blocked but that's not a reason in itself to support a block.
I think it would be better in the very short term for the UK government (national and local) to abandon the site, both in accounts and advertising. Delegitimise and denormalise being on there and undermine it that way.
-
How should I feel about this potential mandated block on X AI? On the one hand I am not in favour of a blocking requirement being dumped on eyeball networks. On the other hand, clearly X AI is making problematic and illegal content.
How to navigate this, morally and ethically?
I feel people are often too willing to bend their preciously-held principles when the target is someone they don’t like. And I’m trying to be careful not to be that kind of person.
@bloor Now what about self-hosted instances? I'm betting there are AI models that you can run on a local machine/random cloud provider to do the same thing.
If you block X AI, how do you stop people running a model? -
@bloor Now what about self-hosted instances? I'm betting there are AI models that you can run on a local machine/random cloud provider to do the same thing.
If you block X AI, how do you stop people running a model?@penguin42 I don’t think x ai releases its model
-
@penguin42 I don’t think x ai releases its model
@bloor Maybe not but there's a whole pile of models you can get which I'd bet could do the same thing if pushed in the right direction; IBM offers Groq for enterprise usage (although my reading it's some fast access to Groq's own cloud)
-
How should I feel about this potential mandated block on X AI? On the one hand I am not in favour of a blocking requirement being dumped on eyeball networks. On the other hand, clearly X AI is making problematic and illegal content.
How to navigate this, morally and ethically?
I feel people are often too willing to bend their preciously-held principles when the target is someone they don’t like. And I’m trying to be careful not to be that kind of person.
@bloor I'm not sure that network level blocking is appropriate, especially when the are alternative options such as criminal prosecution.
-
How should I feel about this potential mandated block on X AI? On the one hand I am not in favour of a blocking requirement being dumped on eyeball networks. On the other hand, clearly X AI is making problematic and illegal content.
How to navigate this, morally and ethically?
I feel people are often too willing to bend their preciously-held principles when the target is someone they don’t like. And I’m trying to be careful not to be that kind of person.
@bloor I think the problem is I don't trust the government's motives either.. they want control of what people see on the internet. X just made it easy - blocking them is a no brainer.. I'd pay very careful attention to how that's done and what powers the government is granting themselves by doing so.
-
R AodeRelay shared this topic
