Online Safety Bill Etc

[Carth]

Originally Posted by RichardCoulter

Within the last hour or so, the PM has given Google & Apple three months to stop children seeing or sending explicit content. If they don't, legislation will be introduced to require it.

https://www.bbc.co.uk/news/live/cr5j43zp2rpt

Tory leader Kemi Badenoch says it's "not enough" to ban certain features and argues that social media "is not for children".

They don't have a clue do they.

[papa smurf]

Originally Posted by Carth

They don't have a clue do they.

i think this is political suicide these "children" will soon be voters.and they won't forget who took their friends and lifestyle away from them

[damien c]

Meanwhile I am doing something to secure my data and network, won't need to worry about the dictator in charge our lord and saviour starmer.

There are things I have noticed in the software for the Ubiquiti Unifi Dream Machine Pro (the white box at the top), that I think all ISP's should be forced to put in the router software and that is the ability to completely, block app's, websites, ip's etc but that won't happen because the Government and there paymasters won't make money that way!

Thought this was a reasonable summary of what is, and isn’t, possible…

https://www.patreon.com/posts/160600382

The first thing to note is the profound and depressing groupthink and lack of critical analysis - there's no apparent realisation that these things will have an impact outside of the narrow scope of children that they are concerned with. When even the government spokesperson is saying things like:

"...new expectations on technology companies to introduce crucial safety measures on children’s phones..."
It is apparent that for many it is still not clear that for this to work, it must impact EVERY phone, every device, even - companies do not have some magical way of distinguishing a child's phone.

Beyond that, the debate (both in parliament and elsewhere) is often characterised by profound misconceptions about what is possible in terms of the internet as a whole. I don't know what this is, although I have theories (the mental model of an effectively anarchic, borderless entity doesn't fit the common experience of most people, particularly politicians steeped daily in national legislation and a sense of complete control, amongst other things).

So, as a reference, here are some presumptions that people make that don't hold...

These things... are not possible

You cannot reliably draw a line between "social media" and "the rest of the internet". If you wall off a chunk which you've labelled, something new will spring up outside of that wall providing the same functionality., because you will have created a market by effectively removing an incumbent.

You cannot reliably prevent people from reaching parts of the internet if they want to enough. Even massive state-level efforts like the restrictions imposed by China are routinely bypassed. The internet, at a technical level, is designed to be resilient to disruption. Some of the technologies used to circumvent some approaches to restriction/censorship are the same technologies required to make aspects of the internet work at a fundamental level (VPNs, and so on). If these essential technologies are available to some, they will be available to all if the desire is high enough.

You cannot reliably impose content standards/classification globally on the creator side. I've seen proposals including things like devices recognising tags and classifications embedded in pages which the creators/owners provide. There is no jurisdiction to make this happen even if it was practically workable - which it isn't, as there is no upside for people to do this at all, let alone accurately and honestly.

You cannot reliably categorise/recognise content on the consumer side either. Despite what companies selling a product claim, any technology to do this is of limited accuracy, with both false positives and false negatives common even when best efforts are applied. These systems routinely determine entirely legitimate content to be disallowed even when free of political interference - and allow content which should be blocked. There is a social cost to this, routinely ignored.

You cannot reliably apply a set of rules to one subset of people without applying it to all people. There is no way to develop a law, a piece of technology, or anything else, which will only apply to children (or any other group). The current direction is guilty until proven innocent - you must prove you are not a child, and therefore any system imposed must apply to every device for every person if it isn't to leave easy loopholes.

You cannot develop technology which can only be used by "the good guys". Any system developed which would be harmful in the hands of the "the bad guys" will end up in those hands at some point. A system which sees everything you do for supposedly benign reasons will at some point see everything you do for malignant reasons, because there is no such thing as 100% secure software or hardware, and the motivation for bad people to exploit such a system is vast. This is not because people aren't trying hard enough, or because laws haven't been passed demanding it - we simply don't have the ability.

In short - you cannot solve problems at the consumer end. If something is out there on the internet, it will be accessible to people who want to find it. If it is something that almost everyone wants to find, finding it will become trivial as systems (both technical and social) spring up to facilitate that.

These things... may be possible

So what can you do with the internet? What IS possible?

You can't reliably block things - but you can make accessing them significantly harder, and that matters enough to large corporations dependent on consistent engagement that it gives you leverage.

You can use that leverage to regulate - you can impose legislation on transparency, on algorithmic behaviour (for everyone, not just children - radicalisation, etc. is a harm for all, not just minors), on moderation, and so on. Nobody will like it, but it is possible.

You can invest more in education, both for children and for parents, around possible harms and how to avoid them.

Taken in concert, you could significantly reduce some of the worst excesses of the common, consumer-focused areas of the internet. You can, from a technical perspective, do almost nothing about the dark corners - although you can invest in better resourced and trained policing.

So?

This list may seem gloomy and defeatist. I have no doubt that politicians would describe it as apologetics or a technologist downplaying the problem or not caring about the real and significant harms caused. It isn't. But it does skew heavily towards realism. You can invest all your money in imaginary solutions which could solve all the problems completely - but don't. Or you can invest in pragmatic solutions which make things better - for some people, some of the time, yes, but still better.

For me, I want politicians to engage with reality, however unpleasant they may find it - the current approach of magical thinking and a refusal to engage with criticism lets everybody down.