Letter to MP

The approach the UK government is taking to online safety is already turning the country into a digital privacy dumpster fire. I believe that the proposals for an Australian-style ban on social media for under sixteen year olds will make things much worse. This has concerned me enough that I responded to a government public consultation yesterday, and I’m going to send the following to my MP…

Dear MP

I am writing to share our experiences resulting from your government’s Online Safety Act, and how, intentionally or not, it has sacrificed the privacy, security and free speech of adults, at the altar of ‘child safety’ while simultaneously failing to address the real problem of algorithmically directed content pushed at everyone, with the explicit aim of causing addiction in social media users of all ages.

Recent updates to the iPhone have added mandatory age verification at device level. One assumes this is in anticipation of further age verification laws being made in regions in which it operates. I imagine other manufacturers will follow suit, if they haven’t already.  Our experience starkly shows how technological approaches are deeply flawed and based on assumptions that are likely to be untrue for parts of the UK population. Apple have decided to enforce age verification at device level by way of their online account ecosystem, which is required for use of their devices.  They verify age based on three factors:  the age of the account, whether there is a credit card attached to the account, and whether you scan a ‘government issued ID card’.  The last of these is restricted, in the UK at least, to a photocard driver’s license.  I hope you can already see that this is inequitable, as it discriminates against those who cannot or do not wish to have a credit card (low income individuals or families for example), or do not drive. I do not feel that either of these are legitimate reasons for restricting access to online services and treating anyone who cannot comply as a child.  This is exactly what has happened to my husband – he neither has, nor wants, a credit card and doesn’t drive, and his passport could not be scanned by the age verification system.

While it is laudable of Apple to attempt to verify age in a way that respects their users’ privacy, as using credit cards is subject to rigorous regulation and is thus low risk, not all manufacturers and services take this approach. Many offload age verification to an emergent, unregulated industry which forces users into supplying valuable identity-related data (including biometrics via face scanning) which could so easily fall into the hands of criminals.  This is a mass identity theft and associated fraud incident waiting to happen.

I know you are supportive of the plans for an Australian-style ban on social media for under-sixteens, but I feel that this, while well-intentioned, would compound the problems of age verification with respect to digital privacy and security.  It also ignores the fact that the harms associated with social media are harmful to everyone, regardless of age. We, as a society, are in danger of producing a generation of children without the skills to navigate digital life when they are suddenly ‘allowed’ to access everything at sixteen or eighteen.  A better approach would be through education, for parents and children, on the risks of the online world, as already exists for the offline world (road safety, stranger danger, RSE etc.) and how to navigate the whole world, online and offline, safely.  While the harms of social media are real (as the recent legal case in California, when Meta and Google were found guilty of creating algorithms designed to cause addiction) the answer is through regulation of these platforms rather than of the users.  I am confident that this is the approach that can provide a better digital experience for all internet users, while protecting privacy, security and legal free speech in an equitable manner.

Thank you for taking the time to read this rather long email. This is an important issue, which could so easily have unintended consequences and cause more long-term harm than it reduces.

Regards

Stewart Tolhurst

My response to the consultation was along similar lines. Sadly the only equitable way to enforce age verification is to have mandatory ID (which would need to be cheap or free to obtain) with strong regulation and not tied to any third party services (Germany trying to tie it to Apple and Google is completely wrongheaded) – which unfortunately probably means governmental (and we know how big technology projects tend to go when the UK government is involved). The technology is there and we can’t afford to leave this to the private sector…

Threads in shreds…

I was very much of the opinion that, given the amount of control Mastodon, and the Fediverse in general, allows over what content you experience, blocking Threads at server level was counterproductive. Sadly yesterday’s announcement from Meta has meant I’ve changed my mind. The lack of any moderation, and an AUP that appears to encourage abuse of minorities, leads me to conclude that Threads is destined to become an X-style hellscape. It’s such a shame as a major player in the social media landscape could have really helped ActivityPub and Fediverse decentralised services reach their full potential.

Some thoughts on Social Media

I’ve been hanging around internet communities for a long time now. Usenet, Yahoo! Groups, IRC, Livejournal, and all the other places to hang out on the early internet. It’s easy to get nostalgic for those days before we were algorithmically spoon-fed cat videos, cleaning ‘hacks’, and hyper-speed cooking videos ending with an oozing chocolate or cheese ‘money shot’, and conveniently forget the never-ending flame wars and general bad behaviour the internet has always seemed to bring out of the most mild-manner netizen. Those early days were always in flux – whole communities would often up and migrate to the latest thing, usually leaving defunct profiles forgotten and fossilizing until either the site went offline or were transferred to increasingly dodgy owners. People weren’t so concerned about digital footprints and online privacy back then… In the mid-2000s the landscape changed completely with the launch of Facebook and Twitter. Both started life quite small, but soon ballooned, making what was once a niche side of online life mainstream. Twitter, in particular, managed to have an impact beyond its user base, becoming an almost real time feed of what was going on around the world. The trouble is that once online communities reach a certain size, social norms become stretched to the limit, and it is beyond any moderation scheme to balance free-speech v acceptable behaviour. It was inevitable that social media would become a toxic hellscape, which appears to have hastened the polarization of society with information overload, ‘deep fakes’, ‘culture wars’ and the endless need for fact checking… If you’re going to venture into the digital badlands, it’s useful to have a bit of a travel guide…

Facebook

Facebook is probably the one that started the modern era of social media, and it is showing its age. It’s the site that everyone and their grandparents are on, but hardly anyone posts to. Almost unusable without some way of controlling the ads and recommendations. On the odd occasion something is posted, you can guarantee someone will be negative within 10 comments. Conversation threading is, as is common to most social media platforms, rudimentary at best, which makes keeping track of things virtually impossible. Like all Meta platforms, there’s a lot of overlap with their other products (Messenger, WhatsApp, Threads and Instagram). Honestly, I’m not sure what Facebook is for these days…

Instagram

Instagram is Facebook’s cooler sibling. Manages to attract a younger crowd than Facebook, but suffers rather from encouraging an over-curated and unrealistic view of life. Morphed from an image centric platform, to more video and ‘reels’ content, presumably with the aim of competing with TikTok (though the conspicuous consumption is toned down somewhat and the more harmful content does seem to get moderated away). The 90-second limit to Instagram videos has led to a very specific style – lots of high speed cooking or gardening pieces with floating captions. Catnip to marketing departments because the algorithm appears to be able to accurately predict the type of content users will engage with.

X (née Twitter)

Twitter has long been a dumpster fire, and Elon Musk’s acquisition seems to have been an exercise in throwing on kerosene because the flames are pretty. The algorithm here seems to be designed specifically to generate outrage under the guise of ‘sparking discussion’. The toxic environment has driven a lot of people away, but somehow it keeps on going…

Bluesky

Touted as a reinvention of Twitter, from pretty much the founding team, but aiming to improve on the original. A decentralized platform (but nothing else uses the protocol yet) with which aims to show users content based on a mix of interests and engagement. In reality, it will, like other platforms, easily lead you down a rabbit hole of similar content. The adult content moderation appears to work quite well (I suspect some machine learning from community moderation is involved), which is fortunate as there are a lot of dick pix! Discovery is horrible – you are very much at the mercy of its ‘Discover’ and ‘For You’ feeds. It originally claimed not to need hashtags, but as most of the users have migrated from Twitter they get used by default and are now properly supported. The community is rather insular, probably due to the invite only beta phase, and does not tolerate violation of unwritten community norms. Block early and often seems to be the general culture, which is odd for a site with little in the way of privacy control.

Fediverse

The Fediverse has been around a while now, with frequent surges in membership whenever X becomes more of a hellscape. It’s very different from other social media as it is a collection of services, some more social-media-like (e.g. Mastodon) and others similar to big services like YouTube (PeerTube). The glue that sticks the Fediverse together is the ActivityPub protocol, which lets users in one platform subscribe to users in a different one. In theory it gives users a lot of control over what content they get – for example you can follow hashtags, meaning if people use them you get a good non-algorthimic way of discovering content, and easily filter your feed based on hashtags and keywords. There is, however, a tendency for the smaller servers to ‘de-federate’ whole servers. Some of these de-federations could be considered perfectly reasonable (Truth Social and a few other far-right networks that have taken advantage of the open-source nature of Mastodon for example), while others are more verging on the censorous (for example a lot of small servers don’t federate with Meta’s Threads or journalist focussed instances). I do think the Fediverse has a lot of potential – the openness means that it isn’t *just* one thing and it’s just as easy to Federate a long form blog (like this one) as it is to microblog on Mastodon (or any of the compatable forks like Hometown) or post videos on PeerTube and for people subscribe to that content.

Final Thoughts

I think the days of centralised Social Media are on the wane. Hopefully The Guardian leaving X is a sign of things to come, but as the landscape becomes more fragmented hopefully interoperable and extensible systems become the norm. Of course the trouble with open systems is that is very hard to monetise them, which means they are unlikely to go mainstream and remain nich community hubs, but one can dream…