You will have read that Mark Zuckerberg has decreed that Donald Trump will be banned from posting on Facebook until at least after the Inauguration. The reason is that Trump used Zuck’s “ platform to incite violent insurrection against a democratically elected government.” This was enough for Zuck and he finally pulled out the naughty step.
So many aspects to discuss here. I am particularly interested in exploring
how this plays out in the 200 or so other countries where Facebook is active. Will the firm apply the same rules? What if a democratic government is acting against a minority, and using Facebook for its messaging?
the balance between private platforms that now have public functions. Do any civic responsibilities arise on those platforms? Should these be global or local standards?
if a platform has a civic infrastructural function, how should it be governed? (We govern civic practice differently to infrastructure differently to run-of-the-mill companies in most nations with strongish political and economic institutions.)
Lots to think about. Emotive topic. Measured, thoughtful and polite responses, please!
Lots of thinking in development orgs and multilaterals at the moment about platforms as civic infrastructure, but unfortunately too much of it still framing platforms as “digital public goods”. We wrote a series of internal policy papers for DFID/FCDO last year including one on digital harms as an attempt to counter this anachronistic thinking. But there hasn’t been a lot of policy engagement of the negative side of social media platforms outside of US & EU, largely because for a decade or so extending access to internet (which in emerging markets basically means social media) was perceived to be something that raised economies and liberal societies - although without much impartial evidence of this, as the World Bank’s WDR 2016 report on Digital Dividends points out.
But what happened in the US this week has already happened in other countries, notably Myanmar but also many others, to limited response from FB et al. But on their doorstep, with a new incoming administration, they couldn’t not react. I agree with Sam here that this is too little to late - what we label as Q Anon conspiracy theories are actually just radicalisation, which is what we call distribution of the same content if it’s Islamist. It should’ve been halted years ago, but the platforms are unwilling to step outside of their get out of jail free platform cards and admit they’re media orgs. They’re happy to take the income of a media org, but not the responsibilities. They will eventually police their platforms in the US because they’ll be forced to, but I can’t see them doing this in the long tail of other markets who contribute less. US is >50% of FB, Twitter revenues. “Rest of World” is tiny in comparison, and much harder to police because the language and culture of emerging markets is less well represented in ML training data, and therefore much harder to manage using AI, and revenues aren’t high enough in these markets to support human moderation costs.
I gave a really depressing presentation to a World Bank group last month that played out the emerging market revenues of social media platforms, their ubiquity in people’s digital lives in these countries, and the lack of knowledge and motivation platforms have to police them properly outside of US and EU. It ends by pointing out we’re likely to see a Rwanda-style genocide at some point in the next few years driven by manipulation of social media in these markets. Arguably we’ve already seen it with the Rohingya.
There’s often a simplistic sense that digital infrastructure is like physical infrastructure, that you lay some out as public good and public benefit derives from this. My old employers the GSMA compounded this thinking with the doubtful claim that for every 10% increase in mobile subs GDP increases 1.4%, and this has been compounded by a Facebook funding research that claimed for every 10 people you connect to Facebook you pull one out of poverty. This has reinforced existing simplistic development narratives on how technology impacts on poverty and societies - the main issue is seen as inclusion, and getting to a magic number of connections per population, rather than a more nuanced look at what people are actually connecting to, and what their digital lives actually are. So the focus is on funding a digital stack, often taking India’s as a template, in the belief this magically leads to positive outcomes and GDP increase rather than engaging with how digital economies and societies are really emerging on the ground.
We’ve successfully challenged this with many clients by getting them to focus on outcomes such as Meaningful Use rather than Inclusion, which considers what digital platforms users are engaging with and whether these are net positive and/or negative for their incomes and livelihoods. Our senior director of research Jonathan Donner frames this well in his 2015 MIT Press book “After Access”.
Yep. So the India digital stack is built around Aadhaar as an ID platform. This has clearly been a public good and has driven rapid adoption of digital payment services enabled by eKYC. But it’s also been used to exclude minorities in some regions. We mapped the digital identity ecosystems in Kenya and Ghana for the Australian Government to get a broader political economy sense of what impact they would have on digital economies and societies, and this allowed us to successfully predict that Huduma Namba would exclude millions from digital ID in Kenya and actually exacerbate digital divides. https://link.medium.com/ft4DgxTCScb
Sorry if I’m going on too much! And we’re some way from the original topic.
This is really fascinating, Chris.and speaks to my current concerns as we try to reframe the BBC as a guarantor of a digital public space rather than a broadcaster with public service remit. I absolutely agree with your point that just providing access does not guarantee good -or even acceptable - outcomes. Thinking about the services you'd want, and how you would assess their impact, is as important as finding ways to move the bits around.
I think a lot of this thinking has been done already by mutual friends of ours (Tony Ageh, Mike Bracken, Tom Loosemore). My twin brother Matt has a nice framing of the public media stack as a way to identify and encourage the use of what are clearly definable digital public good platforms https://publicmediastack.com/
Do you think that will lead Facebook to pull out of some countries where the liabilities > revenue opportunities, or will they still operate in those countries but with no standards?
It’s a tricky one. There are already many countries where we suspect their cost to serve FB customers is greater than the revenue they earn - by our calculations from a couple of years ago maybe most of the last billion users they added are probably not profitable, based on revenue we modelled and the extra costs to serve users who are on lower end devices with no JavaScript or cookie tracking, or are just using WhatsApp, etc. The report is here https://link.medium.com/cSku5ZqoUcb
We suspect this is why Libra was developed. People forget quite how unique the US market is in terms of advertising revenue market size. It’s totally unlike even EU markets, and when you look at India, SE Asia, Africa these are transactional markets rather than attention revenue markets. Pivoting a product based on 99% of revenues coming from advertising to one that can become profitable from transactions is hard, and we suspect Libra was a punt at this, which has failed already. Getting into financial transactions is costly, requires a solid ground game in emerging markets (lack of bank accounts means you need networks of agents for cash in/cash out to digital payment platforms) and brings you into direct contact with regulators. All things FB is allergic to.
Every year I expect the bubble to burst and an analyst to properly ask where the revenue is for the last billion users FB added. And every year ad revenue growth in the US bails out their figures. But at some point these revenues will plateau, and the question will become more immediate. FB had to work out how to monetise WhatsApp users in emerging markets in particularly, and it has the opportunity as our research has shown how integral WhatsApp is to informal trade and marketplaces in emerging markets. But turning informal discovery into a trading platform is still hard.
And also the real challenge for GAFA firms in emerging markets is, will Ant Financial get there first? It’s a battle between Silicon Valley firms with all the reach but no ability to get transactional revenue, and Chinese super platforms that have all the experience in transactional revenue from low income users but none of the reach. This is why Ant has been buying into SE Asian mobile money platforms like it’s going out of fashion (Wave Money, Bkash etc.) - they need to buy the reach before GAFA works out how to turn their attention reach in these markets into transactional revenues.
Many would say this action was six months late. The timing was interesting - coming on the day the Democrats took all three branches of power. And great pressure from Republicans before the 2016 election meant the company were nervous of allegations of censorship. But Zuckerberg wields ultimate power and seems indifferent to the toxicity of Facebook's approach thus far. The share price has grown steadily through the year, and his individual power with it. He can fix this whenever he likes. The reality is that legislation is the only sure way to adapt the way social media companies operate. They have to be treated as editorial publishers, with the same responsibilities. Facebook - like Google - is an advertising sales platform, powered by the ethanol of emotive editorial content. This carries great responsibilities. It's time for legislators to adapt and effectively regulate.
I do. As recently as 2019, Zuckerberg publicly requested new regulation and oversight. This effort is already underway in the UK and elsewhere. A global regulatory approach would be the most practically efficient for the publishers, but unless this is voluntary - like IPSO in the UK - I can't see a legal framework for it. This would also save the publishers a great deal of fragmentation expense. With Facebook et al publicly seeming unwilling to solve their own problems, sadly, this will become a multinational lawyerfest which will ultimately reduce the profits and the power of these companies in the longer term.
Totally agree regulation is coming, but it will take a long time to extend outside of the higher revenue US & EU markets into the countries that are really vulnerable to social media radicalisation because of their weaker civil societies. Over time these markets do copy US & EU regulation, but choose to ignore it where it benefits Government. My old friend Tom Wheeler at Harvard’s Shorenstein Centre writes and talks well on this topic of how FCC policy does, and doesn’t, roll out globally for social media.
Indeed - and that is what will happen. The big shift from governments has to be to disallow the "we aren't publishers, we are just a platform" approach, and to enforce the "we are publishers and advertisers, responsible for what we publish and advertise". Millions of practical and philosophical reasons why FB will struggle with this, but its the right thing for society, and therefore longer term, for FB and their shareholders. Otherwise, they risk increasing reputational decay. Zuckerberg must feel trapped, otherwise he wouldn't go public on wanting more regulation. Hopefully now this gets going at pace and we move into the next phase of social media - less Wild West, more trustworthy, more respected, more useful.
My current homework is to read the Digital Harms legislation going through Parliament and I haven't yet done this. However I don't see why all politicians and political campaigns are either banned from Facebook, after all the White House has it's own platform as does 10 Downing Street and their followers could follow there - in the open! Or they are banned in the 28 days preceeding any elections at any level - there is precedent for this for other political campaigning and advertising in different places. That would be the easy part. Activity in private groups is harder to deal with. I do think the platforms should be regulated as publishers are regulated, and people who post subject to the law of libel, race hatred, incitement to violence wherever they post from. But I guess I will have more to say next week!
This sets many precedents, in my opinion. It shows the power a world leader can build through the use of social media. It shows the lengths social networks must go to police world leaders' social media activity. And it paves the way for lawmakers globally to impose tighter regulations. To date, only in Brazil and Venezuela have we seen a leader's social media posts removed by the social network, as far as I can see from looking into cases. I think however that we will now see far greater proactivity from social networks in this space, and more attention from the global press on this as an issue.
Lots of great thinking in these threads. One aspect I did not see explored as much as I believe is warranted is the competition law vs. freedom of speech basis to challenge/regulate the large social media platforms.
At least in the US, it is quite settled constitutional law that owners of private forums have no obligation to permit all comers to speak. There was a long line of First Amendment caselaw concerning the rights of shopping mall operators to exclude free speech proponents who argued that in the middle of a Minnesota winter, standing on a soap box in a privately owned mall was constitutionally equivalent to Speakers Corner.
To me the argument is not that I should have the right to enter your property or participate as a matter of law in your online community, the societal interest (and the concomitant legitimacy of government in a free market economy to regulate) lies in the size and reach of that platform. If it is so large and dominant that there are no other effective platforms to engage in public debate, then these “essential facilities” need to either be broken-up to allow diverse competition, or regulated as was the case with the early “Fairness Doctrine” that required the holders of broadcast licenses in the US to provide balanced and equitable access to their “platforms.”
The Fairness Doctrine, like Section 230 was not based on competition law, but the power of the FCC to set rules governing the use of “public airwaves.” In the case of FB, TWTR etc. the underlying internet is not regulated as the original broadcast spectrum was and the doctrinal justification for regulation could be taken from competition law. These laws obviously differ from country to country, but concepts such as dominance and abuse of market power translate.
Finally, it should be noted POTUS has always had an effective pulpit from which to be heard: the traditional and official White House Statement or press conference. He should not be heard to complain about deplatforming when he has done so much to promote hate speech and incite violence.
So I think the issue here, and am writing about this in the "you know what", is that we need to reframe anti-trust and competition law to a broader frame that includes the moral and civic space, that the effective privatisation of these new forms of the public space (in the Habermas sense) is a process that warrants the purview of governments on our behalf. In this context, whether as these networks get bigger they do so while maintaining public and civic health.
I have a naive question. If everyone posting to a social media platform is doing so as an individual private citizen, then all can be subject to the same rules against hate speech or inciting violence. Governments could choose to adjust their codes of practice such that officials cannot use social media platforms in an official capacity at all, but the social media platforms could also choose to prevent that usage. Then there would be no privileged @POTUS account, and the @realDonaldTrump account would have no protection from bans. The Twitter algorithm, instead of warning about misleading information, could politely inform government officials to announce their policies on their own government web sites. Implement manual moderation on every government minister or leader across the world, and stop them tweeting policy (e.g. - you can send out a link to an announcement, but not make the original announcement on social media). Problem solved :)
I think one issue was the reverse logic that Twitter and Facebook used, which was to hold public officials (aka Donald Trump) to a lower standard than they would hold other users.
In most places, we expect public officials to behave with greater probity (think about the Parliamentary Standards).
It would have been better for them to adopt a clear standard and happy it very consistently.
It would also make sense to have a sliding set of standards, perhaps as accounts get more followers they might qualify for different (tougher) objective guidelines.
I am sure that twitter went through a process of clarification that was meticulous and involved many lawyers and experts, but ultimately it is an act of fiat, however we describe it.
From Facebook side: There is minimal incentive to apply this approach elsewhere. On home turf, with a lame duck president, and at least four years of corrosive disinformation and lies, they acted only at the very very last minute possible, with the whole world looking on. That circumstance won't repeat anywhere else with same dramatic urgency. Facebook has never shown any real thought on this, and as with Rohingya crisis, only acted when they couldn't not. Removing Trump would only be meaningful beyond the gesture if it was followed up with a root and branch removal of the disinformation spread in his name whether by real people, malicious political actors or AIs. The scale of that work, and the bottom line hit it will mean means there's very little disincentive to do that, and hard political regulation will be very hard to achieve.
From nation state side: Very messily. Only states, or supra-states like the EU, can legislate to provide protections against the mess Facebook and other social media causes. There is no way that legislation can be created or implemented with any evenness across the world. Practically, if the EU, the US and China went after a break-up of Facebook they could in principle unbundle the products, but creating a consensus set of regulatory ethical principles seems super-unlikely, particulary as all these countries try to learn from Russia on the wheres and hows of info war.
There are historical analogies to FB’s dominance of its current day market in things like the East India Company which ran the bulk of world trade over a period spanning centuries and employed private armies to enforce trade terms on various countries, presumably arguing that consumers back home ‘valued’ the access to foodstuffs and opium until the excesses became so egregious it was nationalised and shut down, replaced by government rule. Competition law goes back to at least Roman times too, but usually took the form of limitations on trade practices disadvantageous to the ruler or monarch, by driving up prices on eg. basic foodstuffs to the extent that public discontent would erupt. Most current competition law regimes still assume that the playing field is ex ante level, oddly enough. In searching for a local or national regulatory response, the lack of a country that has solved the thorny issues many have raised here in this discussion to use as a leading light example means that governments will each have to struggle with laborious, thinking slow (a la Kahnemann) type of legal reform that takes years.
I suspect that Trump is a one off, and that they will resist doing this on a wider scale as they want to avoid responsibility, so I doubt they will act without an outcry. To the extent that the platforms can be held accountable, I suspect they will seek to use local standards, as otherwise they will be subject to oversight into how they police transgressions. I suspect that legislation will be required to force the companies to make changes, though if it's not done before the next election cycle, I doubt it'll be done at all. At least in the US.
The recent decision from Facebook to ban Donald Trump from posting on its platform is exceptional and context-specific; not globally applicable.
At the end of the day, Facebook is a (profit-oriented) company and will be proceeding accordingly strategically and tactically. Companies create their rules to support their mission and interest. So, no law and especially no company rule of the kind will be uniformly interpreted and executed the same way for every applicable case. If it does not fit to the current context of interest, the problem at end will be declared to be an exception, which leads to an exceptional rule.
Even law-makers and governments - while acting in the context of their mission - are still shaped by self-interest, power preservation, party, clientele, lobby, etc.
Although Facebook’s - and similar platforms - influence and public impact have taken on worrying dimensions, I don’t believe that any global standard will be applicable here.
Social medias like Facebook reflect the system in place and the underlying society with all its good and the ugly things; in particular, in terms of freedom of expression. The problem is not about regulating Facebook but re-defining and better understanding what we mean by the precious good name freedom of speech/expression. The same applies for a media company, its role, responsibilities, limits, etc. We need a new category of media company and a corresponding regulation framework that is suitable to Facebook & Co.
The point is, justice requires treating equal things equally and unequal things unequally. But who says what is equal and what is unequal? And who determines or establishes whether treatment is equal or unequal? That’s the challenge and injustice we have to cope with everyday.
There is an issue with double standards: why should politicians be exempt from the social media guidelines when other users are barred for misconduct? Why are television and telecommunication providers subject to public oversight and have to adhere to different rules than the social media?
Big tech empires that today have an even stronger impact on public opinion than classic media (boosted by personal data analysis and hyper targeting), and they do need public oversight because corporate voluntary self governance has never worked.
Unlike on TV or printed press, much of political influencing is completely hidden in social networks. An auditable trail of political advertising campaigns (as on TV or printed press) is particularly important in times of filter bubbles to hold politicians, parties and other political influencers accountable to established standards.
It seems that, today, local legislation - see EU or China - is most effective to make social networks adhere to guidelines (even though we may not agree with some national variants). However, given that big tech are supranational organizations and increasingly powerful, I am not sure if a national approach is sufficient in the future.
Govern private platforms as little as possible. Concentrate on opening up more platforms and finance that by taxing the giant profits of the big players FB, Google, etc. Obviously, that is not simple, but here are four simple steps to get started. Taking some billions from FB and Google, 1) give local public radio solid support -- say 30x what they get now; 3) subsidize existing local libraries so they can afford more branches and have local facilities for podcast and video production; 4) subsidies local and small scale publishing operations such as local newspapers, small scale publishers, etc. However, hate speech should not be allowed on social media. How? Start by studying how Germany has handled their Nazi's since 1945. One way or the other, we need to to see to it that a lot of journalists are hired to produce content not just repackage others.
Thank you, Blaine. I needed to read someone widening the aperture. People need information diets that don’t rely on a single channel or format. You highlight how breadth can be restored.
The ethical dimension underpinning the whole system is this: what’s moral is what’s legal and what’s legal is for sale. Instead, think about the $100 million Facebook has spent in DC lobbying in the last 6 years. The only difference between public and private is temporal. How long will it take for Facebook to drill into the legislative marrow and insert its DNA.
This is the key ... and has been for the last 5 years "They have to be treated as editorial publishers, with the same responsibilities. Facebook - like Google - is an advertising sales platform, powered by the ethanol of emotive editorial content." So sad that regulators have been so very slow out of the blocks. And we also need MUCH more focus on getting these big techs to pay taxes in the countries they are serving.
I'll share an observation that has me questioning my usual assumption to follow the money and the business model to get to the underlying problem inherent in platforms.
During the past 4 days every call I've been in composed of eager hyper-networked powerful let's-do-something Americans has ended with a call to see who has one of these men on speed dial: Mark Z, Jack D, Rupert M, Brian R. (Comcast/NBC) and then Mike P. Chuck S. and the one woman in the mix: Nancy P. In many cases someone on the call did have one of these folks on speed dial, and we then resorted to figuring out the talking points.
It's as if through all of these advancements in exponential and less exponential technology we've just recreated vast feudal kingdoms and our only hope for better governance is to appeal to the kings. The difference between now and then is that the kingdoms are interconnected in complex adaptive systems.
Across all three platform types: social media, left/red entertainment television, and the fortified executive wing - we humans (or the subset of humans called Americans) have created the same conditions. While we can fiddle away as philosophers and technocrats to pull and push the levers of social media systems, our selfish genes keep recreating this structure no matter the tech.
Time to rewatch some good Adam Curtis. Monkey in the Machines of Loving Grace and all.
Lots to think about indeed… A core and (to my astonishment) still overlooked issue is the lack of understanding of information within our society. By that I mean the fundamentals - what is information, what are the laws governing its storage, propagation and (most importantly) value. By ‘value’ I don’t mean the sometimes trivial copyright/IP view, but the true impact (or lack thereof), direct and indirect, its true economic value and political consequences. We live in a world of energy and have a decent grasp of laws of thermodynamics. We claim to live in the information age, but we do not have a clear understanding of what (scientific) laws govern the flow of information. We teach the laws of thermodynamics, we teach nutrition, we do not teach the laws of information.
The information flow patterns have undergone a radical evolution, far outpacing our brain evolution - many historical limitations of space, time, persistence and volume have been removed. This has perturbed our human networks and, like any change in the environment, has led to new (side) effects - from new media to new business monopolies to digital mobs, some awesome, some questionable.
What we seem to be struggling with is a framework for understanding the societal (and individual) impact of information in a manner similar to healthcare. We have made good progress in understanding the information flows within biological systems, from molecular biology to population genomics. We figured out how physical messengers (chemicals, food, pathogens etc.) affect individuals and society and built entire sub-fields of medicine around them (environmental impact, nutrition, hygiene, epidemiology, healthcare economics, ...). What seems to be needed next is a similar approach to information, that might lead to a better understanding of our information environment, information economics, information hygiene and allow us to tackle some core issues like information value, monopolies, censorship, etc., many of which lead to familiar healthcare moral dilemmas of public vs. individual good, ownership and control.
A controversial comment - do we remove anonymity and make verification of ID a condition for open social media? Would increase the risk for people using social media to challenge abusive Governments for sure, but I think using international civil society architecture to hold Governments to account for abusing their regulation over platforms and free speech easier than playing whack-a-mole with bot armies and fake new accounts. Maybe worst of potential scenarios to keep social media open and viable.
Verified ID could make sense as one alternative. Not clear that my mum needs to be on a platform that doesn't verify. Parler practically verifies (with phone number, doesn't allow burners).
There can still be an ecosystem of unverified services for the remainder.
Thanks Chris, for making this point. I also had this thought many times before. I feel any public space needs an identifiable public (not a bunch of semi-ironic avatars) to build a baseline of language, values and norms. So people can build upon that to make decisions for the common good. I mean what is a community if you never even see your neighbors?
But Facebook (which seems to be the most problematic of the major networks) already has a "real name" rule that is mostly followed. One of the things that continues to surprise me is how little that seems to matter. Many of the Q followers seem OK with posting as themselves and the expected social consequences (ostracism, job loss, etc.) seem to not be happening.
Valid point, although I doubt that Facebook is where most of the action happened that led people to the Capitol.
Also, on a different note, I suspect that although Internet culture flowed over the barricades into the heart of politics last week, it was those political barricades that led them towards internet culture in the first place. In other words, I feel that this isn't just a tech problem, it's a societal problem. Borne out of the many fault lines that are being aggravated in our rapidly changing society. Perhaps a technological fix isn't enough to stop radicalization - we also societal fixes.
Unfortunately, this narrative plays into Donald's hands and his supporters will lap it up. "The world is against me and is trying to silence me. Rise up and stand with me against such evil tyranny!"... oh the irony.
After the 6th, private platforms did a better job of reflecting the public interest -- including support for constitutional process -- that most public institutions have so far. Excellent work has been done on the intersection of social media and some law enforcement.
Prior to the 6th, it was a different kettle of fish.
In most countries the need for such rapid action would not have been so great -- in the US we have an uncommon number of heavily armed people. As noted elsewhere, I think we need a lot more platforms, many more public spaces, etc. We also need to use education to equip people to participate in social media on a different level.
Business models which are based on maximizing engagement seem to create highly reactive conversations with little time for reflection, investigation, etc. We can probably help establish more reflective, conversational norms internationally, but locally I think we need to develop new business models and restrict existing ones. If the model is centered around selling personal information, the model will value engagement and that leads to frantic, reactive conversations.
More than 150 years ago we took strong steps to ban the sale of humans. I think we need to ban or eliminate the sale of human data as well.
The sharing of human data is a different issue. We are unlikely to deal with pandemics well or have safe civic discourse w/o safe sharing of personal identity and data. One can imagine competition to develop platforms which support a new class of public health apps -- apps with international value -- while preserving identity and personal location information. I think UC Berkeley’s Dawn Song and Oasis Labs are investigating technology which could help. I’m sure there are many others.
In thinking about what these sudden deplatformings portend, it's worth examining why they happened. It's easy to see the attempted coup as the reason Facebook, Twitter, Google (and soon Apple?) suddenly deplatformed Trump and his mob.
But that timeline misses an essential contributor--the runoff election in Georgia on Tuesday that give Democrats control of the Senate. Suddenly facing a blue government, Facebook, Twitter and Google acted either cynically—knowing bans on Trump and Parler might lessen regulatory heat—or with relief that they could finally remove odious speech from their platforms. Without Georgia, it's possible the tech companies wouldn't have needed to (or dared) deplatform the crazy right.
Having written that, I just read Heather Cox Richardson's similar take this AM: "While Twitter officials might well be horrified by the insurrection, the ban is also a sign of a changing government. With the election of two Democratic senators from Georgia this week, the majority goes to the Democrats, and McConnell will no longer be Majority Leader, killing bills. Social media giants know regulation of some sort is around the corner, and they are trying to look compliant fast."
Until everyone is able to disagree without being disagreeable, everyone could benefit from consistent standards. Free speech has limits and social platforms are best when they pay attention to those limits and stop short of testing them in my opinion.
Let's push for two labels on all posts. One label would be that this is a factual statement while the other would be this is the writer's opinion. If all posts could be delayed by some very short period of time, I think technology could do a reasonably good job of labeling all the posts. Of course, if someone wants to post immediately and label what they wrote as their opinion no delay is necessary. And, any posts determined to be hate would not be allowed, come with a suspension or taken down promptly with a suspension if posted immediately as an opinion.
I’ve scanned the comments as of 15:15 GMT and I don’t recall anyone yet commenting about transparency of (a) business model and (b) forensics for revenue.
Most Facebook users are probably not aware that their personal data is a key input and their attention is the product sold to advertisers. Similarly I’m not sure most small businesses and microbusinesses who rely wholly or mostly on Facebook as a *free* platform on which to build their public digital presence form a class of economic agents upon whose participation Facebook’s revenues rely. Consumers and businesses both need better education to make decisions about their data and ad spending that are congruent with their personal values. Many may stay but some, I suspect, would spend less time there if these mechanisms were made plain.
I’m reminded of Ursula Legion’s short thought experiment (1973) called "The Ones Who Walk Away from Omelas"
Values and decency should be universal. However, that "should" is either inappropriate or too aspirational. How about inviting all institutions public or private to post, for all to see, the values they aspire to live into. And let reality score the success or failure of their aspirations. I'm a fan of consistency and as a result think Facebook should do the same thing wherever it operates based upon their values. The one thing that I am stronger about is a clear delineation between facts and opinion. All are welcome as long as they don't incite violence or hate, but all do not have a right to be weaponized.
Private companies should not be able to limit freedom of expression (especially when millions of people decide they would like to listen). If we need bias of any sort this should be governed by democratic institutions. Why does Bolsonaro get censored on Twitter for his policy against Covid19 (herd immunity) and not Boris Johnson for the same (at the time) policy.
Not agreeing with someone is the best starting point to disagree with their censorship. Yesterday, an emblematic fact occurred: the most popular social media platform in the United States preemptively censored the President of the United States for the next two weeks while waiting for the new President to take office.
The reasons can also be understood after the assault on the US Parliament during the confirmation of the new President in charge, which was triggered by an invitation from the outgoing President to protest in front of the doors of Capitol Hill.
But what should give pause for thought should be the fact that today there are democratic protections in place so that everyone can speak and say even things not deemed correct by some, or even many. Freedom of speech is also the freedom to say what one does not want to hear, even if it should be a crime of opinion. Except then be sanctioned if it is committed.
To deny the possibility to a person to speak is the prerogative of the constituted democratic power.
Until today Facebook, like many other social networks, has qualified as an independent software platform, but today perhaps they should qualify as an editorial company, taking responsibility for everything that is made public and specifying on each occasion why a post is tolerated and another is not. If Putin or Xi Jinping were to make statements against U.S. interests or those of the social media, would Zuckerberg decide if it was appropriate to censor them?
You can disagree with someone and even fear their opinions. However, censorship brings with it consequences if dispensed outside the democratic context, especially if linked to millions of people who legitimately choose to follow a person, especially and even more so if done on a tool that has become central to the comparison of ideas.
1- The first consequence is that social networks will become biased and people will start to use the social networks that best represent their opinions. And as always, this a priori polarization will not facilitate the confrontation of ideas, but will crystallize even more any political fan base. It is no coincidence the success of the new social media Parler (https://bbc.in/2L3PmtR) within which are channeling the Americans excluded from the debate on Twitter and Facebook and in general millions of Trump supporters. A social media divide.
2- This division also risks becoming a division between states. What would have happened if it had been TikTok, the Chinese social media outlet, that censored a U.S. president? Censorship mechanisms are already in place (https://bit.ly/3hYsNm7), but applied only to Chinese domestic political contexts. The problem in reality is that Facebook and Twitter already apply political censorship in other countries outside the United States, creating not only a short circuit between politics and private companies, but also a diplomatic one. As it happened after the protests in Hong Kong where Twitter identified and suspended 200 thousand Chinese accounts (https://bit.ly/2XnMzy4) and YouTube suspended over 200 accounts (https://nyti.ms/2LbeqyQ ) accused of propaganda against the autonomy of Hong Kong. The phenomenon is however global and continuous if you think that Facebook closes 3 to 5 billion accounts per year (https://cnn.it/3om11Cx ). Probably most of them are right to be closed, but if even one account was closed without valid reasons it would have been taken away from Facebook or Twitter by a company.
Twitter has also censored tweets of heads of state such as Bolsonaro's when he chose to deal with the Coronavirus with herd immunity and not lockdown describing in his opinion its uselessness (https://bit.ly/2XkfuCX), but has not given the same treatment to the British PM when he had the same positions.
3- Third-party guarantee commissions set up by the social media themselves, for example Facebook's Oversight Board (https://oversightboard.com/), end up being only a mirror for the larks if the owner of the social media of the moment decides and declares censorship, as in this case Zuckerberg (https://bit.ly/3s5ZJOk), then rectified after a few hours by a company communication that reduced the account block to 24 hours (https://bit.ly/3hV9xG8).
Not to mention that preventive censorship seems more similar to the dynamics of preventive justice in the style of Minority Report than to the application of certain and equal rules for all.
In reality, the problem of false information has not yet been solved even in the traditional media. If I think about what has been the biggest fake news of the last thirty years I think about the weapons of mass destruction owned by Saddam Hussein, a false news that has had a devastating impact also from the point of view of the consequences that it has determined and that has been proposed and relaunched by the traditional media.
These days there is much debate about Bitcoin gaining value as a result of the general devaluation of state currencies. The reality is that blockchain technology, on which Bitcoins are based, is leading to the creation of new self-organized and distributed realities where there is no central control and within which people can act within rules established at the beginning and above all verifiable by all. The verifiability of decisions and the distribution of decisions is emerging as a model opposed to the centralized one based on arbitrary choices that are not shared. Maybe a social media based on these principles will be born soon and could solve the eternal dilemma between truth and freedom of speech.
Indeed my 12-year-old daughter's commentary upon hearing the news that Trump was banned from Twitter: "I guess we'll then have to deal with him on TikTok."
Tough call. On one hand, the platforms are meant not to filter content according to Section 230. Therefore they can post broad guidelines, but that's about it. On the other hand, those provisions were designed twenty years ago, before the content platforms achieve near-ubiquitous and near-utility status. It's a fair argument to say they need to be managed differently, however a global explosion in creativity and communication has been facilitated by that approach. Can you minimise negative content without damaging positive content? Especially given that the judgement is profoundly qualitative.
Regarding the events at the capital, a liberal colleague said, "if they really feel that the election was stolen and democracy is being threatened, wasn't that the right thing for them to do?"
It's easy to point the finger at the platforms and say "it's time for you to be regulated" and that will probably happen. However, the reality is that Donald Trump was already impeached by the House and let go by the Senate on a strictly party-line vote. Because at the time, the Republican Party figured that the opportunity for the party was worth the cost of Trump. In fact, at each turn of the Trump administration toward greater degrees of autocracy and abuse, the Republican Party pretty consistently compromised core values for the opportunity to cement things like court justices.
If anything, the safe harbour provisions worked, in the sense that the social networks enabled quite high-fidelity communication between individuals and within/between communities. Everything that the Trump administration did has largely been done in plain sight, in part thanks to the the platforms. While we can kill the messenger, so to speak, this process has shown numerous cracks in the fundamentals of government. Which is in part what science and technology are meant to do. Challenge convention and enable greater understanding of how systems operate.
Whilst Zuckerberg and Dorsey finally had to act, the reason that they had to act in part was due to underlying failures of governance. Or perhaps the success of the system. If we're saying that some people are too stupid to think for themselves and they believe everything they see on the Internet, each side can make that argument. Liberals can say that Trump hijacked the minds of disaffected Americans. Conservatives can say the media at large is destroying the values of American society. Because fundamentally there are at least two Americas now.
Maybe one could argue the platforms helped facilitate those splits and atomisations, however Section 230 was designed to shield the platforms from liability and enable the growth of the network. And it has worked. If we don't like the result, it's worth considering that it's not that the platforms need regulation but perhaps they have created a more transparent view of the underlying society. And it may be that which we do not like and can no longer tolerate.
Regulating Facebook is not going to solve growing systemic inequality in the face of artificial intelligence. It may obscure it to some degree and potentially reduce its appearance, but that is not at all the same as directly addressing the drivers and impact of what led those people to the capital on that day.
Nicholas, is it an either/or? What if social media were not simply displaying the societal/cultural fault lines but also exacerbating them?
Sure, governance structures in the public realm have been revealed/reiterated as faulty. But manipulation through segmented communication on a vast scale has been made affordable and practical by social media in a manner that your description seems to overlook.
I appreciate @Sam Barcroft’s reply to your comment and seek here to extend it.
"if they really feel that the election was stolen and democracy is being threatened, wasn't that the right thing for them to do?"
As a progressive liberal, I agree with this. The problem is that there was no evidence of the election being stolen and there shouldn't be this many people who believe it was. So the problematic bit of this saga is not the final action of the events at the Capitol, it was the spreading of the radicalizing misinformation in the first place.
I disagree. An effectively regulated social media landscape will be fantastic in shining a light on systemic inequality, by embracing guardrails that ensure honest reporting rather than deliberate misinformation.
Parler will not make a difference. It is hard to run a decent service, they won't get the talent. Without moderation, it'll turn truly horrible and for four years the Feds will just use it to track white collar terrorists. For normie on Parler, I can't see the product ever being attractive enough relative to Snapchat or Tiktok of even FB to get them to really stick around.
Yep, usually people run from platforms to get away from the hardcore wing nuts that come with scale. Parler is a unique example of the hardcore wing nuts running away to a new platform themselves.
Lots of thinking in development orgs and multilaterals at the moment about platforms as civic infrastructure, but unfortunately too much of it still framing platforms as “digital public goods”. We wrote a series of internal policy papers for DFID/FCDO last year including one on digital harms as an attempt to counter this anachronistic thinking. But there hasn’t been a lot of policy engagement of the negative side of social media platforms outside of US & EU, largely because for a decade or so extending access to internet (which in emerging markets basically means social media) was perceived to be something that raised economies and liberal societies - although without much impartial evidence of this, as the World Bank’s WDR 2016 report on Digital Dividends points out.
But what happened in the US this week has already happened in other countries, notably Myanmar but also many others, to limited response from FB et al. But on their doorstep, with a new incoming administration, they couldn’t not react. I agree with Sam here that this is too little to late - what we label as Q Anon conspiracy theories are actually just radicalisation, which is what we call distribution of the same content if it’s Islamist. It should’ve been halted years ago, but the platforms are unwilling to step outside of their get out of jail free platform cards and admit they’re media orgs. They’re happy to take the income of a media org, but not the responsibilities. They will eventually police their platforms in the US because they’ll be forced to, but I can’t see them doing this in the long tail of other markets who contribute less. US is >50% of FB, Twitter revenues. “Rest of World” is tiny in comparison, and much harder to police because the language and culture of emerging markets is less well represented in ML training data, and therefore much harder to manage using AI, and revenues aren’t high enough in these markets to support human moderation costs.
I gave a really depressing presentation to a World Bank group last month that played out the emerging market revenues of social media platforms, their ubiquity in people’s digital lives in these countries, and the lack of knowledge and motivation platforms have to police them properly outside of US and EU. It ends by pointing out we’re likely to see a Rwanda-style genocide at some point in the next few years driven by manipulation of social media in these markets. Arguably we’ve already seen it with the Rohingya.
Chris, I don't quite get the issue with your "digital public goods" point. Why is that framing not right?
There’s often a simplistic sense that digital infrastructure is like physical infrastructure, that you lay some out as public good and public benefit derives from this. My old employers the GSMA compounded this thinking with the doubtful claim that for every 10% increase in mobile subs GDP increases 1.4%, and this has been compounded by a Facebook funding research that claimed for every 10 people you connect to Facebook you pull one out of poverty. This has reinforced existing simplistic development narratives on how technology impacts on poverty and societies - the main issue is seen as inclusion, and getting to a magic number of connections per population, rather than a more nuanced look at what people are actually connecting to, and what their digital lives actually are. So the focus is on funding a digital stack, often taking India’s as a template, in the belief this magically leads to positive outcomes and GDP increase rather than engaging with how digital economies and societies are really emerging on the ground.
We’ve successfully challenged this with many clients by getting them to focus on outcomes such as Meaningful Use rather than Inclusion, which considers what digital platforms users are engaging with and whether these are net positive and/or negative for their incomes and livelihoods. Our senior director of research Jonathan Donner frames this well in his 2015 MIT Press book “After Access”.
Ok. In a sense the "digital public goods" can easily have these negative consequences that emerge from their design?
Yep. So the India digital stack is built around Aadhaar as an ID platform. This has clearly been a public good and has driven rapid adoption of digital payment services enabled by eKYC. But it’s also been used to exclude minorities in some regions. We mapped the digital identity ecosystems in Kenya and Ghana for the Australian Government to get a broader political economy sense of what impact they would have on digital economies and societies, and this allowed us to successfully predict that Huduma Namba would exclude millions from digital ID in Kenya and actually exacerbate digital divides. https://link.medium.com/ft4DgxTCScb
Sorry if I’m going on too much! And we’re some way from the original topic.
This is really fascinating, Chris.and speaks to my current concerns as we try to reframe the BBC as a guarantor of a digital public space rather than a broadcaster with public service remit. I absolutely agree with your point that just providing access does not guarantee good -or even acceptable - outcomes. Thinking about the services you'd want, and how you would assess their impact, is as important as finding ways to move the bits around.
I think a lot of this thinking has been done already by mutual friends of ours (Tony Ageh, Mike Bracken, Tom Loosemore). My twin brother Matt has a nice framing of the public media stack as a way to identify and encourage the use of what are clearly definable digital public good platforms https://publicmediastack.com/
Do you think that will lead Facebook to pull out of some countries where the liabilities > revenue opportunities, or will they still operate in those countries but with no standards?
It’s a tricky one. There are already many countries where we suspect their cost to serve FB customers is greater than the revenue they earn - by our calculations from a couple of years ago maybe most of the last billion users they added are probably not profitable, based on revenue we modelled and the extra costs to serve users who are on lower end devices with no JavaScript or cookie tracking, or are just using WhatsApp, etc. The report is here https://link.medium.com/cSku5ZqoUcb
We suspect this is why Libra was developed. People forget quite how unique the US market is in terms of advertising revenue market size. It’s totally unlike even EU markets, and when you look at India, SE Asia, Africa these are transactional markets rather than attention revenue markets. Pivoting a product based on 99% of revenues coming from advertising to one that can become profitable from transactions is hard, and we suspect Libra was a punt at this, which has failed already. Getting into financial transactions is costly, requires a solid ground game in emerging markets (lack of bank accounts means you need networks of agents for cash in/cash out to digital payment platforms) and brings you into direct contact with regulators. All things FB is allergic to.
Every year I expect the bubble to burst and an analyst to properly ask where the revenue is for the last billion users FB added. And every year ad revenue growth in the US bails out their figures. But at some point these revenues will plateau, and the question will become more immediate. FB had to work out how to monetise WhatsApp users in emerging markets in particularly, and it has the opportunity as our research has shown how integral WhatsApp is to informal trade and marketplaces in emerging markets. But turning informal discovery into a trading platform is still hard.
And also the real challenge for GAFA firms in emerging markets is, will Ant Financial get there first? It’s a battle between Silicon Valley firms with all the reach but no ability to get transactional revenue, and Chinese super platforms that have all the experience in transactional revenue from low income users but none of the reach. This is why Ant has been buying into SE Asian mobile money platforms like it’s going out of fashion (Wave Money, Bkash etc.) - they need to buy the reach before GAFA works out how to turn their attention reach in these markets into transactional revenues.
This and your last comment are beyond helpful in terms of data-driven and informed perspective. Thanks for sharing yours here.
If Azeem doesn’t mind me posting it here (it’s subscription free! Not poaching EV subscribers) our subsidiary company DFS Lab did a short 6 episode email primer on emerging market fintech for investors. A useful primer - https://us13.campaign-archive.com/?u=665090c913eeda856cec81ec8&id=bb97098b0c
Many would say this action was six months late. The timing was interesting - coming on the day the Democrats took all three branches of power. And great pressure from Republicans before the 2016 election meant the company were nervous of allegations of censorship. But Zuckerberg wields ultimate power and seems indifferent to the toxicity of Facebook's approach thus far. The share price has grown steadily through the year, and his individual power with it. He can fix this whenever he likes. The reality is that legislation is the only sure way to adapt the way social media companies operate. They have to be treated as editorial publishers, with the same responsibilities. Facebook - like Google - is an advertising sales platform, powered by the ethanol of emotive editorial content. This carries great responsibilities. It's time for legislators to adapt and effectively regulate.
Do you think this means national legislation in each country, and then local compliance?
I do. As recently as 2019, Zuckerberg publicly requested new regulation and oversight. This effort is already underway in the UK and elsewhere. A global regulatory approach would be the most practically efficient for the publishers, but unless this is voluntary - like IPSO in the UK - I can't see a legal framework for it. This would also save the publishers a great deal of fragmentation expense. With Facebook et al publicly seeming unwilling to solve their own problems, sadly, this will become a multinational lawyerfest which will ultimately reduce the profits and the power of these companies in the longer term.
Totally agree regulation is coming, but it will take a long time to extend outside of the higher revenue US & EU markets into the countries that are really vulnerable to social media radicalisation because of their weaker civil societies. Over time these markets do copy US & EU regulation, but choose to ignore it where it benefits Government. My old friend Tom Wheeler at Harvard’s Shorenstein Centre writes and talks well on this topic of how FCC policy does, and doesn’t, roll out globally for social media.
Maybe Tom will reply to this thread! Hi Tom!
Most other industries are regulated locally, aren't they?
Indeed - and that is what will happen. The big shift from governments has to be to disallow the "we aren't publishers, we are just a platform" approach, and to enforce the "we are publishers and advertisers, responsible for what we publish and advertise". Millions of practical and philosophical reasons why FB will struggle with this, but its the right thing for society, and therefore longer term, for FB and their shareholders. Otherwise, they risk increasing reputational decay. Zuckerberg must feel trapped, otherwise he wouldn't go public on wanting more regulation. Hopefully now this gets going at pace and we move into the next phase of social media - less Wild West, more trustworthy, more respected, more useful.
My current homework is to read the Digital Harms legislation going through Parliament and I haven't yet done this. However I don't see why all politicians and political campaigns are either banned from Facebook, after all the White House has it's own platform as does 10 Downing Street and their followers could follow there - in the open! Or they are banned in the 28 days preceeding any elections at any level - there is precedent for this for other political campaigning and advertising in different places. That would be the easy part. Activity in private groups is harder to deal with. I do think the platforms should be regulated as publishers are regulated, and people who post subject to the law of libel, race hatred, incitement to violence wherever they post from. But I guess I will have more to say next week!
Totally agree. The digital harms paper is good, but very broad. Will be easy to dodge turning it into the regulation that’s needed.
This sets many precedents, in my opinion. It shows the power a world leader can build through the use of social media. It shows the lengths social networks must go to police world leaders' social media activity. And it paves the way for lawmakers globally to impose tighter regulations. To date, only in Brazil and Venezuela have we seen a leader's social media posts removed by the social network, as far as I can see from looking into cases. I think however that we will now see far greater proactivity from social networks in this space, and more attention from the global press on this as an issue.
In Myanmar there were bans but after the genocide.
Lots of great thinking in these threads. One aspect I did not see explored as much as I believe is warranted is the competition law vs. freedom of speech basis to challenge/regulate the large social media platforms.
At least in the US, it is quite settled constitutional law that owners of private forums have no obligation to permit all comers to speak. There was a long line of First Amendment caselaw concerning the rights of shopping mall operators to exclude free speech proponents who argued that in the middle of a Minnesota winter, standing on a soap box in a privately owned mall was constitutionally equivalent to Speakers Corner.
To me the argument is not that I should have the right to enter your property or participate as a matter of law in your online community, the societal interest (and the concomitant legitimacy of government in a free market economy to regulate) lies in the size and reach of that platform. If it is so large and dominant that there are no other effective platforms to engage in public debate, then these “essential facilities” need to either be broken-up to allow diverse competition, or regulated as was the case with the early “Fairness Doctrine” that required the holders of broadcast licenses in the US to provide balanced and equitable access to their “platforms.”
The Fairness Doctrine, like Section 230 was not based on competition law, but the power of the FCC to set rules governing the use of “public airwaves.” In the case of FB, TWTR etc. the underlying internet is not regulated as the original broadcast spectrum was and the doctrinal justification for regulation could be taken from competition law. These laws obviously differ from country to country, but concepts such as dominance and abuse of market power translate.
Finally, it should be noted POTUS has always had an effective pulpit from which to be heard: the traditional and official White House Statement or press conference. He should not be heard to complain about deplatforming when he has done so much to promote hate speech and incite violence.
So I think the issue here, and am writing about this in the "you know what", is that we need to reframe anti-trust and competition law to a broader frame that includes the moral and civic space, that the effective privatisation of these new forms of the public space (in the Habermas sense) is a process that warrants the purview of governments on our behalf. In this context, whether as these networks get bigger they do so while maintaining public and civic health.
I have a naive question. If everyone posting to a social media platform is doing so as an individual private citizen, then all can be subject to the same rules against hate speech or inciting violence. Governments could choose to adjust their codes of practice such that officials cannot use social media platforms in an official capacity at all, but the social media platforms could also choose to prevent that usage. Then there would be no privileged @POTUS account, and the @realDonaldTrump account would have no protection from bans. The Twitter algorithm, instead of warning about misleading information, could politely inform government officials to announce their policies on their own government web sites. Implement manual moderation on every government minister or leader across the world, and stop them tweeting policy (e.g. - you can send out a link to an announcement, but not make the original announcement on social media). Problem solved :)
I think one issue was the reverse logic that Twitter and Facebook used, which was to hold public officials (aka Donald Trump) to a lower standard than they would hold other users.
In most places, we expect public officials to behave with greater probity (think about the Parliamentary Standards).
It would have been better for them to adopt a clear standard and happy it very consistently.
It would also make sense to have a sliding set of standards, perhaps as accounts get more followers they might qualify for different (tougher) objective guidelines.
I am sure that twitter went through a process of clarification that was meticulous and involved many lawyers and experts, but ultimately it is an act of fiat, however we describe it.
I said the same thing this morning!
1) How does it play out? From
From Facebook side: There is minimal incentive to apply this approach elsewhere. On home turf, with a lame duck president, and at least four years of corrosive disinformation and lies, they acted only at the very very last minute possible, with the whole world looking on. That circumstance won't repeat anywhere else with same dramatic urgency. Facebook has never shown any real thought on this, and as with Rohingya crisis, only acted when they couldn't not. Removing Trump would only be meaningful beyond the gesture if it was followed up with a root and branch removal of the disinformation spread in his name whether by real people, malicious political actors or AIs. The scale of that work, and the bottom line hit it will mean means there's very little disincentive to do that, and hard political regulation will be very hard to achieve.
From nation state side: Very messily. Only states, or supra-states like the EU, can legislate to provide protections against the mess Facebook and other social media causes. There is no way that legislation can be created or implemented with any evenness across the world. Practically, if the EU, the US and China went after a break-up of Facebook they could in principle unbundle the products, but creating a consensus set of regulatory ethical principles seems super-unlikely, particulary as all these countries try to learn from Russia on the wheres and hows of info war.
There are historical analogies to FB’s dominance of its current day market in things like the East India Company which ran the bulk of world trade over a period spanning centuries and employed private armies to enforce trade terms on various countries, presumably arguing that consumers back home ‘valued’ the access to foodstuffs and opium until the excesses became so egregious it was nationalised and shut down, replaced by government rule. Competition law goes back to at least Roman times too, but usually took the form of limitations on trade practices disadvantageous to the ruler or monarch, by driving up prices on eg. basic foodstuffs to the extent that public discontent would erupt. Most current competition law regimes still assume that the playing field is ex ante level, oddly enough. In searching for a local or national regulatory response, the lack of a country that has solved the thorny issues many have raised here in this discussion to use as a leading light example means that governments will each have to struggle with laborious, thinking slow (a la Kahnemann) type of legal reform that takes years.
I suspect that Trump is a one off, and that they will resist doing this on a wider scale as they want to avoid responsibility, so I doubt they will act without an outcry. To the extent that the platforms can be held accountable, I suspect they will seek to use local standards, as otherwise they will be subject to oversight into how they police transgressions. I suspect that legislation will be required to force the companies to make changes, though if it's not done before the next election cycle, I doubt it'll be done at all. At least in the US.
The recent decision from Facebook to ban Donald Trump from posting on its platform is exceptional and context-specific; not globally applicable.
At the end of the day, Facebook is a (profit-oriented) company and will be proceeding accordingly strategically and tactically. Companies create their rules to support their mission and interest. So, no law and especially no company rule of the kind will be uniformly interpreted and executed the same way for every applicable case. If it does not fit to the current context of interest, the problem at end will be declared to be an exception, which leads to an exceptional rule.
Even law-makers and governments - while acting in the context of their mission - are still shaped by self-interest, power preservation, party, clientele, lobby, etc.
Although Facebook’s - and similar platforms - influence and public impact have taken on worrying dimensions, I don’t believe that any global standard will be applicable here.
Social medias like Facebook reflect the system in place and the underlying society with all its good and the ugly things; in particular, in terms of freedom of expression. The problem is not about regulating Facebook but re-defining and better understanding what we mean by the precious good name freedom of speech/expression. The same applies for a media company, its role, responsibilities, limits, etc. We need a new category of media company and a corresponding regulation framework that is suitable to Facebook & Co.
The point is, justice requires treating equal things equally and unequal things unequally. But who says what is equal and what is unequal? And who determines or establishes whether treatment is equal or unequal? That’s the challenge and injustice we have to cope with everyday.
There is an issue with double standards: why should politicians be exempt from the social media guidelines when other users are barred for misconduct? Why are television and telecommunication providers subject to public oversight and have to adhere to different rules than the social media?
Big tech empires that today have an even stronger impact on public opinion than classic media (boosted by personal data analysis and hyper targeting), and they do need public oversight because corporate voluntary self governance has never worked.
Unlike on TV or printed press, much of political influencing is completely hidden in social networks. An auditable trail of political advertising campaigns (as on TV or printed press) is particularly important in times of filter bubbles to hold politicians, parties and other political influencers accountable to established standards.
It seems that, today, local legislation - see EU or China - is most effective to make social networks adhere to guidelines (even though we may not agree with some national variants). However, given that big tech are supranational organizations and increasingly powerful, I am not sure if a national approach is sufficient in the future.
Govern private platforms as little as possible. Concentrate on opening up more platforms and finance that by taxing the giant profits of the big players FB, Google, etc. Obviously, that is not simple, but here are four simple steps to get started. Taking some billions from FB and Google, 1) give local public radio solid support -- say 30x what they get now; 3) subsidize existing local libraries so they can afford more branches and have local facilities for podcast and video production; 4) subsidies local and small scale publishing operations such as local newspapers, small scale publishers, etc. However, hate speech should not be allowed on social media. How? Start by studying how Germany has handled their Nazi's since 1945. One way or the other, we need to to see to it that a lot of journalists are hired to produce content not just repackage others.
Thank you, Blaine. I needed to read someone widening the aperture. People need information diets that don’t rely on a single channel or format. You highlight how breadth can be restored.
The ethical dimension underpinning the whole system is this: what’s moral is what’s legal and what’s legal is for sale. Instead, think about the $100 million Facebook has spent in DC lobbying in the last 6 years. The only difference between public and private is temporal. How long will it take for Facebook to drill into the legislative marrow and insert its DNA.
Or extract our DNA!
The DNA of your intangible net worth. "Facebook's Conduct Shows Not All Network Effects Are Positive" https://simonschase.co/facebooks-conduct-shows-not-all-network-effect-are-positive/
This is the key ... and has been for the last 5 years "They have to be treated as editorial publishers, with the same responsibilities. Facebook - like Google - is an advertising sales platform, powered by the ethanol of emotive editorial content." So sad that regulators have been so very slow out of the blocks. And we also need MUCH more focus on getting these big techs to pay taxes in the countries they are serving.
I'll share an observation that has me questioning my usual assumption to follow the money and the business model to get to the underlying problem inherent in platforms.
During the past 4 days every call I've been in composed of eager hyper-networked powerful let's-do-something Americans has ended with a call to see who has one of these men on speed dial: Mark Z, Jack D, Rupert M, Brian R. (Comcast/NBC) and then Mike P. Chuck S. and the one woman in the mix: Nancy P. In many cases someone on the call did have one of these folks on speed dial, and we then resorted to figuring out the talking points.
It's as if through all of these advancements in exponential and less exponential technology we've just recreated vast feudal kingdoms and our only hope for better governance is to appeal to the kings. The difference between now and then is that the kingdoms are interconnected in complex adaptive systems.
Across all three platform types: social media, left/red entertainment television, and the fortified executive wing - we humans (or the subset of humans called Americans) have created the same conditions. While we can fiddle away as philosophers and technocrats to pull and push the levers of social media systems, our selfish genes keep recreating this structure no matter the tech.
Time to rewatch some good Adam Curtis. Monkey in the Machines of Loving Grace and all.
Lots to think about indeed… A core and (to my astonishment) still overlooked issue is the lack of understanding of information within our society. By that I mean the fundamentals - what is information, what are the laws governing its storage, propagation and (most importantly) value. By ‘value’ I don’t mean the sometimes trivial copyright/IP view, but the true impact (or lack thereof), direct and indirect, its true economic value and political consequences. We live in a world of energy and have a decent grasp of laws of thermodynamics. We claim to live in the information age, but we do not have a clear understanding of what (scientific) laws govern the flow of information. We teach the laws of thermodynamics, we teach nutrition, we do not teach the laws of information.
The information flow patterns have undergone a radical evolution, far outpacing our brain evolution - many historical limitations of space, time, persistence and volume have been removed. This has perturbed our human networks and, like any change in the environment, has led to new (side) effects - from new media to new business monopolies to digital mobs, some awesome, some questionable.
What we seem to be struggling with is a framework for understanding the societal (and individual) impact of information in a manner similar to healthcare. We have made good progress in understanding the information flows within biological systems, from molecular biology to population genomics. We figured out how physical messengers (chemicals, food, pathogens etc.) affect individuals and society and built entire sub-fields of medicine around them (environmental impact, nutrition, hygiene, epidemiology, healthcare economics, ...). What seems to be needed next is a similar approach to information, that might lead to a better understanding of our information environment, information economics, information hygiene and allow us to tackle some core issues like information value, monopolies, censorship, etc., many of which lead to familiar healthcare moral dilemmas of public vs. individual good, ownership and control.
A controversial comment - do we remove anonymity and make verification of ID a condition for open social media? Would increase the risk for people using social media to challenge abusive Governments for sure, but I think using international civil society architecture to hold Governments to account for abusing their regulation over platforms and free speech easier than playing whack-a-mole with bot armies and fake new accounts. Maybe worst of potential scenarios to keep social media open and viable.
Verified ID could make sense as one alternative. Not clear that my mum needs to be on a platform that doesn't verify. Parler practically verifies (with phone number, doesn't allow burners).
There can still be an ecosystem of unverified services for the remainder.
Thanks Chris, for making this point. I also had this thought many times before. I feel any public space needs an identifiable public (not a bunch of semi-ironic avatars) to build a baseline of language, values and norms. So people can build upon that to make decisions for the common good. I mean what is a community if you never even see your neighbors?
But Facebook (which seems to be the most problematic of the major networks) already has a "real name" rule that is mostly followed. One of the things that continues to surprise me is how little that seems to matter. Many of the Q followers seem OK with posting as themselves and the expected social consequences (ostracism, job loss, etc.) seem to not be happening.
Valid point, although I doubt that Facebook is where most of the action happened that led people to the Capitol.
Also, on a different note, I suspect that although Internet culture flowed over the barricades into the heart of politics last week, it was those political barricades that led them towards internet culture in the first place. In other words, I feel that this isn't just a tech problem, it's a societal problem. Borne out of the many fault lines that are being aggravated in our rapidly changing society. Perhaps a technological fix isn't enough to stop radicalization - we also societal fixes.
Unfortunately, this narrative plays into Donald's hands and his supporters will lap it up. "The world is against me and is trying to silence me. Rise up and stand with me against such evil tyranny!"... oh the irony.
I think censorship is good except when it isn't.
rotfl
After the 6th, private platforms did a better job of reflecting the public interest -- including support for constitutional process -- that most public institutions have so far. Excellent work has been done on the intersection of social media and some law enforcement.
Prior to the 6th, it was a different kettle of fish.
In most countries the need for such rapid action would not have been so great -- in the US we have an uncommon number of heavily armed people. As noted elsewhere, I think we need a lot more platforms, many more public spaces, etc. We also need to use education to equip people to participate in social media on a different level.
Business models which are based on maximizing engagement seem to create highly reactive conversations with little time for reflection, investigation, etc. We can probably help establish more reflective, conversational norms internationally, but locally I think we need to develop new business models and restrict existing ones. If the model is centered around selling personal information, the model will value engagement and that leads to frantic, reactive conversations.
More than 150 years ago we took strong steps to ban the sale of humans. I think we need to ban or eliminate the sale of human data as well.
The sharing of human data is a different issue. We are unlikely to deal with pandemics well or have safe civic discourse w/o safe sharing of personal identity and data. One can imagine competition to develop platforms which support a new class of public health apps -- apps with international value -- while preserving identity and personal location information. I think UC Berkeley’s Dawn Song and Oasis Labs are investigating technology which could help. I’m sure there are many others.
This is a valuable discussion. Thanks everyone!
In thinking about what these sudden deplatformings portend, it's worth examining why they happened. It's easy to see the attempted coup as the reason Facebook, Twitter, Google (and soon Apple?) suddenly deplatformed Trump and his mob.
But that timeline misses an essential contributor--the runoff election in Georgia on Tuesday that give Democrats control of the Senate. Suddenly facing a blue government, Facebook, Twitter and Google acted either cynically—knowing bans on Trump and Parler might lessen regulatory heat—or with relief that they could finally remove odious speech from their platforms. Without Georgia, it's possible the tech companies wouldn't have needed to (or dared) deplatform the crazy right.
Having written that, I just read Heather Cox Richardson's similar take this AM: "While Twitter officials might well be horrified by the insurrection, the ban is also a sign of a changing government. With the election of two Democratic senators from Georgia this week, the majority goes to the Democrats, and McConnell will no longer be Majority Leader, killing bills. Social media giants know regulation of some sort is around the corner, and they are trying to look compliant fast."
Until everyone is able to disagree without being disagreeable, everyone could benefit from consistent standards. Free speech has limits and social platforms are best when they pay attention to those limits and stop short of testing them in my opinion.
Let's push for two labels on all posts. One label would be that this is a factual statement while the other would be this is the writer's opinion. If all posts could be delayed by some very short period of time, I think technology could do a reasonably good job of labeling all the posts. Of course, if someone wants to post immediately and label what they wrote as their opinion no delay is necessary. And, any posts determined to be hate would not be allowed, come with a suspension or taken down promptly with a suspension if posted immediately as an opinion.
I’ve scanned the comments as of 15:15 GMT and I don’t recall anyone yet commenting about transparency of (a) business model and (b) forensics for revenue.
Most Facebook users are probably not aware that their personal data is a key input and their attention is the product sold to advertisers. Similarly I’m not sure most small businesses and microbusinesses who rely wholly or mostly on Facebook as a *free* platform on which to build their public digital presence form a class of economic agents upon whose participation Facebook’s revenues rely. Consumers and businesses both need better education to make decisions about their data and ad spending that are congruent with their personal values. Many may stay but some, I suspect, would spend less time there if these mechanisms were made plain.
I’m reminded of Ursula Legion’s short thought experiment (1973) called "The Ones Who Walk Away from Omelas"
Values and decency should be universal. However, that "should" is either inappropriate or too aspirational. How about inviting all institutions public or private to post, for all to see, the values they aspire to live into. And let reality score the success or failure of their aspirations. I'm a fan of consistency and as a result think Facebook should do the same thing wherever it operates based upon their values. The one thing that I am stronger about is a clear delineation between facts and opinion. All are welcome as long as they don't incite violence or hate, but all do not have a right to be weaponized.
Private companies should not be able to limit freedom of expression (especially when millions of people decide they would like to listen). If we need bias of any sort this should be governed by democratic institutions. Why does Bolsonaro get censored on Twitter for his policy against Covid19 (herd immunity) and not Boris Johnson for the same (at the time) policy.
P.S. I went in depth with an article I published today (Italian): https://www.facebook.com/Casaleggio.Davide/posts/248375173320506
Automatic translation (DeepDL):
Not agreeing with someone is the best starting point to disagree with their censorship. Yesterday, an emblematic fact occurred: the most popular social media platform in the United States preemptively censored the President of the United States for the next two weeks while waiting for the new President to take office.
The reasons can also be understood after the assault on the US Parliament during the confirmation of the new President in charge, which was triggered by an invitation from the outgoing President to protest in front of the doors of Capitol Hill.
But what should give pause for thought should be the fact that today there are democratic protections in place so that everyone can speak and say even things not deemed correct by some, or even many. Freedom of speech is also the freedom to say what one does not want to hear, even if it should be a crime of opinion. Except then be sanctioned if it is committed.
To deny the possibility to a person to speak is the prerogative of the constituted democratic power.
Until today Facebook, like many other social networks, has qualified as an independent software platform, but today perhaps they should qualify as an editorial company, taking responsibility for everything that is made public and specifying on each occasion why a post is tolerated and another is not. If Putin or Xi Jinping were to make statements against U.S. interests or those of the social media, would Zuckerberg decide if it was appropriate to censor them?
You can disagree with someone and even fear their opinions. However, censorship brings with it consequences if dispensed outside the democratic context, especially if linked to millions of people who legitimately choose to follow a person, especially and even more so if done on a tool that has become central to the comparison of ideas.
1- The first consequence is that social networks will become biased and people will start to use the social networks that best represent their opinions. And as always, this a priori polarization will not facilitate the confrontation of ideas, but will crystallize even more any political fan base. It is no coincidence the success of the new social media Parler (https://bbc.in/2L3PmtR) within which are channeling the Americans excluded from the debate on Twitter and Facebook and in general millions of Trump supporters. A social media divide.
2- This division also risks becoming a division between states. What would have happened if it had been TikTok, the Chinese social media outlet, that censored a U.S. president? Censorship mechanisms are already in place (https://bit.ly/3hYsNm7), but applied only to Chinese domestic political contexts. The problem in reality is that Facebook and Twitter already apply political censorship in other countries outside the United States, creating not only a short circuit between politics and private companies, but also a diplomatic one. As it happened after the protests in Hong Kong where Twitter identified and suspended 200 thousand Chinese accounts (https://bit.ly/2XnMzy4) and YouTube suspended over 200 accounts (https://nyti.ms/2LbeqyQ ) accused of propaganda against the autonomy of Hong Kong. The phenomenon is however global and continuous if you think that Facebook closes 3 to 5 billion accounts per year (https://cnn.it/3om11Cx ). Probably most of them are right to be closed, but if even one account was closed without valid reasons it would have been taken away from Facebook or Twitter by a company.
Twitter has also censored tweets of heads of state such as Bolsonaro's when he chose to deal with the Coronavirus with herd immunity and not lockdown describing in his opinion its uselessness (https://bit.ly/2XkfuCX), but has not given the same treatment to the British PM when he had the same positions.
3- Third-party guarantee commissions set up by the social media themselves, for example Facebook's Oversight Board (https://oversightboard.com/), end up being only a mirror for the larks if the owner of the social media of the moment decides and declares censorship, as in this case Zuckerberg (https://bit.ly/3s5ZJOk), then rectified after a few hours by a company communication that reduced the account block to 24 hours (https://bit.ly/3hV9xG8).
Not to mention that preventive censorship seems more similar to the dynamics of preventive justice in the style of Minority Report than to the application of certain and equal rules for all.
In reality, the problem of false information has not yet been solved even in the traditional media. If I think about what has been the biggest fake news of the last thirty years I think about the weapons of mass destruction owned by Saddam Hussein, a false news that has had a devastating impact also from the point of view of the consequences that it has determined and that has been proposed and relaunched by the traditional media.
These days there is much debate about Bitcoin gaining value as a result of the general devaluation of state currencies. The reality is that blockchain technology, on which Bitcoins are based, is leading to the creation of new self-organized and distributed realities where there is no central control and within which people can act within rules established at the beginning and above all verifiable by all. The verifiability of decisions and the distribution of decisions is emerging as a model opposed to the centralized one based on arbitrary choices that are not shared. Maybe a social media based on these principles will be born soon and could solve the eternal dilemma between truth and freedom of speech.
Translated with www.DeepL.com/Translator (free version)
Indeed my 12-year-old daughter's commentary upon hearing the news that Trump was banned from Twitter: "I guess we'll then have to deal with him on TikTok."
Tough call. On one hand, the platforms are meant not to filter content according to Section 230. Therefore they can post broad guidelines, but that's about it. On the other hand, those provisions were designed twenty years ago, before the content platforms achieve near-ubiquitous and near-utility status. It's a fair argument to say they need to be managed differently, however a global explosion in creativity and communication has been facilitated by that approach. Can you minimise negative content without damaging positive content? Especially given that the judgement is profoundly qualitative.
Regarding the events at the capital, a liberal colleague said, "if they really feel that the election was stolen and democracy is being threatened, wasn't that the right thing for them to do?"
It's easy to point the finger at the platforms and say "it's time for you to be regulated" and that will probably happen. However, the reality is that Donald Trump was already impeached by the House and let go by the Senate on a strictly party-line vote. Because at the time, the Republican Party figured that the opportunity for the party was worth the cost of Trump. In fact, at each turn of the Trump administration toward greater degrees of autocracy and abuse, the Republican Party pretty consistently compromised core values for the opportunity to cement things like court justices.
If anything, the safe harbour provisions worked, in the sense that the social networks enabled quite high-fidelity communication between individuals and within/between communities. Everything that the Trump administration did has largely been done in plain sight, in part thanks to the the platforms. While we can kill the messenger, so to speak, this process has shown numerous cracks in the fundamentals of government. Which is in part what science and technology are meant to do. Challenge convention and enable greater understanding of how systems operate.
Whilst Zuckerberg and Dorsey finally had to act, the reason that they had to act in part was due to underlying failures of governance. Or perhaps the success of the system. If we're saying that some people are too stupid to think for themselves and they believe everything they see on the Internet, each side can make that argument. Liberals can say that Trump hijacked the minds of disaffected Americans. Conservatives can say the media at large is destroying the values of American society. Because fundamentally there are at least two Americas now.
Maybe one could argue the platforms helped facilitate those splits and atomisations, however Section 230 was designed to shield the platforms from liability and enable the growth of the network. And it has worked. If we don't like the result, it's worth considering that it's not that the platforms need regulation but perhaps they have created a more transparent view of the underlying society. And it may be that which we do not like and can no longer tolerate.
Regulating Facebook is not going to solve growing systemic inequality in the face of artificial intelligence. It may obscure it to some degree and potentially reduce its appearance, but that is not at all the same as directly addressing the drivers and impact of what led those people to the capital on that day.
Nicholas, is it an either/or? What if social media were not simply displaying the societal/cultural fault lines but also exacerbating them?
Sure, governance structures in the public realm have been revealed/reiterated as faulty. But manipulation through segmented communication on a vast scale has been made affordable and practical by social media in a manner that your description seems to overlook.
I appreciate @Sam Barcroft’s reply to your comment and seek here to extend it.
This needs to be highlighted:
"if they really feel that the election was stolen and democracy is being threatened, wasn't that the right thing for them to do?"
As a progressive liberal, I agree with this. The problem is that there was no evidence of the election being stolen and there shouldn't be this many people who believe it was. So the problematic bit of this saga is not the final action of the events at the Capitol, it was the spreading of the radicalizing misinformation in the first place.
I disagree. An effectively regulated social media landscape will be fantastic in shining a light on systemic inequality, by embracing guardrails that ensure honest reporting rather than deliberate misinformation.
Parler will not make a difference. It is hard to run a decent service, they won't get the talent. Without moderation, it'll turn truly horrible and for four years the Feds will just use it to track white collar terrorists. For normie on Parler, I can't see the product ever being attractive enough relative to Snapchat or Tiktok of even FB to get them to really stick around.
Yep, usually people run from platforms to get away from the hardcore wing nuts that come with scale. Parler is a unique example of the hardcore wing nuts running away to a new platform themselves.