This article is part of a series in which OECD experts and thought leaders—from around the world and all parts of society—discuss and develop solutions now and for the future. Aiming to foster the fruitful exchange of expertise and perspectives across fields to help us rise to this critical challenge, opinions expressed do not necessarily represent the views of the OECD.
Join the Forum Network for free using your email or social media accounts to share your own stories, ideas and expertise in the comments.
Paying and receiving attention drives the attention economy’s business model. Users, platforms and various online-service providers circulate information that other users spend their attention on. Spending attention generates engagement and traffic, which in turn creates the user data that social platforms and tech-companies collect, curate, analyse and subsequently package and sell as finely masked and segmented ad-packages to interested parties. It is from this simple target advertisement model that the tech industry has spun its fortunes. In essence, it rests on capitalising on our attention: our online behaviour is translated into behavioural data, which in aggregated form is sold to companies for the purpose of predicting the future behaviour of different segments and for mapping the potential to influence them.
BigTech business model
This monetary setup based on attention allocation is also known as the business model of surveillance capitalism (Shoshana Zuboff 2019, see figure 1.0); it is one of the most potent and pervasive ever invented by mankind. While industrial capitalism exploits reserves of natural resources such as oil, coal and gas and processes raw materials, surveillance capitalism relies on the cognitive resources of humans for its exploitation. As Zuboff notes, Google was the first platform to implement the model of surveillance capitalism, even though they had declared in the 1990s that they would never be ad-driven. But around 2001, after the dot-com bubble burst and the economic downturn that ensued, Google began generating ad-revenue via its enormous data bank repositories (and subsequently knowledge) of user behaviour, linking them to algorithm-based and automated systems that effectively targeted advertisements at users. During their 2004 stock listing, Google's financial statement revealed that their earnings had risen by 3,590% from 2001 to 2004 (Zuboff 2019a).
The data users voluntarily, involuntarily, knowingly and unknowingly provide to platforms free of charge, in exchange for attention, is what platforms profit from by renting or selling these data to third party vendors. As users we tend to think of ourselves as the consumer, but in reality we are the product itself. User profiling and predicting behaviour is not only interesting for a company that has a product to sell, but also for institutions, organisations or nation states that seek to influence the behaviour of their citizens e.g. stopping the spread of COVID-19 in populations or consolidating autocratic ambitions.
Data sales are massive and concentrated in very few hands. Meta/Facebook and Google in particular have sat on more than half of all digital advertisement budgets over the past years, and are now locked in a bidding war to sweep up the remaining competition.
Figure 1.0. The attention economic business model of tech platforms.
Besides Meta/Facebook and Google/YouTube, Instagram, WhatApp and Amazon, all the major players in tech run on this business model—including TikTok and Twitter, the latter of which was recently sold for a mind-blowing USD 44 billion .
Tesla and SpaceX founder Elon Musk has just acquired a tech platform for himself, and not only an ad-based one but also what is considered the new public square and speaker’s corner: Twitter. In what The Wall Street Journal has described as one of the biggest deals in tech history, Twitter accepted Musk’s USD 44 billion bid to gain control over a platform with more than 200 million users world-wide. His take-over came with substantial promises of change rendered in Musk’s tweet immediately after the deal was struck (figure 2.0).
🚀💫♥️ Yesss!!! ♥️💫🚀 pic.twitter.com/0T9HzUHuh6— Elon Musk (@elonmusk) April 25, 2022
Figure 2.0. The tweet from Elon Musk announcing the take-over of Twitter on 25 April, 2022.
As a “free speech absolutist”, Musk’s vision is of a largely unregulated information market liquid with opinion, which will eventually find some suitable state of equilibrium given supply and demand. By way of example, Musk vehemently opposed former president Donald Trump’s ban from Twitter under the current community standards of Twitter, and many are speculating as to which accounts may be reinstated under Musk’s new broom. The two game-changing promises launched in the tweet include:
- "Free speech is the bedrock of a functioning democracy, and Twitter is the digital town square where matters vital to the future of humanity are debated”.
- “Making the algorithms open source to increase trust, defeating the spam bots and authenticating all humans”.
To realise the central promises of democratisation and trust, in a series of interviews and tweets Musk has as of late specified a number of concrete initiatives to be implemented on the platform under his reign and rule:
- In an interview during a TED conference on 14 April, 2022, Musk the “free speech absolutist” calls Twitter the de facto town square, where “it’s very important for there to be an inclusive arena for free speech”. On such a square, Twitter is to soften its stance on content moderation, exercise extreme caution while handing out permanent bans, instead enfocing timeouts and upholding full transparency if shadow-banning is the policy pursued on specific users.
- An edit button on tweets should be installed, a long-time request among users.
- Twitter should allow for longer tweets, and come with encrypted-messaging functionality.
- Twitter should be transformed into a private company and taken off the public stock exchange, since “Twitter has extraordinary potential—I will unlock it”, as Musk noted in a regulatory securities filing.
- Twitter’s algorithmic architecture should be open source and made available on GitHub.
- Twitter should wage war against bot armies and install authentication procedures for users.
- In terms of the business model, the global public digital square should rely less on advertising and more on subscriptions.
The political landscape in the United States has responded to Musk’s acquisition of Twitter as one would expect according to party lines, emphasising either the freedom of expression pace the Republicans…
“This is a great day to be conservative on Twitter”. – Sen. Marsha Blackburn (R., Tenn.);
“Congratulations, @elonmusk! Looking forward to a free-speech oriented twitter”. – Rep. Yvette Herrell (R., N.M.)
… or the Democrats’ concerns related to the public space in private hands:
“This deal is dangerous for our democracy. Billionaires like Elon Musk play by a different set of rules than everyone else, accumulating power for their own gain. We need a wealth tax and strong rules to hold Big Tech accountable”. – Sen. Elizabeth Warren (D., Mass.);
“I think in general, it's really bad for democracy to have a handful of rich people owning everything that matters to us”. – Sen. Chris Murphy (D., Conn.)
Both of these tweets from Democrats were topped by Rep. Alexandria Ocasio-Cortez (D., N.Y.) who, following a scathing tweet, got into a heated Twitter exchange with Musk, the latter of whom used reflexive show-it-don’t-tell-tactics (“Stop hitting on me, I’m really shy" ). Ocasio-Cortez later deleted her response, which in the first place was seemingly intended for Musk, after suggesting the criticism was for Meta's Mark Zuckerberg (figure 3.0).
Figure 3.0. The Twitter-exchange between Alexandria Ocasio-Cortez and Elon Musk on April 29, 2022.
Musk’s take over of Twitter comes only days after the European Parliament and European Commission converged on the Digital Services Act (DSA), which itself only post-dates the Digital Markets Act (DMA) by a measly month, both of which in return are of interest—and concern—to Musk.
While the DMA is to ensure level playing fields among businesses in the information market ruled by an attention economy, the DSA is to ensure a safe, sustainable and accountable online environment for users. European Commission President Ursula von der Leyen recapped the basics in a press release on 23 April, 2022:
"The DSA will upgrade the ground rules for all online services in the EU. It will ensure that the online environment remains a safe space, safeguarding freedom of expression and opportunities for digital businesses. It gives practical effect to the principle that what is illegal offline, should be illegal online".
Essentially, the DSA comes with the following obligations and provisions:
- All targeted advertising related to an individual’s sexual orientation, religion or ethnicity is banned. In addition, minors are not allowed to be subjected to targeted advertising.
- What is referred to as “dark patterns” are prohibited by the new legislation e.g. discombobulating or deceptive user interfaces nudging users to making certain choices.
- BigTech platforms are to make the nuts and bolts of their recommender-based algorithmic architecture transparent to platform users.
- Instead of profiling, users should be offered recommender-based systems of a different nature, the results of which should be generic ads or chronological feeds.
- Platforms of substantial size have to provide researchers and the scientific community at large with key data to “provide more insight into how online risks evolve”.
- Any online marketplace must keep basic records of their vendors such that, if needed, traders of illegal goods and/services may be tracked down.
- Platforms are to be held accountable for their decisions to remove illegal content; consequently, they need to explain their decisions to intervene and ought to make available to users the ability to appeal such actions. The definition of illegal content lays with the respective countries of the EU.
- Finally, a constraint—in no small way inspired by the recent invasion of Ukraine—dictates that large-size platforms must come up with measures to battle misinformation, fake news and “bs” in times of crisis.
Musk’s Twitter visions and the DSA seem to somewhat coincide when it comes to (1) limiting the ad-driven model and targeted advertising; (2) making the recommender-based algorithmic architecture transparent and possibly available to users; (3) removing illegal content where legal/illegal refers to the rule of law of the country in question. That’s about half of the provisions of the DSA they agree on up front, taking into account that Musk has not explicitly voiced his opinion on many of the remaining obligations and provisions.
Be that as it may, items 7 and 8 on the DSA agenda may be subject to some dispute between Musk and the EU. Although agreeing that illegal content should be removed, the leniency with respect to the ruling principles of the free marketplace of ideas, and best practice principles when it comes to borderline content, may in the end differ quite substantially. According to Musk, the laissez-faire liquid opinion market is the way to go, ideally resulting in a natural equilibrium of “less hate and more love” while users are giving their consent to certain points of view based on supply and demand chains for them. But this assumes level playing fields in the marketplace of ideas, which is not immediately the case as demonstrated elsewhere. We should add to this that a public space in private hands does not guarantee equal and fair access for all users, and tech platforms do not immediately answer back to public like values truth, democracy or even the personal preferences of users, but rather to their respective community standards, advertisers and shareholders.
Even Musk inadvertently has had to admit and experience as much—even prior to his acquisition of Twitter. Musk has routinely blocked and tried to have accounts banned on Twitter over the years...especially people criticizing Musk. He apparently remains committed to the belief that even his most virulent critics should be allowed under his new rule (figure 4.0).
Thanks, @elonmusk. As your best, blocked critic, I will definitely remain on Twitter.— Alex Epstein (@AlexEpstein) April 25, 2022
And I am confident I am less likely to get kicked off with you owning Twitter.
Although it’s pretty weak to block someone for rigorously refuting your claims. Within your rights, but weak. pic.twitter.com/gaVkOWEPga
Figure 4.0. Musk vowing that even users critical of him are to remain on Twitter after his take over.
Of course, Musk is free to block users as he sees fit on his personal account—including the critical voices, no matter how “weak” his reason may be—but Musk’s argument for having a particular account closed is to be sure a different story. Since June 2020, a 19-year-old freshman from the University of Central Florida in Orlando, Jack Sweeney, has been maintaining a Twitter account, @ElonJet, keeping track of a Gulfstream G650ER identified as Musk’s private jet (Figure 5.0). On 30 November, 2021, Sweeney received a message from Musk himself asking him to deactivate the account citing safety and security risks. Sweeney has continuously denied to do so, and has of 23 January, 2022 been blocked from sending messages to Elon Musk on Twitter.
Figure 5.0. A screenshot from @ElonJet, the Twitter account keeping track of Elon Musk’s private jet.
To reiterate, the reason for having @ElonJet shut down was conveyed to Jack Sweeney in a message from Musk: “Can you take this down? It’s a security risk”. If nothing else, this is an admission that even a “free speech absolutist” is committed to thinking about particular community standards when, according to Musk, it is about the safety and security…of himself (and perhaps others with private jets). Safety and security is part and parcel of a plethora of issues related to questions of allowed content, illegal content and borderline content, the latter of which is up for multiple interpretations and framings as well as other underdetermined and exceedingly difficult features of content moderation. The global public square is not only in private hands, it is literally in Musk’s two hands—which is all for the better according to Musk, as he will see to it that Twitter’s “extraordinary potential” will be unlocked with his personal checks-and-balances in place. Even so, Musk is under the terms of the USD 44 billion dollar buy-out, and is himself banned from criticising Twitter. So much for his first promise of free speech and the bedrock of democracy.
While we're at it: Musk’s second promise in the 25 April tweet to wage war against bot armies, spam bots and other forms of automated vandalism may turn into a predicament. Research from the Observatory on Social Media of Indiana University, Bloomington (OSOME) show that tweets expressing positive sentiments were subject to bot amplification immediately after Musk’s buy-out Twitter announcement (figure 6.0).
Quick analysis on bot activities of tweets mentioning 'Elon Musk' Apr 23-26. Two notable patterns— Kevin Yang (@yang3kc) April 28, 2022
- Twitter&Musk deal news out -> higher volume -> lower bot activities
- Tweets expressing overall positive sentiment began to show higher bot amplification levels after Mon 4pm ET
Figure 6.0. Bot amplification of positive sentiment following Musk’s announcement to buy Twitter.
Earlier this week, Shoshana Zuboff referred to the DSA as “the first comprehensive declaration of a digital future founded on the legitimate authority of democratic rights and the rule of law, and a signal that the principles of a self-governing demos might survive the digital century”, but she also mentions that “a great deal of work remains to be done. Much of what occurs in our information spaces today is profoundly illegitimate, but because it is unprecedented it is not yet illegal".
Please read our book Ministry of Truth: BigTech’s Influence on Facts, Fictions and Feelings, which, by the way, is now ready for order…on Amazon…complete with ads worldwide. We’re all neck deep in it!
Please sign in or register for FREE
If you are a registered user on The OECD Forum Network, please sign in