Welcome to the TechStream newsletter! If this email was forwarded to you, sign-up here. We love to hear from readers. Get in touch: techstream@brookings.edu. A programming note: The TechStream newsletter will be on hiatus for the next two weeks and back in your inbox on Feb. 3.
Hacker Moxie Marlinspike on web3's empty promises
Over the past year, venture capital firms and investors have poured money into building out web3—a version of the internet that is based on the blockchain and that promises a return to the decentralized vision of the original web. If web2 was the era of big platforms built around user-generated content and data—think Facebook, Twitter, and YouTube—web3’s evangelists argue that decentralized protocols have the potential to return online to the ordinary user. Whether those evangelists can deliver on the promise of web3, however, remains tentative and deeply controversial. Early movers in the space may have become wildly wealthy, but the technology remains immature and has yet to demonstrate the viability of a fully decentralized internet stack.
The debate over the viability of blockchain applications took a turn this week when Moxie Marlinspike, the hacker best known for founding the encrypted messaging app Signal, published a blog post about his recent experience building a decentralized application, or dApp. Marlinspike details his experiments creating a dApp that minted a non-fungible token (NFT)—a digital piece of art whose ownership information is stored on the blockchain—and finds himself deeply skeptical of web3.
The core of Marlinspike’s argument is that web3’s promises of decentralization are almost entirely empty. In order to mint and trade NFTs, users have almost no choice but to rely on centralized platforms, like the NFT marketplace OpenSea. And because most dApps need to quickly query blockchain data to feel as responsive as web2 applications, developers to tend to build them using centralized sources of blockchain data—such as Infura or Alchemy—rather than the querying decentralized blockchain data directly, which can take far longer.
This results in a situation where an ostensibly distributed technology falls victim to the centralizing forces of the tech industry writ large. Users and developers of web3 applications don’t want to deal with the hassle of managing the infrastructure necessary to interact with the blockchain, just as most users and developers of web2 prefer not to host their own servers. Instead, they rely on companies like Infura or Alchemy that have built application programming interfaces to interact with the blockchain. Consequently, the winners of the early web3 economy are the platforms that are making blockchain technologies accessible to users. These include exchanges like Coinbase and Binance, NFT marketplaces like OpenSea, and the smart contracts platform Etherscan. It is of course deeply ironic that centralized applications and interfaces are the early winners of the web3 economy, which its evangelists are pitching as a way to escape the power of the current crop of dominant platforms.
Although the Graph Protocol and similar efforts to decentralize web3 infrastructure are under active development, his post stands out as a clarion call for greater discussion of the key ways in which the decentralized web is failing to deliver on its guiding vision. Web3 evangelists will need to heed that call if they ever aim to turn their aspirations into reality.
Why is Marlinspike’s post worthy of such consideration? First, because the reception to his essay has been nothing short of astounding. Believers in web3 technologies tend to respond to criticism with hand-waving and platitudes: The technology is at an early stage, they say, and the critic is missing the point of what might be a game-changing technology. Marlinspike has generated a very different response. Vitalik Buterin, one of the developers behind Ethereum, took to Reddit to observe that Marlinspike’s points are basically correct. Brian Armstrong, the CEO of Coinbase, said much the same on Twitter. Second, Marlinspike is one of the most respected and important cryptographers and developers of his generation. He has done perhaps more than anyone in the world to make end-to-end encryption available to ordinary users.
Marlinspike’s intervention this week isn’t the first time he’s offered his thoughts on decentralizing technologies, and it’s worth returning to his 2019 talk at the Chaos Communication Congress (CCC) to get a sense of where Marlinspike stands in this debate. Marlinspike’s development work has been oriented toward making privacy-preserving technologies easily accessible to ordinary users, and he sees decentralization as standing in the way of this project. In his CCC talk, Marlinspike argues that decentralization doesn’t have much to offer in terms or privacy protection, security, availability, or control—and arguably undermines it. In the 2019 talk, Marlinspike’s central point was something like this: If software is to be improved for these qualities, the solution doesn’t lie in technically arcane decentralized technologies, but in making software development easier, so that the platforms that currently make development possible don’t hold so much power.
China’s appropriation of foreign technologies and products is only one part of the way in which its tech industry has developed innovative products in recent years. As the Chinese government pushes its tech industry to move beyond copycat methods, it has bolstered its domestic innovation capabilities and adapted them to fit within the Chinese model. Chinese technological innovation is a system of “re-innovation”, or zaichuangxin; it does not mirror other global paradigms. Discussions surrounding China as a strategic competitor have been shaped by the notion that only democracy can promote innovation and creativity, that a market economy provides the only route to success, and that innovation derives from the private sector. Every day, China is disproving this line of thinking. From a U.S. policy perspective, a more holistic understanding of China’s innovation strategy—beyond a myopic focus on technology transfer—will be key to effectively competing with China in emerging and foundational technologies, Emily Weinstein writes.
Industrial revolutions of any kind have real and severe implications for economic development, national security, social cohesion, and human rights, and the one we are experiencing now poses even greater risks along these lines, given the speed and scale at which digital applications and systems can be deployed across traditional borders. Structural shifts in the sources and wielders of economic, social, and political power—and the urgent threats that accompany such changes—suggests a need for new forms of regulation and governance that ensure common social values survive and thrive. Most importantly, values such as fairness, inclusiveness, and accountability need to be consciously and carefully built into both our governance systems and the technologies themselves to ensure that their direct risks to users and negative externalities to others are well managed, Nicholas Davis, Landry Signé, and Mark Esposito write.
Deepfakes. China ramped up its effort to police deepfakes on online platforms. New rules issued by Chinese internet regulators set up limits on the content shown by recommendation algorithms that powered many of China's biggest platforms, like the news aggregator Toutiao, microblogging site Weibo and Douyin—China's version of TikTok. The new rules ban the platforms’ recommender systems from showing "synthetic” content, including deepfakes and "fake news.”
Antitrust. A federal judge is allowing the Federal Trade Commission’s antitrust lawsuit against Meta (née Facebook) to proceed after dismissing it last year. The FTC lawsuit, the agency’s signature antitrust lawsuit against Meta, alleges the company violated antitrust law by acquiring Instagram and WhatsApp. Judge James Boasberg is now allowing the complaint to go forward after the government revised its suit to include more evidence as to why Facebook has a monopoly in social media. His ruling is a victory for the government in what is expected to be a protracted legal battle. Boasberg cautioned that the FTC “may well face a tall task down the road in proving its allegations.”
NFTs. As the digital artworks known as non-fungible tokens move further into the mainstream, Politico reports that the Canadian firm Dapper Labs, a major NFT player, will become the first company registered to lobby Congress on the issue. Lawmakers have shown increased interest in recent months about how to best regulate blockchain technologies, and Dapper Labs decision to lobby on matters related to NFTs is likely a preview of an intense effort by the nascent industry to shape government regulation.
Social media manipulation. India’s ruling party, the Bharatiya Janata Party, appears to use an app dubbed “Tek Fog” to manipulate trending topics on Facebook and Twitter and amplify the party’s propaganda, according to an investigation in the Indian outlet The Wire. The investigation finds that a significant amount of social media replies to prominent local journalists between January and April 2021 can be traced back to the app, and the contents of many of these replies are classified as hate speech.
South Korea. Apple submitted plans to South Korean regulators detailing how it will allow third-party payments on its App Store in compliance with a ruling from South Korean anti-trust regulators. Allowing third-party payment systems on app stores has become a central conflict in attempts by anti-trust regulators to rein in the power of firms like Apple and Google. South Korea’s efforts to force alternative payment systems are now the most advanced test-case for how to do so.
Metaverse. As companies ranging from game publishers to social media giants rush to build virtual worlds, experts are beginning to examine how they will affect user well-being. A recent feature from the Wall Street Journal shows a mixed view from experts, with some researchers underlining the potential for such worlds to foster greater connection and others raising concerns about the potential for harm. Those wary of the metaverse cite, among other things, the potential loss of a sense of human identity as well as the difficulty of ensuring that virtual worlds are safe for young people, who are poised to be early adopters of this technology.
Reports we’re reading
Misinformation. A new paper in Nature Reviews Psychology examines why some people believe in misinformation and why they are resistant to having those views corrected.
Machine learning. A recently published paper examines the use of benchmark datasets in machine learning and raises concerns about how the field has standardized on a few central datasets created by a small number of institutions.
A final point
“I wouldn’t bet my house on it.”
— A staffer in the U.S. Congress assesses the likelihood of antitrust reform passing before the November election.
Facebook and Google provide financial support to the Brookings Institution, a nonprofit organization devoted to rigorous, independent, in-depth public policy research.
The Brookings Institution, 1775 Massachusetts Ave NW, Washington, DC 20036