New study outlines four recommendations to end the harmful dominance of “big tech” AI platforms (e.g. Google, Facebook, Amazon, Tencent, Baidu, Alibaba)
Push for corporate breakups not appropriate and somewhat arbitrary (e.g. Google and TikTok cases in the U.S., Amazon case in the EU); instead, new regulations at the data level (e.g. EU’s Digital Services Act) promise to level the playing field and restore healthy competition
Decentralised platforms that use blockchain technology instead of AI represent promising alternative
The regulatory competition between U.S. and China to achieve data supremacy is feeding the lack of competition among digital platforms
The new study, Decentralized vs. Distributed Organization: Blockchain, Machine Learning, and the Future of the Digital Platform, is the latest research from UCL Associate Professor JP Vergne, exploring the growth of centralised digital platforms.
A handful of platforms dominate the global economy in the 21st century mainly by using machine learning as their core technology to transform enormous amounts of personal data into prediction services. These businesses cannot be competitively rivalled under the current regulatory frameworks because the datasets they have amassed cannot easily be shared or accessed by potential competitors, let alone be split as a result of a potential corporate breakup. Government appetite to regulate these companies is also tempered by the geopolitical benefits they deliver.
The study compares these centrally controlled platform companies with truly decentralised approaches to ascertain what lessons can be learned. Blockchain, for example, provides an interesting alternative technological blueprint by enabling transparent, decentralised platforms that are truly neutral.
The paper observes that the centralised platforms need a different understanding and regulatory approach to correctly address the elements that allow them to dominate. By following the four key recommendations in the research paper, there is potential to level the playing field without having to break these centralised companies up, which this study believes would also be counterproductive:
- Regulate data from the bottom up – this new approach would require regulators to work from the data up as opposed to reviewing at the corporate level first (e.g. the Google anti-trust investigation). A bottom up approach would mean regulating the properties of a company’s data so the capture and use of it is more tightly restricted.
- Regulators must clearly define and reward the benefits of decentralisation – Antitrust laws based on the 1890 Sherman Act are unfit for regulating machine learning platforms. Regulators need to a robust definition and measurement of decentralisation and should consider favourable tax regimes for decentralised platforms whose users have an enforceable right to vote service terms using digital tokens.
- Establish a platform utility regulation – The big tech corporations currently benefit from exemptions and certain protections in the same way as utility companies (e.g. telco, electricity, water distribution), however they don’t face the same level of regulation and restriction so have an unfair competitive advantage. Viewing dominant digital platforms as essential infrastructure would enable a “utility” designation with new obligations (e.g. common carriage, non-discrimination, interoperability, fair competition).Overall, slow progress on the regulatory front could be due to U.S.-China competition. Both countries seek to shield their homegrown platform monopolies from measures that could curb their growth and ultimately advantage foreign competitors.
- Education needs to keep pace with the technology – without an intermediate understanding of technologies such as blockchain and machine learning, managers’ abilities to successfully lead teams of developers and engineers will decrease with time. At the same time, digital platforms do not rely on managers as much as traditional corporations and there may be fewer job opportunities for them going forward. As a result, business schools adapt their curriculum or many of them will disappear.
JP Vergne said, “Today, centralisation is about data, not decision-making. Centralised corporations such as Facebook can distribute all the decision-making authority they want to platform users but as long as all the data is held on corporate servers, these firms are empowering nothing and no one. Without data, a decision-maker is just a puppet. This is the trick that centralised (yet distributed) platforms have been playing on their users for years and years. And this has been feeding a dystopian scenario built on top of an oligopoly. That said, trying to understand platform regulation without mentioning US-China geopolitics is like discussing nuclear arms regulation in the 1940s without mentioning WWII. And introducing new regulation that ignores the existence of decentralised platforms based on blockchain is like introducing new traffic regulations that ignore the existence of pedestrians and cyclists.”
Note. This research is at the crossroads of management, law, and computer science. It is the result of years of investigation by the paper’s author, who has held several leadership positions at research institutions focusing on the digital economy. The paper itself is based on a thorough review of recent scholarship on the business applications of blockchain and AI.