This article is the third in the A.I. and Blockchain mini-series by George Polzer. In the first two articles, he separately explained A.I., then Blockchain and touched on how they interconnect. In this article, he details a key driver, decentralization, that interconnects those two transformative technologies.
As explained in article #2 “What is Blockchain”, it was Satoshi Nakamoto in 2009 for the first time in history to offer a mathematical solution that automates reaching consensus among a peer-to-peer computer network using economic incentives (Bitcoin mining) to come to a single truth of the ledger state (recording value exchange) without a centralized authority.
Disintermediation vs. Decentralization?
It is important to understand that the Blockchain as implemented by Satoshi with the introduction of Bitcoin was not premised on a decentralized computer network but a trust-minimized system without a centralized authority, i.e., disintermediating institutionalized trust/inequities. Decentralization does not equal disintermediation. This distinction is critical to not only understand the Blockchain A.I. interconnection but also because many Blockchain projects had promoted themselves to develop decentralized apps. However, the real value they planned to offer was disintermediation — neither promises have been realized. Consider this short article titled “Deconstructing the Myth of Decentralization” . It breaks down four essential components of decentralization: architectural, data, procedural, and political.
What is Decentralization?
There have always been centralized vs. decentralized computing cycles: the end of the mainframe era, into user desktops, back into centralized computing with X terminals and other “thin clients”, back out onto the desktops again with the rise of extremely powerful, extremely low cost commodity hardware and back into giant centralized clusters now called “Clouds”. These cycles were based on economic factors of the changing cost of computing power and hardware. In contrast, the impetus for the introduction of Blockchain technology as Bitcoin after the 2009 financial crisis was the demand for disintermediation (automating trust), not the economics of centralization/decentralization cycles.
A.I. has become the “victim” of the centralization of the Internet. The Internet was originally intended by the US ARPANET military project to withstand unforeseen disasters like war, designed without a central server but peer-to-peer interconnectivity. After the ARPANET military project was decommissioned in early 1990, centralization settled in when Internet Service Providers (ISP) entered the picture, i.e., Web 1.0. Users no longer had to dial-up different servers, each hosting a particular service which made the Internet chaotic and highly decentralized. Fast forward to the emergence of cloud computing, a virtual collection of servers controlled by giants such as Facebook, Google, Amazon, hence the centralized Internet and its centralized data, a.k.a. Web 2.0.
A.I. Creation is a Data Problem
Creating A.I. or intelligent systems to automate prediction is fundamentally a data problem. Extremely large data sets or Big Data is needed to train machine learning algorithms. Those large datasets are currently centralized and controlled by a few BigTech companies. This “rich get richer” cycle of centralized data, and machine intelligence has become one of the most pressing issues in the evolution, potential abuse, and promise of A.I. An excellent book to read on this topic is Amy Webb’s recently published “The Big Nine: How the Tech Titans and Their Thinking Machines Could Warp Humanity” — a call-to-arms about the broken nature of artificial intelligence, and the powerful corporations that are turning the human-machine relationship on its head.
The rapidly maturing Blockchain technology and its related decentralized vision are the catalysts to enable the first phase of increased awareness and beginning to offer partial solutions that have launched the decentralization of A.I. journey. If A.I., as Nvidia CEO Jensen Huang recently stated, is the “single most powerful force of our time,” its value creation model must be democratized. It must not be trusted to a single authority. Disintermediation is essential. This need for disintermediation makes the intersection of Blockchain and A.I. a perfect use case: Blockchain as the operating system for A.I.
With the potentially existential threat of centralized A.I., we can learn from similar centralization inequities of the industrial revolution which created a 150-year gap between countries that industrialized and those that didn’t. Moreover, we can learn from the benefits of Open Source software development leveraging thousands of software engineers to produce superior code rather than a few that were driven by corporate interests.
In the next and fourth A.I. and Blockchain mini-series article, I explain the current state of Blockchain projects attempting to obtain the benefits of an inclusive and open method of developing A.I.