This week, S&P Global Ratings released a report on the crossover between AI and crypto titled “The Question is WHEN, not IF.” With the launch of ChatGPT almost two years ago, the topic of AI is now much bigger than blockchain and cryptography. But the intersection of the two is not new.
Before diving into the report, it is worth exploring some examples of collaboration between the two technologies. Artificial intelligence requires machine learning, which in turn needs data. So, an application uses cryptography to pay people or devices that generate data.
Since 2019, German manufacturer Bosch has been working with Fetch.ai to use IoT devices and Fetch’s idea of Autonomous Economic Agents (AEA). These are sensor-driven (IoT) devices that are both self-learning and capable of automatically carrying out transactions using cryptography. One can easily imagine a decentralized network of weather stations receiving compensation for data. These types of applications are called decentralized physical infrastructure networks (DePIN).
Fetch recently struck a deal with GameSwift Launcher that lets users use their computer’s unused graphics card capacity for AI applications like text-to-speech or music creation. In other words, they run extended language models (LLMs) locally, creating a decentralized network that addresses the capacity gap. Of course, compensated using crypto.
S&P Report on the Intersection of AI and Crypto
Let’s return to the S&P Global Ratings report, which highlights the potential of AI, but also the risks it creates. These include data traceability, cyber threats and data center energy consumption.
The report considers three potential future scenarios. At one extreme, there is modest progress in the areas of AI and cryptography. At the other end, the vision includes a decentralized internet powered by cryptography and AI. In this scenario, the immutability and traceability of blockchain contributes to transparency and auditability. It will be possible to understand why the AI had this crazy idea. Blockchain can also help distinguish between humans and robots.
Another scenario is the rapid expansion of AI, as is currently the case, with risks of centralization. In other words, the governance of these large publishers is relatively opaque. There are concerns about bias, privacy and censorship.
The report then takes these three scenarios and applies them to different use cases, including cybersecurity, financial markets and supply chain.
“Synergies between technologies are expected to support their growth, mitigate the risks of centralization and give rise to high-impact applications ranging from supply chain management to smart cities,” said Andrew O’Neill, managing director of digital assets at S&P Global Ratings. “The speed at which these applications will emerge and the pace of their adoption remains uncertain. Yet we believe the question is not if adoption will happen, but when it will happen.
