In the media, the advent of ChatGPT in November 2022 was described as mind-blowing, astonishing and a game changer. Before OpenAI opened its doors to the public, exposure to the prompt engineering of large language learning models was available only to a small subset of technology experts. The subsequent launch of similar tools and integrations has raised awareness of the future influence of AI on our lives and economies, and the ways in which it will revolutionize industries through process automation and job replacement.
Much like any newly introduced technology, its impact isn’t inherently positive or negative but it depends on how the technology is used. With this in mind, ImpactScope recently embarked on a journey to research and understand the potential impact of AI as it relates to our domain of knowledge: Blockchain and sustainability. As such – under the aegis of the Sustainability Working Group of the Crypto Valley Association – the report Sustainable AI: How can blockchain help? was published in July 2023.
The authors, Maja Kehic (CMO at DeepSquare), Carmen Cucul (Healthcare consultant) and our own Chief Sustainability Officer, Michele Soavi, wrote a report that explores the intricate relationship between AI and Blockchain technology within the transversal subject of sustainability. The report is based on the review of the topic conducted at the beginning of 2023, a survey proposed to the members of the Crypto Valley Association as well as an analysis of its members involved in different ways at the intersection of AI, Blockchain and Sustainability. Finally, it provides a framework – based on the acronym RESTART (Restrainability, Effectiveness, Security, Transparency, Accessibility, Representativity and Trust) – to narrow down to the most important ethical and sustainability principles and values when it comes to AI development and usage.
The conclusions of the report drawn on the need for a sustainable AI approach are multifaceted. First, there’s an emphasis on the crucial role of open-source software. This ensures that the benefits of AI are reaped by all of humanity, rather than being monopolized by a select group of corporations or individuals. Secondly, there is a significant focus on strategic planning regarding the objectives to be achieved through AI. The RESTART framework, in this context, emerges as a valuable tool to guide these requirements. By adhering to it, we can anticipate and sidestep potential negative repercussions of AI, such as the displacement of jobs, or mental health issues stemming from an overreliance on AI. Lastly, it is imperative to build the groundwork for an environmentally conscious AI. This means developing software that is not only efficient but also prioritizes reduced energy consumption and minimizes electronic waste production.
In this context, Blockchain can play a crucial role in sustainable AI. The two technologies exhibit a notable synergy. While Blockchain offers trust and transparency with its permanent and unalterable transaction records, AI maintains significant data processing capabilities, though its decisions can be opaque. This black box issue in AI can be mitigated by integrating it with Blockchain. By storing AI models and inputs on a Blockchain, it’s possible to trace AI decisions, providing clarity on its reasoning. Additionally, using Blockchain can help monitor and address inherent biases in AI models by offering a transparent review mechanism.
The report represents the first effort from ImpactScope concerning the possibility of complementing Blockchain – and more generally Web3 tools – with AI to achieve sustainability goals. We are currently exploring different approaches with such an objective both from a research and implementation perspective.
The full report can be downloaded here.