Machines and Influence

Policymakers face a broader challenge of how to view AI capabilities today and where does society stand in terms of those capabilities. This paper surveys AI capabilities and tackles this very issue, exploring it in context of political security in digital societies. We introduce a Matrix of Machine Influence to frame and navigate the adversarial applications of AI, and further extend the ideas of Information Management to better understand contemporary AI systems deployment as part of a complex information system. Providing a comprehensive review of man-machine interactions in our networked society and political systems, we suggest that better regulation and management of information systems can more optimally offset the risks of AI and utilise the emerging capabilities which these systems have to offer to policymakers and political institutions across the world. Hopefully this long essay will actuate further debates and discussions over these ideas, and prove to be a useful contribution towards governing the future of AI.

https://arxiv.org/pdf/2111.13365.pdf

Sample:

…The recent development of Open AI’s Codex (Chen et al., 2021) which lets users accurately generate and complete computer programs, forms an incipient indication of a future where the most challenging reasoning tasks which require a logical generation of information, could be turned over to the machine. To any actor with enough computing power and access to advanced models like these, this would mean the capability to develop novel software tools which tilt the balance of information dominance in cyberspace. But in terms of influence, there could be far deeper implications of the rise of generative models, such as the process of training these models could also generate in these models an understanding (Karpathy et al., 2016) of the world. Another concerning and emerging reality is that the generative AI models based on Transformer architecture, like the GPT series, will be feeding into the rise of “no code platforms” and highly optimised computational generation of advertising content (Floridi & Chiriatti, 2020)…

About
Latest Posts

Ryan Watkins

Professor at George Washington University

I am a Professor with Human-Technology Collaboration and Educational Technology programs at George Washington University in Washington DC. I have written 12 books and more than 100 articles, and I co-host of the Parsing Science podcast where scientists tell the stories behind their research. I am also the developer of the WeShareScience.com online platform for sharing research videos, and SciencePods.com where researchers can create free podcasts about their science. My research interests include human interactions with intelligent machines, needs, needs assessments, and instructional design.