Analysis by AI of Reports of UN Debate on Artificial Intelligence

TRANSCEND MEMBERS, 14 Oct 2024

Anthony Judge | Laetus in Praesens - TRANSCEND Media Service

Eliciting a Coherent Meta-Pattern of Connectives of Strategic Relevance

Introduction

14 Oct 2024 – The United Nations has just approved an historic Pact for the Future and a Declaration for Future Generations — on the occasion of its Summit of the Future. That event was explicitly inspired by the possibility and necessity of “turbocharging” the Sustainable Development Goals with its 169 tasks, as discussed separately (Turbocharging SDGs by Activating Global Cycles in a 64-fold 3D Array, 2024). On that occasion it also adopted a Global Digital Compact to frame a regulatory response to the perceived threat to the future of humanity of AI. The drafts of the documents of the Summit rendered possible an early analysis by AI, as presented separately (AI analysis of connectives in the UN’s Pact for the Future and its Global Digital Compact, 2024).

The concern in what follows continues to be the extent to which debate on future global strategy, and the relevance of AI, is adequately articulated in the light of the logical connectives fundamental to the operation of AI and a computer-based knowledge civilization, as previously discussed (Comprehensible Mapping of the Variety of Fundamental Governance Functions, 2024). There is therefore a case for using AI to analyze the UN General Assembly debate which adopted those documents by acclamation — in the questionable absence of wider input, despite evolution of technical possibilities (Multi-option Technical Facilitation of Public Debate, 2019).

The debate had been summarized on a a daily basis in the DiploAI reports by the Digital Watch Observatory of the Geneva Internet Platform, written by their AI reporting tool (UNGA79: AI-powered insights and human-curated analysis). The summaries were described as “generated by humans” to provide “a comprehensive overview of how digital issues were tackled”. These enabled the procedure described below whereby those daily reports could be converted into PDF files and presented to ChatGPT 4 and Claude 3 for an analysis of their use of logical connectives and commentary on their possible implications for the future.

The use of “connectives” in debate — whether logical, emotional, “spiritual”, or action-oriented — is necessarily fundamental to the articulation of coherent strategy, its comprehensibility and its wider uptake. This consideration could be considered fundamental to any preoccupation with “turbocharging” the SDGs. The Digital Watch Observatory notably offers an interactive knowledge graph facility into which issues of the debate articulated by 215 speakers and 227 speeches have been incorporated (Visual mapping of UNGA 79 (Arguments and Statements), 2024). There it is noted that 254 arguments were presented on which there were 26 agreed points and 22 disagreed points.

The concern in what follows is how coherence and its comprehension is engendered by summit debates of strategic significance for the future — and the need for debate of a higher order (Second-order Dialogue and Higher Order Discourse for the Future, 2023). The initiative of the Digital Watch Observatory is therefore especially valuable in framing the advantages and disadvantages of one approach — evoking the question of how such possibilities might be usefully developed further. Some possibilities with that emphasis have been articulated previously in the light of the potential role of AI (Facilitating Global Dialogue with AI? 2024; Pathways in Governance between Logic, Emotion, Spirituality and Action, 2024; AI-enabled Mapping and Animation of Learning Pathways, 2024; Reframing Challenges of Governance of SDGs through Music, 2024).

As evident in the documents of the Summit of the Future and in the debate, the possibility that AI might be of considerable value in response to this global crisis is obscured by relatively ill-informed fear-mongering regarding the threat of AI to the future of human civilization. Little attempt is seemingly made to explore and demonstrate in detail how AI might be used to mitigate the challenges to the governance of a knowledge-based civilization — emotion to be recognized as “trumping” logic? The outcome of the UN-organized AI for Good Summit (2023) does not seem to have contributed to more balanced understanding. Unfortunately the fear-mongering has effectively been embodied in the Global Digital Compact, as approved at the Summit of the Future. Somewhat ironically even the possibility of use of AI — to summarize, analyze and render comprehensible the documents emanating from the Summit of the Future — has been avoided by the UN.

Provocatively it is appropriate to ask whether it is even possible to detect new strategic possibilities in a debate in which speakers tend to rehash old ideas — for which there has been limited uptake, as i evident by the challenges faced by the SDGs.

As in the previous experiments, the responses of ChatGPT 4o are distinctively presented below in grayed areas, in parallel with those of Claude 3.5. Given the length of the document to which the exchange gives rise, the form of presentation has itself been treated as an experiment — in anticipation of the future implication of AI into research documents. Web technology now enables the whole document to be held as a single “page” with only the “questions” to AI rendered immediately visible — a facility developed in this case with the assistance of both ChatGPT and Claude 3 (but not operational in PDF variants of the page, in contrast with the original). Reservations and commentary on the process of interaction with AI to that end have been discussed separately (Methodological comment on experimental use of AI, 2024). Whilst the presentation of responses of two AIs could be readily considered excessive, it offers a “stereoscopic” perspective highlighting the strengths and limitations of each.

As in previous uses of this approach, the question evoked is what can be “gleaned” from interaction with AIs, given their unprecedented access to information generated by a vast array of authors and authorities. The approach also offers the possibility that similar questions could be asked of any AI facility to which readers may have access, currently or in the future. The questions could well be refined, and the responses challenged, given the proactive responses of AI to such interaction.

READ THE FULL PAPER –> laetusinpraesens.org


Tags: ,

This article originally appeared on Transcend Media Service (TMS) on 14 Oct 2024.

Anticopyright: Editorials and articles originated on TMS may be freely reprinted, disseminated, translated and used as background material, provided an acknowledgement and link to the source, TMS: Analysis by AI of Reports of UN Debate on Artificial Intelligence, is included. Thank you.

If you enjoyed this article, please donate to TMS to join the growing list of TMS Supporters.

Share this article:

Creative Commons License
This work is licensed under a CC BY-NC 4.0 License.

There are no comments so far.

Join the discussion!

We welcome debate and dissent, but personal — ad hominem — attacks (on authors, other users or any individual), abuse and defamatory language will not be tolerated. Nor will we tolerate attempts to deliberately disrupt discussions. We aim to maintain an inviting space to focus on intelligent interactions and debates.

+ 71 = 77

Note: we try to save your comment in your browser when there are technical problems. Still, for long comments we recommend that you copy them somewhere else as a backup before you submit them.

This site uses Akismet to reduce spam. Learn how your comment data is processed.