Stephen Downes

This is in response to a contribution to an OAS meeting distributed in my office this morning. It is of course my set of opinions only, and not reflective of any official policy or practice, though I would add that most of these have been undertaken to one degree or another by various levels of Canadian government organizations.

I want to say something like 'this paper describes a core tenet in connectivism' although of course we never conceived of it in anything like the richness and detail collective intelligence across 'scales and substrates' described here. This, in particular, is crucial: "collective intelligence is not only the province of groups of animals, and... an important symmetry exists between the behavioral science of swarms and the competencies of cells and other biological systems at different scales." The way networks work is tied up in the way evolution works, and these are tied up in how we describe learning and cognition generally. Or - how we should describe learning and cognition (as most people still labour under the mythology of folk-psychological information processing types of pictures such as 'executive function' and 'cognitive load' theories).

Drawing on an emerging new picture of animal consciousness,  a group of researchers have signed this declaration recognizing animal consciousness. "Subjective experience requires more than the mere ability to detect stimuli. However, it does not require sophisticated capacities such as human-like language or reason. Phenomenal consciousness is raw feeling—immediate felt experience, be it sensory or emotional—and this is something that may well be shared between humans and many other animals." My similar sentiment is expressed here. It may be thought that an ethics of animal rights and welfare follows immediately, but given the way humans treat each other, we need not fear the recognition of animal consciousness forces any new behaviours on our part (though it probably should).

I mean, it should be obvious why we don't want advertising in AI chatbots, right? They're supposed to be trusted advisors. Imagine you went to your lawyer for advice on selling a house and they said, "I'd be glad to help you, but first, let me take this opportunity to recommend a McDonald's hamburger." Yeah, no. This applies doubly as the technology, still under development, already has trust issues. " Advertising in these generative AI chatbot experience won't be a "sustainable model" long term, according to Jaffe, unless CPMs 'go way up.' Building a subscription model for Ingenio's chatbots also means the publisher will have more control over revenue, the user relationship and distribution, he added." 

A recent meeting of UNESCO on digital education futures is notable for the resources distributed: "its AI Readiness Assessment tool to translate the Recommendations on the Ethics of AI into actions, as well as its Guidelines for the Governance of Digital Platforms to enable freedom of expression and inclusion while promoting a healthy information ecosystem in a digital era." Additionally, UNESCO's Guidance for the Use of Generative AI in Education and Research contains "a roadmap for regulating AI in education and strategies to address its profound risks and impact on teaching and learning," and a recent research report "which revealed gender biases and prejudices found in Large Language Models."

I've been doing some work recently looking at task modeling. The time of linear and even circular models (like OODA) has past. Today we're looking at learning and working in complex environments that requires new tools. I stated by thinking of circular network diagrams, looked at chord diagrams, but this seems to be getting closer to the reality. "The CL aims to capture the entanglement and dynamics between the Covid-19 crisis and the Swedish food system through data-based visualization. The methodology includes a creation of a constantly evolving database with systemic trends captured through qualitative research methods i.a. interviews with food system actors, or literature review." The tool they use is called Kumu.