What’s of interest? Open education principles: Resisting the metrics of AI black boxes
Tell me more!
By Tel Amiel, Glenda Cox and Colin de la Higuera
As the current trajectory of big tech, proprietary artificial intelligence (AI) threatens to widen educational divides rather than narrow them. The financial and environmental costs of commercial AI tools, and their concerns with profit and development primarily for affluent markets risk leaving behind precisely the communities that could benefit most from AI’s incorporation as an educational technology.
We suggest a strategy of “non-stupid optimism” in relation to educational technology (Facer & Selwyn, 2021). This approach acknowledges technology’s potential benefits while remaining critically aware of how techno-solutionist attitudes often serve economic growth and capital accumulation rather than genuine educational needs. We must avoid the trap of seeing technological fixes (AI included) as complete solutions to complex educational challenges. Still, we can be optimists in the sense that AI systems can embody openness: digital public goods that are designed to be transparent in their structures and functioning and embody collaborative governance structures.
The emphasis on closed models based on efficiency and economic gains has greatly intensified a historical narrative of educational achievement as measurable in metrics of performance (grades, graduates, papers) and return on investment (Luke, 2023). We begin by presenting two examples of these logics: traditional, closed educational resources and current AI applications in education. Second, we briefly discuss how aspects of openness – technology, practices, resources, and governance – can help us counter the current trajectory in AI in education.
This is one among many items I will regularly tag in Pinboard as oegconnect, and automatically post tagged as #OEGConnect to Mastodon. Do you know of something else we should share like this? Just reply below and we will check it out.
Re: Open Education Principles: Resisting the Metrics of AI Black Boxes
Dear Tel, Glenda, Colin, and OEG Connect Team,
Thank you for sharing this insightful article. I found the discussion on the risks of proprietary AI in education and the concept of “non-stupid optimism” particularly relevant.
As part of practical experiences in higher education, I would like to highlight the Open University of Sudan as an example of open higher education practices. Their approach demonstrates how openness in technology, resources, and governance can provide equitable educational opportunities, even under challenging circumstances.
I believe sharing such real-world examples can complement the theoretical perspectives presented in your article and help illustrate the impact of open practices in diverse contexts.
Thank you again for sharing this valuable resource.
Best regards,
Wisal Alim
Almayzab Digital Library
Sudan
Obrigado, Tel- the translated versions are now handy as I seem to be having connection problems to the UNESCO site, so I am re-reading a translated version.
What is the way forward around the black boxes? I read of ideas to call for open LLMs and efforts to make more use of local installed LLMS, yet so much attention and force seems exerted by the large commercial box holders.
yes, strangely the UNESCO site is down here periodically, too.
i think there are many strategies we point to, some easier than others. first is that the idea of “AI literacy” only goes so far and makes us reactive to extremely large forces (big tech et. al). So learning how to ‘use the tool’ (better, ethically, etc.) is not effective. we have to fight the idea that educators need to learn how to use ChatGPT “critically”.
we are really pushing (not new, off course) for more participatory design of AI policies, policy iterations (faster decision making in fast pace AI times), making sure we get educators, students engaged in thinking about how AI is integrated in their institutions. collective action is key - individuals fighting alone become - at best - martyrs for openness. but for this we need folks to see tech in education as a political fight…
and of course, moving away from big tech in our ed practice, using and promoting good FOSS and engaging in open practices/OER is an act of resistance against metrics of efficiency. small language models, local models, other types of AI (not extractive/error-prone chatbots), and effectively open tools (use only ethically sourced data, open code, weights etc) are some things cropping up.
