This section of the article appears to drop the theory all together and use real data from YRDSB and PDSB. It exposes the 'culture of fear' where white administrators used nepotism, favouritism and unposted jobs to hire their friends while deliberately passing over skilled and qualified BIPOC candidates. This again establishes a power dynamic used to control the narrative within their systems.

From this, the takeaway is that Ontario has talked about inclusion and diversifying their teacher populations since 1993, but their main strategy to use neutral hiring doesn't appear to work. The province identified that they provided zero proof that this method closes the diversity gap. Instead, they have continued to ignore the real systemic racism that keeps BIPOC educators out while continuing to feed the privileges that favour white candidates.

This whole section is interesting because the official leadership framework in Ontario is providing guidance on how to bring changes through administration by determining their own conscious and unconscious biases, but nothing actually forces them to look at their own privilege or question why they keep promoting the way they are.

I understand the use of CRT and ACL to argue that we need to stop pretending the school system is totally 'colour-blind', but leaders in these higher power positions cannot simply follow passive rules; rather, they need to actively look within their biases to fix the unfair setup of the system. As situated further down, there needs to be a collaboration to shift the narrative.

This seems to be a key stat within this article, 90% of principals and VPs are white and middle class even though almost 30% of the population identifies as BIPOC. This speaks to the power that is held, and means the power to hire and promote is stuck with one specific group .

While Ontario is lauded as highly diverse, this section of the article specifically calls to question the education infrastructure and how it remains a system that privileges whiteness. The lack of representation has real-world consequences such as BIPOC students being suspended at way higher rates because the individuals in charge don't understand the communities in which they work.

The way that the war of position is explained here suggests that it is a slow struggle where common sense once existed. Zuckerman, in my opinion, is appearing to suggest that AI is not longer a simple technology used in settings such as education, but that AI is becoming an institution itself. It does bring about more questions than answers in terms of control. If AI continues to control what, how and why we read is it participating in the war of position or changing the field in which the struggle is taking place?

This word makes me curious. Zukerman describes the inference that is happening a means of encoding knowledge through patterns in the language/word prediction. But, if those patterns are drawn from existing texts, does or will AI generate new knowledge or is it just reproducing the ways of knowing it has been trained to do? I understand that the argument here is that LLMs do not simply know facts because they have been programmed with these language patterns. Seems to me that these patterns simply represent the statistical relationship given the ability to predict the next word. I suppose it is through these relationships that the encoded worldviews of LLMs are held.

My understanding is that hegemony isn't just about censorship or control, rather, it's about making one worldview appear to be natural. This suggests that LLMs are doing more than share their inherent bias, and moves in the direction embedded cultural assumptions and perspectives. If LLMs are trained primarily through these WEIRD cultural text outputs, will they not just normalize those perspectives while marginalizing others? Can or will AI ever move beyond the reproduction of dominant culture assumptions, or will it be limited by the perspectives represented within its training data? At this juncture, what responsibility do we carry as emerging scholars, researchers, and educators to recognize and challenge those assumptions?

If LLM's assume the 'average human' is WEIRD, what happens to learners that are already marginalized. In this case, I'm referring to those with disability, trauma histories, BIPOC populations, and all others outside dominant educational settings. It would seem to me that establishing 'average human' already positions these individuals as "invisible" based on AI technology and generated knowledge through AI.

It was fascinating for me to read through this blog post, and to recognize the lack of diversity that exists within the AI age. Through these vulnerabilities, the biases that exist contribute to the propagation of power imbalances through a set of "standard" values and beliefs in AI tools. If, in the age of AI, we were to embrace the many cultures, values and beliefs that exist, would we still be viewing AI from a premise of only texts contributed by "WEIRD" people?