Contrarian by Necessity
- janalumi
- May 30, 2024
- 2 min read
Where AI gets it wrong and how it promotes popular ideas and group-think.
I recently read, or rather listened to, Michael O'Niell's article on Survivorship Bias in Curriculum Design. Just as in educational circles, AI is also being designed by a group of people who have survived the education system and social spaces that drive and influence AI's development: often at the detriment of those around them. The values of a lot of those in tech end up driving innovation that is isolating, immobility and centres a lot of European aristocratic cultural norms. There are very few apps that encourage movement beyond "exercise" apps. Convenience is not about making shopping and meals within walking distance, but rather promotes a logistics nightmare where drivers are under paid and people can just order what they want, when they want regardless of the cost of people and planet.
The narcissistic demands that are driving the consumer towards unsustainable choices and behaviours are at the centre of technological development and those who are coming up with the "ideas". When it comes to AI, we need a way to evaluate the data to identify trends, biases, paths of least resistance, majorities, group-think, popularity, etc. In turn, we need to find opposing or alternate options so we don't follow echo chambers into oblivion.
Yesterday, I watched World War Z:
If nine of us who get the same information arrived at the same conclusion, it’s the duty of the tenth man to disagree. No matter how improbable it may seem. The tenth man has to start thinking about the assumption that the other nine are wrong. - Mossad Chief Jurgen Warmbrunn, World War Z
We need to listen to our doubters and contrarians.
Our lives might well depend on them.
So what might that look like? How can we evaluate data? I haven't yet looked into what others are doing, but I've been mulling over this problem space for most of my life. My life long journey has included studies of technology, systems, information, knowledge discovery, AI, algorithms, data mining, programming, ethnography and anthropology. Additionally I've mainly listened to literature from people with female first names or non-binary or female identities who come from Indigenous, African, Black American, Asian, Indian and other minoritised groups and individuals. These groups are minoritised by our historical and present day white Christian European male centred hierarchy.
I strive to find the holes and gaps in my experiences and thinking, weaving their experiences into mine, becoming a whole with, rather than a self-centred isolated whole without. I did this also by acknowledging the privileges I've had at the expense of others, and the disadvantages I've had that others may well have profited from.
This is only one page in a life long journey, I don't aim to find answers, but to respond to a problem that is increasingly pressing.
Feature image generated by Stable Diffusion in Dreamstudio.ai
