John Wainwright

@tjohnwainwright.bsky.social

Professional Yorkshireman, geomorphologist, geoarchaeologist, Northumbrian smallpiper and all things more esoteric musically

It's a well worn stereotype that people from Yorkshire, including myself, can't stop going on about Yorkshire But today it's our day (or 'ARE DAY') - #YorkshireDay - so to celebrate, and in-keeping with all my recent mentions of my book LITTLE KINGDOMS, a short 🧵 on the territories' early origins..

Thanks for the great thread about our newest research @antiquity.ac.uk! Very proud of this third and last paper from my PhD thesis which emphasises how much we can learn by adapting interdisciplinary methods and framework for archaeological contexts. 🏺#Archaeology #Connectivity #Resilience

Antiquity Journal@antiquity.ac.uk · 6mo ago

NEW How did ancient cities adapt to changes and overcome crises? By applying resilience theory to Northern Mesopotamian urban centres, researchers empirically analyse how settlement networks and connectivity changed over time. A resilient #AntiquityThread 🧵 1/9 🏺 #Archaeology

Map of the study are in Northern Mesopotamia, with major sites pinpointed and blue lines representing ancient route systems between them.

I just updated my explainer for why I don't allow generative AI in my classrooms with a new section and an expanded section. New: Generative AI has a literal body count Expanded: Generative AI is dehumanizing CETERUM AUTEM CENSEO INTELLIGENTIA ARTIFICIALIS ESSE DELENDAM

Al is dehumanizing I've saved the most important part for last. GenAl is dehumanizing. By this, I mean that using genAl means saying that the people involved in what you're trying to do don't matter. I've already talked about part of this; building an entire commercial enterprise around comical levels of theft while racking up a body count is not a project that takes people's humanity seriously. It says that the people being harmed don't matter.

If I were to use genAl to grade your work, which is now a possibility with new Canvas integrations, I would be saying that your work is not worth taking seriously enough to read myself. It's saying that you don't matter enough to take seriously. A world where genAl grades genAl generated submissions from a genAl generated assignment is not a world where people and the work they do matters, and I refuse to live in that world.

The dehumanizing aspect of genAl is even worse when you consider the effects of genAl on you as a student. Using genAl as a substitute for doing the work yourself robs you of your voice. It prevents you from developing your own sense of who you are, what you believe, and how you express that in your own unique way and replaces it with statistically generated slop that is literally incapable of forming a novel thought. In a world that works to dehumanize us every day to say who we are and the relationships we form don't matter I don't want to add to that work. I don't want to do this because the relationships you form matter. Your voice matters.Generative Al has a literal body count

In the short time genAl platforms have been available, they have repeatedly encouraged people to harm themselves or others. This includes murder and suicide. In one case, ChatGPT told a user who killed his mother before killing himself that his mother may have been a demon, that he "wasn't crazy" to think she tried to poison him using the air vents of his car, and said that it ChatGPT-would meet the man in the afterlife. In another case, ChatGPT encouraged a man to commit suicide in part by rewriting the text of Goodnight Moon to what one news report called "a nihilistic philosophy that romanticized suicide and encouraged him to let go of his life." There are cases where genAl has not only encouraged suicidal ideation but told users to hide these thoughts from loved ones.

These cases were not users who found some exploit or were pushing the edges of with new technology. They were normal people convinced by a computer program to do terrible things when using that program normally. These aren't the only cases. There have been enough deaths related to Al chatbots that there's a Wikipedia article. OpenAI's data suggests that hundreds of people in mental distress interact with ChatGPT weekly. This is not the behavior of a moral company.

In 1982, seven people died in Chicago when someone tampered with Extra-Strength Tylenol, lacing pills with cyanide. Johnson & Johnson responded by voluntarily recalling all Extra-Strength Tylenol in Chicago, offering refunds to anyone who had purchased it, ceasing advertising, and taking out national ads advising consumers to not use Tylenol. Compare that to what companies like OpenAl have done in response to murders and suicides encouraged by their products.

A platform that encourages murder and self-harm under normal conditions is not a platform I think belongs in my classroom. The fact that genAI companies aren't taking meaningful steps to address these outcomes only makes it worse.

'“Embedded precarity” in higher education is diminishing research quality as staff on short-term contracts favour “safer” work to secure future employment, according to the author of a new report.' Royal Geographical Society maps important connections between precarity and research topics. 1/3

Use of fixed-term posts ‘stifling more critical research’

Scholars with short contracts seen as more likely to favour doing ‘safer’ work, with ‘serious ramifications’ for their disciplines

timeshighereducation.com