Ambrose

Oh thank you, working my way through Cognition and Intractability, and read some of these, but will add these to the list.

They're right, but AI makes this worse as many orgs are using AI as moral cover as pointed out by Borinca here. "Moral disengagement helps explain how users maintain ethical self-image while enabling discrimination. AI tools create psychological distance that weakens moral accountability."

My employer, Dartmouth College, today boasts it's 1st Ivy "to launch AI at an institutional scale." It is doing this by partnering--"more than a collaboration"--with Anthropic, a company that stole the books of many faculty, me included, which many of us are suing.

Such a good choice in book too!

Am I experiencing a hermenuetical harm because I have no idea what this dude is talking about, but I'm pretty sure it is hurting my brain. (Much like Heideggers philosophical legacy of colonial domination through technology nearly a century later).

Well there are LTNs and other neurosymbolic methods (My lab does a lot of work in neurosymbolic). But these don't work with LLMs in the way people intuit. And the part that the tech bros care about: they don't scale/generalize.

If your business model requires dependency, then deskilling is far-more than just a convenient side effect.

Guess which age japanese kids start taking the train alone.

Hoping more countries, but more realistically parents, realize how much harm unfettered access to these technologies is doing to these kids brains and follow suit.

TikTok's algorithm directs 13-year-olds to porn TikTok suggests sexually explicit search terms to 13-year-olds directing them to porn, in apparent breach of UK law, Global Witness finds

One of my favorites is this report from Global Witness which found that TikTok's algorithm directs children to explicit content and sexually explicit searches continuously.

The Decline in Adolescents’ Mental Health with the Rise of Social Media: A Narrative Review - Kristin Burgess, 2025 Mental illness in adolescence has seen an unprecedented increase over the past decade. This has left parents, teachers, health-care providers, and other stakeho...

This paper from May does an excellent look into the many studies on this journals.sagepub.com/doi/10.1177/...

Just yesterday a friend pointed out to me how different today's teens are, like they're forced to grow up. We have lots of literature that pretty conclusively shows that "brain rot" isn't just a meme. Social media platforms are literally designed to erode mental health and change behaviour.

What purpose and benefit do data brokers serve, cause it seems like they're solely a vehicle of surveillance and profit. Meanwhile, LLMs struggle for the latter. I would also like to point out that this issue would be easily mitigated by blacklisting these brokers in RAG. This is a design choice.

What's keeping you reading it, out of curiosity?

loved this conversation. Anyone need a crash course on why the intersection of AI, fascism and climate breakdown? We tried to cover it all!

Um, if an LLMs's tendancy to lie and cheat isn't something that can be stopped, maybe this isn't a great product to incorporate at every level of our professional and personal lives?

Like I genuinely cannot wait to be able to do this with my own research.

Yeah, sorry I was just joking. Cause like the funding environment right now is so tragic. I 100% agree with you.

Last week @theverge.com published my essay exploring the limitations of large-language models. This week, that same essay is cited by a federal judge in Michigan to distinguish the process of human reasoning from what these models do. Very, very gratifying.

Please compare and contrast the Critical AI Literacies framework offered by @olivia.science @marentierra.bsky.social @irisvanrooij.bsky.social with the OECD/Euro Commission/Code.org AILit framework.

Do you have any doubt which one will leave students better prepared for the future?

Need funding for this.

The point being: it should be clearer to able bodied people that the price of your suspicions of whether someone who says they are disabled is disabled or not are paid overwhelmingly by actually disabled people themselves. And what you are doing is making sure that disabled students are not helped

“HIS DOG IS HIS ‘EYE’”: The Guide Dog in Public Spaces and Transit, 1930-1970 - All of Us Writing the first editorial for AllOfUs has not only been a joy, but it has also been an honor. What has always delighted me about disability history is how much our field owes its very existence to t...

Anyway, if you think it’s EASY having a service dog in public and people aren’t suspicious and sometimes abusive, let me tell you otherwise. The sad thing it’s always been like this, I wrote about it a while back for the @disability-history.bsky.social’s blog AllOfUs:

allofusdha.org/research/his...

Thank you so much for writing this. There are so many misconceptions about service dogs that just make everything so much more difficult. It's especially difficult when the disabilities they're supporting aren't super visible.

As a disabled person in higher education... I would not be able to survive, much less succeed without my accomodations. That's the point. I see students who need more support and they're not getting it because of lack of diagnosis or stigmas.

Colonial thinking, and colonial research investment is more interested in an imaginative "us vs them" threat, than the immediate harms of the climate crisis against all.

Read this, this morning and it's amazing work. Had a thought on Slow Science.

Much of the drive to research quickly comes from strategic fears of adversaries achieving advances that will disadvantage us. But climate change is a bigger threat to the west than the east, yet also is a threat to both.

If AI wants epistemic credibility, it has to reconnect data to the practices and conditions that give knowledge meaning.

LLMs also lack both of the epistemic foundations that Bender and Rogoff suggest are necessary for meaningful knowledge production. Which may be unsurprising given the research behind their continued development in industry also seems to largely lack these foundations.