Chris Manion (he/him)
LightNews LightNews
banner
jazzmanion12.bsky.social
Chris Manion (he/him)
@jazzmanion12.bsky.social
49 followers 94 following 40 posts
Writing Across the Curriculum Coordinator, Center for the Study and Teaching of Writing at The (I know I know) Ohio State University. Jazz dad. Don't use DMs here for reasons.
Posts Replies Media Videos
Reposted by Chris Manion (he/him)
timnitgebru.bsky.social
Wildly different things, tasks, techniques, subspecialties being lumped into "AI" and then being conflated with each other, doesn't help. Different types of models vs the techniques to train them vs the tasks they are supposed to accomplish, all under "AI".
December 22, 2025 at 5:21 PM
Reposted by Chris Manion (he/him)
ecourtem.bsky.social
Amazing 😳 digital archaeology

How AI’s annoying blandness “accidentally replicated the linguistic ghost of the British Empire”
December 16, 2025 at 7:47 PM
Reposted by Chris Manion (he/him)
justinbrake.bsky.social
This is the 2nd time a major government report from Deloitte has been found to contain errors likely generated by AI.

First in Australia and now, as The Independent has confirmed, in a major healthcare policy paper for the Government of Newfoundland and Labrador #nlpoli #AI #deloitte
November 22, 2025 at 3:35 PM
Reposted by Chris Manion (he/him)
mattbaume.bsky.social
To anyone dismayed about not getting Muppet Knives Out don't worry because they basically already did that in 1979 with Liza Minnelli and it's a perfect (Emmy-nominated!) episode of TV that you can enjoy right now!
lisa looking at a dead kermit kermit talking to suspects backstage lize really going for it liza slaying in a 40s noir outfit
November 19, 2025 at 1:31 AM
Reposted by Chris Manion (he/him)
hagenblix.bsky.social
I do, of course, have a paragraph from "Why We Fear AI" for this
An LLM is an abstraction over the status quo (or rather, the text of the status quo) that has, at its core, no other aim than producing more text that is just like the text already in existence. One result is that the dominant ideas, the dominant textual patterns of our time, are recast as “intelligence” per se. We might call the complex mass of concepts and ideas, the channels into which ideas flow often and easily, or the boundaries of what is reasonable to debate in so-called “polite society”—the dominant ideology of an age. An age’s dominant ideology tends to presuppose that the power of that age is natural; the hallmark of ideology, in this sense, is that it is easy to recognize only when it’s past its time. We may no longer think that an almighty god appointed a king to rule us, that women are naturally unfit to vote, or that sex outside of a heterosexual marriage is immoral. But at various times, to think otherwise (to think of replacing the monarchy with a republic, to affirm that women are people, or being in favor of sexual or gay liberation) was the far more difficult, even dangerous thought—if not to have, then at the very least to put in writing. Ideology, of course, is never complete, other thoughts are always possible. But ideology is easy to think, it’s what our thoughts are already marinated in. And hence, ideology will be dominant in the textual patterns of an age. Seen in this light, what is called “intelligence” in a language model, is really that—ideology as a thing that can pretend to be a person.
November 11, 2025 at 8:36 PM
Reposted by Chris Manion (he/him)
carlzimmer.com
Largest study of its kind shows AI assistants misrepresent news content 45% of the time – regardless of language or territory. www.bbc.co.uk/mediacentre/...
Largest study of its kind shows AI assistants misrepresent news content 45% of the time – regardless of language or territory
An intensive international study was coordinated by the European Broadcasting Union (EBU) and led by the BBC
www.bbc.co.uk
October 23, 2025 at 5:17 PM
Reposted by Chris Manion (he/him)
tressiemcphd.bsky.social
Okay so a I have a true little story about how the episcopal church funded the local BPP chapter’s ambulance program and all I’m saying is, Episcopalians: they may surprise you!
hollyanderson.bsky.social Holly Anderson 🍇 @hollyanderson.bsky.social · Oct 15
you fucked around and now the Episcopalians are doing memes. are you happy now. are you
Screenshot of a Facebook post featuring three of the Portland protesters wearing inflatable frog costumes with the following text: Episcopalians on Facebook Elizabeth Rose Elrod • 22h • Exodus 8:2-6 "But if you refuse to let them go, I will plague your whole country with frogs... The frogs shall come up on you and on your people and on all your officials."
October 15, 2025 at 5:18 PM
Reposted by Chris Manion (he/him)
doctorwaffle.substack.com
In honor of National Poetry Day, the greatest parody rewrite of all time:
Screen cap of parodic version of William Blake's "The Tyger" that begins: Tyger! Tyger! Burning bright (Not sure if I spelled that right) What immortal hand or eye Could fashion such a stripy guy? What the hammer that hath hewn it Into such a chonky unit? Did who made the lamb make thee, Or an external franchisee?
October 2, 2025 at 3:17 PM
Reposted by Chris Manion (he/him)
markriedl.bsky.social
If you use any of my How LLMs Work blog posts in your classes, could you please let me know. I'm applying for a thing 🙏.

The Intuition Behind How Large Language Models Work
medium.com/@mark-riedl/...

A Very Gentle Introduction to Large Language Models without the Hype
medium.com/@mark-riedl/...
The Intuition Behind How Large Language Models Work, Part I
Large Language Models (LLMs) are fancy artificial neural networks. But you don’t have time to learn the math or engineering. Unfortunately…
medium.com
September 30, 2025 at 1:31 AM
Reposted by Chris Manion (he/him)
thetattooedprof.bsky.social
Lmaooooooo
Education report calling for ethical AI use contains over 15 fake sources
Experts find fake sources in Canadian government report that took 18 months to complete.
arstechnica.com
September 23, 2025 at 8:42 PM
Reposted by Chris Manion (he/him)
derekbruff.bsky.social
If you've been trying to get a read on "AI literacy," the UVA Teaching Hub has a great new collection for you. "Frameworks and Activities for Fostering AI Literacy" has been curated by librarian Bethany Mickel and instructional designer Fang Li. teaching.virginia.edu/collections/...
Frameworks and Activities for Fostering AI Literacy — UVA Teaching Hub
<p>This collection features frameworks for understanding AI literacy—including one framework developed here at UVA—as well as classroom activities that support the development of AI literacy in both s...
teaching.virginia.edu
September 23, 2025 at 1:32 PM
Reposted by Chris Manion (he/him)
anildash.com
Okay, this one’s a little bit diffferent — a quick video from me, wondering if “Taylor’s Version” might give us a glimpse to a way forward in the war against AI robots scraping up all the content of the internet without consent or compensation. youtu.be/X7jbRY3MvpQ
Taylor's Version vs the AI Robot Invasion
YouTube video by Anil Dash
youtu.be
September 23, 2025 at 5:04 PM
Reposted by Chris Manion (he/him)
annakornbluh.bsky.social
in education, creative industries, caring professions, and LABOR ORGANIZING, people are doing things, there's a movement

against-a-i.com

www.bloodinthemachine.com/p/the-luddit...
The Luddite Renaissance is in full swing
This fall, the new luddites are rising
www.bloodinthemachine.com
September 21, 2025 at 2:56 PM
Reposted by Chris Manion (he/him)
profmarciniak.bsky.social
Reading papers is a basic skill (and dare I say duty?) of scientists, and I include medical doctors in that group. Keeping abreast of the literature is a foundational part of our professions. There aren’t good shortcuts. In any case, reading papers regularly is fun.
arstechnica.com/ai/2025/09/s...
Science journalists find ChatGPT is bad at summarizing scientific papers
LLM “tended to sacrifice accuracy for simplicity” when writing news briefs.
arstechnica.com
September 21, 2025 at 6:22 AM
Reposted by Chris Manion (he/him)
blmckean.bsky.social
“OpenAI’s own advanced reasoning models actually hallucinated more frequently than simpler systems. The company’s o1 reasoning model ‘hallucinated 16% of the time’ when summarizing public information, while newer models o3 and o4-mini ‘hallucinated 33% and 48% of the time, respectively.’”
September 20, 2025 at 12:09 PM
Reposted by Chris Manion (he/him)
ncecire.bsky.social
I'm on the fence about some of this, but it takes the current problems seriously. Even the early, idealistic ideology of OA was premised on the idea that access above all meant 1. digital and 2. access above all to stuff.
September 11, 2025 at 10:37 AM
Reposted by Chris Manion (he/him)
osuaaup.bsky.social
www.wosu.org/politics-gov...
Ohio State University prohibits land acknowledgements under most circumstances
A land acknowledgement is a verbal or written statement that recognizes that land was taken from Indigenous people. Ohio State's policy says they are considered statements "on behalf of an issue or ca...
www.wosu.org
September 9, 2025 at 7:47 PM
Reposted by Chris Manion (he/him)
keptsimple.bsky.social
like so many products to come out of silicon valley in the last decade or so, AI is a legal innovation masquerading as a tech innovation. the legal theory seems to be that AI is entitled to everything but liable for nothing.
arstechnica.com Ars Technica @arstechnica.com · Aug 27
AI advocates have warned that if every author in the class action filed a claim, it would "financially ruin" the entire industry.
Authors celebrate “historic” settlement coming soon in Anthropic class action
Advocates fear such settlements will “financially ruin” the AI industry.
arstechnica.com
August 27, 2025 at 6:10 PM
Reposted by Chris Manion (he/him)
anildash.com
This is a very good post to send to your friends and colleagues who use AI, teaching them how to learn how to get more correct answers out of LLMs. AI haters should read this too, as it’s research-based and you can see how these systems work when you encounter them.
September 7, 2025 at 2:33 PM
Reposted by Chris Manion (he/him)
olivia.science
Finally! 🤩 Our position piece: Against the Uncritical Adoption of 'AI' Technologies in Academia:
doi.org/10.5281/zeno...

We unpick the tech industry’s marketing, hype, & harm; and we argue for safeguarding higher education, critical
thinking, expertise, academic freedom, & scientific integrity.
1/n
Abstract: Under the banner of progress, products have been uncritically adopted or even imposed on users — in past centuries with tobacco and combustion engines, and in the 21st with social media. For these collective blunders, we now regret our involvement or apathy as scientists, and society struggles to put the genie back in the bottle. Currently, we are similarly entangled with artificial intelligence (AI) technology. For example, software updates are rolled out seamlessly and non-consensually, Microsoft Office is bundled with chatbots, and we, our students, and our employers have had no say, as it is not considered a valid position to reject AI technologies in our teaching and research. This is why in June 2025, we co-authored an Open Letter calling on our employers to reverse and rethink their stance on uncritically adopting AI technologies. In this position piece, we expound on why universities must take their role seriously toa) counter the technology industry’s marketing, hype, and harm; and to b) safeguard higher education, critical thinking, expertise, academic freedom, and scientific integrity. We include pointers to relevant work to further inform our colleagues. Figure 1. A cartoon set theoretic view on various terms (see Table 1) used when discussing the superset AI (black outline, hatched background): LLMs are in orange; ANNs are in magenta; generative models are in blue; and finally, chatbots are in green. Where these intersect, the colours reflect that, e.g. generative adversarial network (GAN) and Boltzmann machine (BM) models are in the purple subset because they are both generative and ANNs. In the case of proprietary closed source models, e.g. OpenAI’s ChatGPT and Apple’s Siri, we cannot verify their implementation and so academics can only make educated guesses (cf. Dingemanse 2025). Undefined terms used above: BERT (Devlin et al. 2019); AlexNet (Krizhevsky et al. 2017); A.L.I.C.E. (Wallace 2009); ELIZA (Weizenbaum 1966); Jabberwacky (Twist 2003); linear discriminant analysis (LDA); quadratic discriminant analysis (QDA). Table 1. Below some of the typical terminological disarray is untangled. Importantly, none of these terms are orthogonal nor do they exclusively pick out the types of products we may wish to critique or proscribe. Protecting the Ecosystem of Human Knowledge: Five Principles
September 6, 2025 at 8:13 AM
Reposted by Chris Manion (he/him)
kottke.org
Apologies: You Have Reached the End of Your Free-Trial Period of America! “We are retaining some features for premium users. Want rule of law? That’s premium. The right to run your company without government interference? That’s a paid feature now.” [theatlantic.com]
Apologies: You Have Reached the End of Your Free-Trial Period of America!
Want rule of law? That’s premium.
www.theatlantic.com
September 3, 2025 at 1:08 PM
Reposted by Chris Manion (he/him)
barilleon.com
When you're working on a solo project and you lose so much momentum that everything grinds to a screeching halt, what you are suffering from is a lack-of-feedback loop. It's time to show what you have to someone who is not you.

Is what I'm doing worth it? Is it good? Entertaining? GO FIND OUT!
September 1, 2025 at 3:59 PM
Reposted by Chris Manion (he/him)
anildash.com
If people want some of the things AI can do, have those technologies be provided by tools & platforms created by cooperatives, by unions, by universities and governments, by municipalities and by individual creators, or as open source owned by nobody. Destroy the economic value of proprietary tools.
August 26, 2025 at 8:32 PM
Reposted by Chris Manion (he/him)
chanda.blacksky.app
It enrages me how useful grant proposals are lol

Also, related, one of the ways that cancelling future grant opportunities harms science is that it scales back a process that often helps scientists really think through our ideas

When we do less of it, the science can suffer
August 27, 2025 at 1:21 AM
Reposted by Chris Manion (he/him)
jeffgreene.bsky.social
Great article with one of the best brief layperson introductions to AI v. LLMs that I’ve seen. And I love the exploration of whether and how LLMs’ are useful depends on the user’s level of expertise. #PsychSciSky #AcademicSky #EduSky
doi.org/10.1037/mac0...
Screenshot of title page of article published in the Journal of Applied Research in Memory and Cognition titled "Expert Thinking With Generative Chatbots."
August 25, 2025 at 12:02 PM