James Mackay
@jamesmackaycyp.bsky.social
Associate Professor of Literature and Digital Cultures. European University Cyprus. GenAI-curious.
Writers + AI: tool, co-author, or threat? 90-min live panel with @aijiang.bsky.social sky.social, Argyris Loizou, @hannahsilva.bsky.social , Haris Ioannides, @julietemckenna.bsky.social, and Tim Boucher. 16:00 EET (09.00 EST). Email j.mackay@euc.ac.cy for the link. #WritingCommunity #AIEthics
Yet another interesting essay from our November issue: @jamesmackaycyp.bsky.social and Polina Mackay share on the practice of creating cut-ups with commercial AI electronicbookreview.com/essay/experi...
Experiments in Generating Cut-up texts with Commercial AI
Most serious writers refuse to make themselves available to the things that technology is doing. I’ve never been able to understand this sort of fear. Many of them are afraid of tape recorders and the...
electronicbookreview.com
"AI is an automated amplifier of social norms and assumptions" - well put by Selena Savić, speaking at "Rethinking the Inevitability of AI" conference today. Such a productive series of talks.
Good to see ebr on here!
electronic book review@electronicbr.bsky.social · 2y agoHi, #bsky! Here's an online, open access, peer-reviewed journal of critical writing on #digitalculture, #criticalAI, #e-literature, #gaming and all forms of digital creativity aka "digital futures of literature, theory, criticism, and the arts". You can check us out here: electronicbookreview.com
Interesting piece on the non takeup of GenAI by educators. This is a group that has been heavily targeted by edutech startups, yet they clearly don't see any huge benefit. www.criticalinkling.com/p/students-u...
"Students Use AI. Teachers Do Not."
Two years later, neither ChatGPT nor generative AI live up to the hype.
criticalinkling.com
Fascinating. “Given the general sentiment about AI, scraping the web for content and the disintermediating of content creators, it’s not hard to imagine enterprising individuals adding ‘David Mayer’ …to their own content or website in an effort to make it disappear from ChatGPT’s output.” 👀
The Curious Case of David Mayer, the man ChatGPT cannot name.
ChatGPT users have noticed a curious behavior: It refuses to talk about a 'David Mayer'. We have the explanation and point out less obvious implications.
centreforaileadership.org
It seems that someone called David Mayer has succeeded in bringing OpenAI to heel. It's impossible to get the chatbot to generate that name.
This isn't true. In fact, only the part of the Internet curated by algorithm will fill up w/ slop. Human curated databases (libraries) are just fine, and still full of wonderful art. It's just that we've all got so used to having a machine do our thinking for us, and now we need #criticalAI skills.
Anthony Oliveira@meakoopa.bsky.social · 2y agoyou know, let me tell you at the vanguard of this sort of thing: the place where AI is just OBLITERATING public resources and cultural history is in Christian art it's becoming impossible to search and access artistic representations of biblical scenes that aren't slop
"Not a single Whitehall department has registered the use of AI systems since the government said it would become mandatory, prompting warnings that the public sector is “flying blind” about the deployment of algorithmic technology affecting millions of lives." www.theguardian.com/technology/2...
UK government failing to list use of AI on mandatory register
Technology secretary admits government departments are not being transparent over way they use AI and algorithms
theguardian.com
Very reassuring piece. Though I'm not sure that "2% of global energy use" is all that minimal.
Anuj Gupta@mettalrose.bsky.social · 2y agoExcellent article by Hannah Ritchie, a geoscientist, who contextualizes energy use by GAI using emerging data. I found it very helpful to think beyond the 1 prompt = 1 glass of water comparisons. Thanks for sharing @nic221.bsky.social www.sustainabilitybynumbers.com/p/ai-energy-...
Has a political party anywhere in the world ever won a democratic election with manifesto promising to cut services while raising taxes? (Honestly curious, and it's tricky to research.)
This is only a light little article, but it's the first time I've seen anyone in mainstream media talk about Ethiopian perspectives on Western charity drives. www.theguardian.com/music/2024/n...
‘Everyone was happy, but it became annoying’: Ethiopians look back on Band Aid
London cafe owner Yared Markos, 48, saw ‘aeroplanes, biscuits, sweets and cake all over Addis Ababa’ after song’s release
theguardian.com
Some observations from Digital Forms: The Long and Short of It, an interdisciplinary enquiry into length as a feature of the digital. I'll concentrate on digital cultures and #GenAI, though a lot more was discussed.
If you want to feel the lovely warm glow of human connection, go read the comments under the AI generated Coca Cola advert. "Uncanny" "Disturbing" "Dead," not to mention "Made me a lifelong Pepsi fan." youtu.be/IQWUKWM2JrQ?...
Coca-Cola - Secret Santa (AI-Generated Christmas Ad 2024)
YouTube video by LLLLITL
youtu.be
today! #AI #philosophy #empiricism #cogsci #DeepLearning
Centre for Cognition, Computation, and Modelling, Birkbeck@cccm-bbk.bsky.social · 2y agoJoin us online for tomorrow's seminar in the Cognitive Science of Generative AI series: Cameron Buckner Donald F. Cronin Chair in the Humanities, Univ. of Florida on: LLMs as models of human reasoning (21 Nov 2024, 16:00 GMT, online) registration: psyc.bbk.ac.uk/cccm/cccm-se... #LLMs #cogsci
Got to admit, I'm genuinely shocked at how good ChatGPT has got at generating poetry. Even a few months ago, it was still insistent on making poems rhyme and iterative promptinh was needed. I mean, this isn't a "great" poem. But it's also far from terrible. chatgpt.com/share/673dc8...
ChatGPT - Elevator Poem Request
Shared via ChatGPT
chatgpt.com
Recent research shows that #learning from #AI #feedback is not automatic. Furthermore, despite its individual-level benefits, AI feedback can have significant population-level downsides including loss of intellectual #diversity and an increasing #skills gap. #FutureOfWork arxiv.org/abs/2409.18660
Effects of AI Feedback on Learning, the Skill Gap, and Intellectual Diversity
Can human decision-makers learn from AI feedback? Using data on 52,000 decision-makers from a large online chess platform, we investigate how their AI use affects three interrelated long-term outcomes...
arxiv.org
Each of these documents is 10,000-15,000 words long. I've spent today proofing, standardising, cross checking them all. I'd have gone mad if it wasn't for Ants Canada videos on Youtube (www.youtube.com/@AntsCanada) (PS Ignore the absolutely insane nesting folders - it's been a long couple of weeks)
I finally saw snails mating today... such a bizarre process.
What is the very best example of art or music produced wholly by #GenAI that you've seen? Just looking for pictures / audio clips, but with source if possible.
To researchers doing LLM evaluation: prompting is *not a substitute* for direct probability measurements. Check out the camera-ready version of our work, to appear at EMNLP 2023! (w/ @rplevy.bsky.social) Paper: arxiv.org/abs/2305.13264 Original thread: twitter.com/_jennhu/stat...
One of our universities in Cyprus just launched an "MSc in Metaverse." Which may not be the best timing, given that the hype bubble around that concept disappeared about 18 months ago...
The point isn't that GenAI can write good poems now, the point is that most people have no idea what a poem is or what it's for, preferring a vague concept of "poeticness" ("poeticality"?) instead.
Nicole Hennig@nic221.bsky.social · 2y agoAI-generated poetry is indistinguishable from human-written poetry and is rated more favorably - Scientific Reports https://www.nature.com/articles/s41598-024-76900-1 #AI #poetry
I made a free course about deepfakes, hosted by my deepfake. The entire course is AI generated but based on real research. It’s sort of fun/weird/terrifying #artificialintelligence #aieducation leonfurze.teachable.com/p/this-cours...
This Course is AI Generated
leonfurze.teachable.com
The #AI feed - roughly 1/3rd entrepreneurs pitching some kind of tool with wild overpromises, 1/3rd academics/commentators saying "read my article please", and 1/3rd people showing off dodgy/exploitative softcore images they've made with GenAI. Pretty accurate reflection of the conversation, then.
It was a fun conversation at the launch of our book Reading #Instapoetry! Lots of interesting suggestions for future directions from the attendees. Many thanks to @eliterature.bsky.social for organizing this event.
People can't reliably detect academic abstracts generated by AI. But of course: abstracts are written to a formula, AI does formulaic writing well, as everywhere AI replaces drudge work. Wake me up when it writes the article (inc research) undetectably. www.sciencedirect.com/science/arti...
Can linguists distinguish between ChatGPT/AI and human writing?: A study of research ethics and academic publishing
There has been considerable intrigue surrounding the use of Large Language Model powered AI chatbots such as ChatGPT in research, educational contexts…
sciencedirect.com
Testing out the new ChatGPT-o1 preview, and am finding its supposedly enhanced reasoning facilities extremely funny. Although the conversation has mostly been about maths, I'm looking at it for literary analysis/creative writing. It's barely changed from 4, but there is 1 v funny difference.