C. E. Aubin
I have seen a lot of cursed stuff in my time in academia but this is among the *most* cursed. Grammarly is generating miniature LLMs based on academic work so that users can have their writing ‘reviewed’ by experts like David Abulafia, who died less than two months ago.
Mar 3 at 11:58amWeb
Show threadC. E. AubinMar 3
If this were me I’d be haunting their asses so hard. Every cabinet flying open, candles snuffed out, hot breath on the back of their necks, chains rattling in the night, the whole shebang
Show threadC. E. AubinMar 3
If anyone tries to “um actually 🤓” me about these being LLMs to score goober points instead of engaging with the very obviously unethical substance of the problem, I hope you get haunted too
Show threadC. E. AubinMar 3
This is from @[email protected] !!!!!!
Show threadV For ViennettaMar 3
It really upsets me that there's no hell for someone to rot in if they do something that despicably gruesome.
Show threadOmnivoreMar 3

@ceaubin.com

#alttext

Vanessa Heggie 2nd Associate Professor of the History...

I don't know where to start with this but... Grammerly is now offering "expert review" of your work by living and dead academics. Yes, dead ones - without anyone's explicit permission it's creating little LLMs based on their scraped work and using their names and reputations.
Obscene.

It may be possible to opt out of having YOUR work used to train genAl - https://lnkd.in/eXZB6TY8

Expert Review
Selected experts
DA
David Abulafia
Historian specializing in Mediterranean and global medieval history, author of 'The Great Sea'
FF
Finbarr Barry Flood
Art historian and expert in medieval material mobility, author of 'Objects of Translation'
CW
Chris Wickham
Historian of medieval Europe and author of 'Framing the Early Middle Ages'

Grammarly's new TOS: How to opt out of AI training | Shari Berg posted on the topic | LinkedIn

If you're not keen on tools and apps using your content to train their AI, you'll want to pay attention to Grammarly's new TOS. As part of its rebranding to Superhuman (yes, that's a real thing), the editing app will use anything you upload to it to train its AI. For those of us who may work with proprietary information or other sensitive data, you can turn off this permission. Here's how. - Log in to your Grammarly account - Go to your privacy settings (shortcut is https://lnkd.in/gtZrrDMt) - Find the toggle labeled "Product Improvement and Training" and switch it to the off position Grammarly promises not to use your content for AI model training and product improvement if you do this. However, some non-content data (like usage stats) may still be collected by the app. If an organization manages your account, they'll need to make these changes for you. While we're chatting about Grammarly, let's also discuss whether rebranding to Superhuman is a good thing or off the mark. I'll share my thoughts in the comments after you've had a chance to weigh in. Image Description: A screenshot of the privacy settings for Grammarly, with a red arrow indicating where to toggle off the AI data-sharing permissions.

Show threadSamuelJohnsonMar 3
@ceaubin.com Jeez, what if we could get a Jesus LLM and have it do some coaching for Republicans?
Show threadhartreefoxMar 3

@ceaubin.com

Similar vibe: People on YouTube offering up AI generated vids of Cal Tech physicist Leonard Susskind (For example. There are many such including Hawking ones) talking on physics subjects.
No indication given that it's not him or even reviewed by him and saying things of questionable veracity.

Show threadStephen Brooks 🦆Mar 4

@ceaubin.com Right but the main use case of this is going to be getting the LLM of yourself to do your work for you.

That, and academics who feel ignored can pretend to have attention from famous people. 😆