“Even the term 'Artificial Intelligence' … is widely misused, with conceptual unclarity coopted to advance industry agendas and undermine scholarly discussions. It is our task to demystify and to challenge 'AI' in our teaching, research & engagement with society.” openletter.earth/open-letter-... - ThreadSky

irisvanrooij.bsky.social • 2 days ago

“Even the term 'Artificial Intelligence' … is widely misused, with conceptual unclarity coopted to advance industry agendas and undermine scholarly discussions. It is our task to demystify and to challenge 'AI' in our teaching, research & engagement with society.” https://openletter.earth/open-letter-stop-the-uncritical-adoption-of-ai-technologies-in-academia-b65bba1e

Comments

zencker.bsky.social•1 day ago

Sounds so much better than “broken search engine connected to defective speech interpreter”

paolosandro.bsky.social•1 day ago

📌

michae.lv•2 days ago

I broadly agree with those five points (and they are in line with the direction our Faculty has published and expressed on AI) but fwiw the first page of breaching ‘the spirit of the AI Act’ is really incorrect and makes me as a legal academic kind of unable to sign.

olivia.science•2 days ago

oh, I missed this, do you want to share more?

michae.lv•2 days ago

the AI act is explicitly designed to promote the unfettered uptake of tech like these in the uni sector by banning any other AI regulation in the EU other than that which the act proscribes—it is about the free movement and uptake of AI services as long as high risk ones self certify as trustworthy

michae.lv•2 days ago

it is centred on the idea that it is the characteristics of the tool, not the use of the tool (beyond ‘follow the instructions on the label’), which needs regulating. it is the exact opposite of your letter and principles.

olivia.science•2 days ago

Ah, very good to know (and disappointing) — I did not know this. Thank you. In future writing I will not mention this law. FWIW what I meant is that the EU AI act (obviously I might be wrong) wants AI to preserve human dignity and the tech sector doesn't care about that.

michae.lv•2 days ago

that’s how it’s been sold but those are effectively PR words in the non binding recitals unfortunately. it only has that substantive effect with the misconception that you can make a good tool that is good regardless of how it is integrated/used — clearly not a misconception you operate under

kissemjolk.bsky.social•2 days ago

I mean, where I'm at, human dignity is inviolable, and it is the duty of all state authority to respect and protect it.

But sometimes yeah, the state disagrees on exactly how fair human dignity extends.

emmalbriant.bsky.social•2 days ago

And ChatGPT is not considered high risk

olivia.science•2 days ago

Laughable if not horrific.

jahbottle.bsky.social•1 day ago

wish they would do that in NZ instead of running toward it with open arms

katjathieme.bsky.social•6 hours ago

📌

benecal.bsky.social•2 days ago

📌

vicozandy.bsky.social•2 days ago

It should be standard for teachers to say "If you use LLMs for writing, you are wasting your time (it's probably wrong), crippling your brain (research skills need practise), and making yourself dumber (rote work makes you dull). You have been warned!"
Also, ability is fun, being dependant sucks.

progpk.bsky.social•2 days ago

📌Academia letter on AI policies. Looking forward to watching it progress

aknupsky.bsky.social•1 day ago

👏🏻👏🏻

duhe.bsky.social•2 days ago

https://bsky.app/profile/duhe.bsky.social/post/3lpqwcxxkxs2b

gurkon.bsky.social•5 hours ago

what about the use of AI among teaching staff? should professors also be prohibited to use AI? is there a need to include this point in the letter?

stopgenai.com•3 hours ago

Obviously yes. 🤦🏻‍♀️

roneda.bsky.social•2 days ago

📌

runefar.bsky.social•1 day ago

I think you are right when it comes to demystify and challenging AI, but I also do think you are ignoring the impact and benefit that aspects of teachers at least attempting to implement AI has for disabled researchers like me. You arent fighting the machine, you are fighting people like me

agnesbookbinder.bsky.social•1 day ago

In what way?

runefar.bsky.social•1 day ago

Those of us who have to utilize features like speech to text, benefit from more accessibility in the form of alt text and content summaries that fuffil WGA3 accessibility standards are often fought aganist specifically because of how people think about AI

agnesbookbinder.bsky.social•1 day ago

Those are accessibility features & they’re fairly standard these days. Are you saying people are confusing those with LLMs/GenAI?

runefar.bsky.social•1 day ago

People justify fighting against those things quite commonly because they associate them with AI

agnesbookbinder.bsky.social•1 day ago

That’s unfortunate & misguided. There’s an education piece that needs to happen, & it is really hard to get people to think differently about disability when they think they get it.

runefar.bsky.social•1 day ago

More additionally, many of us use and build on AI as a expression of ourselves in a more accessible way because of the aspects of ourselves that are disabled. This isn't about doing less work but often requires a distinct form of creativity of its own. More so we need to discuss how socratic method

runefar.bsky.social•1 day ago

no matter if you like ai could be beneficial simply because of what we know from the advances in learning science and metacognition. Even if you disagree with AI, there is a need for change that itself is being resisted through this discussion of AI. It is just another form of system justification

agnesbookbinder.bsky.social•1 day ago

If I understand you correctly, the issue is with academia & approaches to making sure disabled folks are included while they’re critically evaluating the new technology? Because academia wasn’t overly disability friendly before now. That should be a consideration in the evaluation of the tools.

misseli.bsky.social•2 days ago

📌

ronaustin.bsky.social•2 days ago

AI?
Oh, you mean 'spell check' on steroids?

olivia.science•2 days ago

Hey don't be rude to steroids. Steroids work. 🤣

partickle.bsky.social•2 days ago

👀

irisvanrooij.bsky.social•2 days ago

If you agree with our 5 requests to our universities, please sign 🖊️ the open letter and don’t forget to confirm your email! ☺️🙏

https://openletter.earth/open-letter-stop-the-uncritical-adoption-of-ai-technologies-in-academia-b65bba1e

drgbuckingham.bsky.social•2 days ago

Have you come across any well articulated policies for banning ai use in coursework? I suspect our permissive policy is driven by worries about being unable to show ai use beyond reasonable doubt (fake references seems to be the only useful diagnostic indicator at the moment)

irisvanrooij.bsky.social•2 days ago

I think it is a mistake to think that if we cannot detect then we cannot forbid. Fraud & plagiarism are forbidden. Often they may go undetected (like people may drive too fast or through a red light), does not mean they are not violations of scientific integrity (or law, in case of the examples).

drgbuckingham.bsky.social•2 days ago

Fair point, and I don't disagree. But I've not seen this well articulated by a university yet (perhaps none have, which is why you wrote this in the first place)

michae.lv•2 days ago

we specifically articulate it here throughout the assessment section (it’s woven through but most prominent in the screenshotted part) https://bsky.app/profile/michae.lv/post/3lois5uhia22o

olivia.science•2 days ago

amazing work; consider signing in solidarity 🥰
https://bsky.app/profile/olivia.science/post/3lslki6gm7c2r

valoisdubins.bsky.social•2 days ago

UCL AI guidelines 🔖

valoisdubins.bsky.social•2 days ago

Does the ban on AI use extend into eg AI assisted literature searches?

irisvanrooij.bsky.social•2 days ago

Not happy with the first sentence, but just sharing as example.

irisvanrooij.bsky.social•2 days ago

Our policy is that genAI is not an authorised tool for assessments. It is as simple as that. Just as you also are not allowed to let someone write your thesis. People could do that, could even hire people to do it for them. But that would be fraud (even if undetected).

drgbuckingham.bsky.social•2 days ago

How does the academic misconduct narrative play out in practice? For people getting others to write their work, we conduct a viva to examine understanding, but that's a relatively rare problem because of the cost that it's not a big part of the narrative. That, I sense, is the fear from up top

irisvanrooij.bsky.social•2 days ago

Even if one passes viva, one could still had help w/ writing, no? I think in practice we trust that people don’t lie, like we trust scientists do not fake data (even though sometimes they do, and when detected consequences are large). I have students sign a declaration that their work is their own.

fernandoblancopsy.com•2 days ago

I was wondering if we can change something in our incentive system so that students themselves do not resort to these tools. Nobody uses a forklift for training at the gym for example 🤣
Do you have any idea?

irisvanrooij.bsky.social•2 days ago

I wrote this in a similar spirit https://irisvanrooijcogsci.com/2022/12/29/against-automated-plagiarism/

drartanalytics.bsky.social•4 hours ago

The ARMS framework is a good starting point for AI in assessments. The risks and mitigations. We teach it with a view to well critiqued understanding. Covering everything from energy and water use through to plagiarism and security. Plus the essential - don't outsource your thinking!

olivia.science•3 hours ago

Ooh, link?

gpetrick.bsky.social•2 days ago

Do you want academics from other countries? Happy to sign, but not if it isn't helpful.

irisvanrooij.bsky.social•2 days ago

Yes, please feel free to sign, thank you!

neilturkewitz.bsky.social•2 days ago

What about those of us not associated with formal institutions of learning?

irisvanrooij.bsky.social•1 day ago

You are very welcome to sign, thank you! 🙏

gpetrick.bsky.social•2 days ago

Done!

gpetrick.bsky.social•2 days ago

Will do!

bergha2.bsky.social•2 days ago

I may have missed it, but is there anything about commitments around academics not using AI in the preparation of course materials and assessments?

irisvanrooij.bsky.social•2 days ago

It is not explicitly stated but certainly included in our considerations and the spirit of the letter, see eg second sentence about faculty and leadership and see comments about standards and principles. Also point 1 is about staff being pushed to use AI in their software too.

bergha2.bsky.social•2 days ago

I think it would be strengthened by an explicit statement

taw6.bsky.social•3 hours ago

If you really think industry has never had anything to do with what Universities teach, I have a bridge to sell you.

leia4u.bsky.social•1 day ago

📌

merzmensch.bsky.social•2 days ago

Totally agree with the request number 5. Critical thinking about AI is inevitable. Yet AI is inevitable as well, becoming electricity and OS of 21st Century. It's academia (not TechCorp) responsibility to create ethical, IP-friendly and sustainable AI. Banning entire technology is not academic imho.

irisvanrooij.bsky.social•2 days ago

FWIW, I think this is not a correct reading of the letter. Note that the people leading this initiative are AI researchers and teachers. Also, “AI” as marketed is not inevitable, even if the AI industry wants to make us believe that.

merzmensch.bsky.social•2 days ago

Agreed, that industry is pushing, but we don't have to forget that industry just adapted and monetarized what researchers and academia were working on since last century. We as academia should take over, and I see in Europe many good initiatives. But, imho: developing, not banning. And rethinking AI

Comments

Posting Rules

Reply