#generative ai critical

[sempermoi]

Rant about generative AI in education and in general under the cut because I'm worried and frustrated and I needed to write it out in a small essay:

So, context: I am a teacher in Belgium, Flanders. I am now teaching English (as a second language), but have also taught history and Dutch (as a native language). All in secondary education, ages 12-16.

More and more I see educational experts endorse ai being used in education and of course the most used tools are the free, generative ones. Today, one of the colleagues responsible for the IT of my school went to an educational lecture where they once again vouched for the use of ai.

Now their keyword is that it should always be used in a responsible manner, but the issue is... can it be?

1. Environmentally speaking, ai has been a nightmare. Not only does it have an alarming impact on emission levels, but also on the toxic waste that's left behind. Not to mention the scarcity of GPUs caused by the surge of ai in the past few years. Even sources that would vouch for ai have raised concerns about the impact it has on our collective health. sources: here, here and here

2. Then there's the issue with what the tools are trained on and this in multiple ways:

Many of the free tools that the public uses is trained on content available across the internet. However, it is at this point common knowledge (I'd hope) that most creators of the original content (writers, artists, other creative content creators, researchers, etc.) were never asked for permission and so it has all been stolen. Many social media platforms will often allow ai training on them without explicitly telling the user-base or will push it as the default setting and make it difficult for their user-base to opt out. Deviantart, for example, lost much of its reputation when it implemented such a policy. It had to backtrack in 2022 afterwards because of the overwhelming backlash. The problem is then that since the content has been ripped from their context and no longer made by a human, many governments therefore can no longer see it as copyrighted. Which, yes, luckily also means that ai users are legally often not allowed to pass off ai as 'their own creation'. Sources: here, here

Then there's the working of generative ai in general. As said before, it simply rips words or image parts from their original, nuanced context and then mesh it together without the user being able to accurately trace back where the info is coming from. A tool like ChatGPT is not a search engine, yet many people use it that way without realising it is not the same thing at all. More on the working of generative ai in detail. Because of how it works, it means there is always a chance for things to be biased and/or inaccurate. If a tool has been trained on social media sources (which ChatGPT for example is) then its responses can easily be skewed to the demographic it's been observing. Bias is an issue is most sources when doing research, but if you have the original source you also have the context of the source. Ai makes it that the original context is no longer clear to the user and so bias can be overlooked and go unnoticed much easier. Source: here

3. Something my colleague mentioned they said in the lecture is that ai tools can be used to help the learning of the students.

Let me start off by saying that I can understand why there is an appeal to ai when you do not know much about the issues I have already mentioned. I am very aware it is probably too late to fully stop the wave of ai tools being published.

There are certain uses to types of ai that can indeed help with accessibility. Such as text-to-voice or the other way around for people with disabilities (let's hope the voice was ethically begotten).

But many of the other uses mentioned in the lecture I have concerns with. They are to do with recognising learning, studying and wellbeing patterns of students. Not only do I not think it is really possible to data-fy the complexity of each and every single student you would have as they are still actively developing as a young person, this also poses privacy risks in case the data is ever compromised. Not to mention that ai is often still faulty and, as it is not a person, will often still make mistakes when faced with how unpredictable a human brain can be. We do not all follow predictable patterns.

The lecture stated that ai tools could help with neurodivergency 'issues'. Obviously I do not speak for others and this next part is purely personal opinion, but I do think it important to nuance this: as someone with auDHD, no ai-tool has been able to help me with my executive dysfunction in the long-term. At first, there is the novelty of the app or tool and I am very motivated. They are often in the form of over-elaborate to-do lists with scheduled alarms. And then the issue arises: the ai tries to train itself on my presented routine... except I don't have one. There is no routine to train itself on, because that is my very problem I am struggling with. Very quickly it always becomes clear that the ai doesn't understand this the way a human mind would. A professionally trained in psychology/therapy human mind. And all I was ever left with was the feeling of even more frustration.

In my opinion, what would help way more than any ai tool would be the funding of mental health care and making it that going to a therapist or psychiatrist or coach is covered by health care the way I only have to pay 5 euros to my doctor while my health care provider pays the rest. (In Belgium) This would make mental health care much more accessible and would have a greater impact than faulty ai tools.

4. It was also said that ai could help students with creative assignments and preparing for spoken interactions both in their native language as well as in the learning of a new one.

I wholeheartedly disagree. Creativity in its essence is about the person creating something from their own mind and putting the effort in to translate those ideas into their medium of choice. Stick figures on lined course paper are more creative than letting a tool like Midjourney generate an image based on stolen content. How are we teaching students to be creative when we allow them to not put a thought in what they want to say and let an ai do it for them?

And since many of these tools are also faulty and biased in their content, how could they accurately replace conversations with real people? Ai cannot fully understand the complexities of language and all the nuances of the contexts around it. Body language, word choice, tone, volume, regional differences, etc.

And as a language teacher, I can truly say there is nothing more frustrating than wanting to assess the writing level of my students, giving them a writing assignment where they need to express their opinion and write it in two tiny paragraphs... and getting an ai response back. Before anyone comes to me saying that my students may simply be very good at English. Indeed, but my current students are not. They are precious, but their English skills are very flawed. It is very easy to see when they wrote it or ChatGPT. It is not only frustrating to not being able to trust part of your students' honesty and knowing they learned nothing from the assignment cause you can't give any feedback; it is almost offensive that they think I wouldn't notice it.

5. Apparently, it was mentioned in the lecture that in schools where ai is banned currently, students are fearful that their jobs would be taken away by ai and that in schools where ai was allowed that students had much more positive interactions with technology.

First off, I was not able to see the source and data that this statement was based on. However, I personally cannot shake the feeling there's a data bias in there. Of course students will feel more positively towards ai if they're not told about all the concerns around it.

Secondly, the fact that in the lecture it was (reportedly) framed that being scared your job would disappear because of ai, was untrue is... infuriating. Because it already is becoming a reality. Let's not forget what partially caused the SAG-AFTRA strike in 2023. Corporations see an easy (read: cheap) way to get marketable content by using ai at the cost of the creative professionals. Unregulated ai use by businesses causing the loss of jobs for real-life humans, is very much a threat. Dismissing this is basically lying to young students.

6. My conclusion:

I am frustrated. It's clamoured that we, as teachers, should educate more about ai and it's responsible use. However, at the same time the many concerns and issues around most of the accessible ai tools are swept under the rug and not actively talked about.

I find the constant surging rise of generative ai everywhere very concerning and I can only hope that more people will start seeing it too.

Thank you for reading.

[cyber-harpie]

Hilariously enough, the rise of openai has driven me to bookstores and libraries more in search of good reference material. I'm working on some levels and world building but I know I don't have a good enough handle on architectural styles, especially anything outside of my immediate experiences (which is very limited, as I'm born and raise usamerican and as we all know all our architecture is either somewhat recent or has been razed to the ground during colonization, and doesn't include the widest variety of cultural influences regardless). But I've been having trouble finding good references, especially since I'm looking for some very specific styles in large quantities and also looking for architectural styles I may have never encountered.

And, like. Pinterest has been dead and useless for a very long time but the AI glut has completely obliterated the remaining shreds of its usefulness. I could maybe find an interesting article on one specific building, or an article on a specific style which might include 5-15 photos, but what I really needed was a DELUGE of indexed reference photos, and preferably something which indicated the philosophy or purpose behind some elements of the style, a timeline of development, etc etc etc. I'm sure there's something out there on the Internet I could use, somewhere. But now to find anything I have to slog through a fuckton of generative AI content to even find it.

And before someone tried to tell me it's useful in this context - it absolutely is not. A well-designed home or commercial street is created by someone who made deliberate choices for deliberate reasons, subconcious or otherwise. So it's absolutely essential to work from real references, personal beef with genAI aside.

But anyway. I did some research and went to the bookstore and found two incredible books - one of which that breaks down and defines architectural elements, explaining their source and purpose and where they're used and by whom. And another which is a architectural world tour, and small taste of different architectural styles from all over the world, with tons of photos from real buildings. Perfect for a research starting point

But also the only reason I went is because I'm having trouble finding these resources online because AI clogs up all these spaces and there's no way to filter it out. It's not just Pinterest. Is it just me?

[polturn]

Todays scribblings

[probablyasocialecologist]

A new paper from researchers at Microsoft and Carnegie Mellon University finds that as humans increasingly rely on generative AI in their work, they use less critical thinking, which can “result in the deterioration of cognitive faculties that ought to be preserved.” “[A] key irony of automation is that by mechanising routine tasks and leaving exception-handling to the human user, you deprive the user of the routine opportunities to practice their judgement and strengthen their cognitive musculature, leaving them atrophied and unprepared when the exceptions do arise,” the researchers wrote. 

[...]

“The data shows a shift in cognitive effort as knowledge workers increasingly move from task execution to oversight when using GenAI,” the researchers wrote. “Surprisingly, while AI can improve efficiency, it may also reduce critical engagement, particularly in routine or lower-stakes tasks in which users simply rely on AI, raising concerns about long-term reliance and diminished independent problem-solving.” The researchers also found that “users with access to GenAI tools produce a less diverse set of outcomes for the same task, compared to those without. This tendency for convergence reflects a lack of personal, contextualised, critical and reflective judgement of AI output and thus can be interpreted as a deterioration of critical thinking.”

[...]

So, does this mean AI is making us dumb, is inherently bad, and should be abolished to save humanity's collective intelligence from being atrophied? That’s an understandable response to evidence suggesting that AI tools are reducing critical thinking among nurses, teachers, and commodity traders, but the researchers’ perspective is not that simple.

10 February 2025

[hausdellamorte]

these sketches have not left my mind since i saw them because the way the subplot with zara and illario was just dropped is genuinely insane

we know from the wigmaker job that zara was after lucanis because of his work as the demon of vyrantium. it's not mentioned if she already knew illario at this point and it doesn't appear as if illario was out for lucanis' neck already. sure, he was frustrated that lucanis was the favorite but he jokingly introduced himself as "master dellamorte the lesser" and when he asks about caterina it seems like illarios anger was directed at caterina, not lucanis. the illario we see in wigmaker job and the wake does NOT seem like the illario in the game.

and that change is just... dropped? not adressed at all.

we legit never find out why illario switched up on lucanis, how he met zara, what the nature of his relationship is with zara (which knowing that she is an old blood mage who uses the blood of her slaves to get rid of stretchmarks, i think we can guess)

these scenes were so very needed, because illario is a fascinating character and to have him be the shifty, jealous cousin does not seem enough, especially after we saw an entirely different illario in the other material he was in

[unohdukseen]

BioWare uses AI in Dragon Age: The Veilguard codex cards

And I think it's not only there - most of the armor looks like the use AI for references.

But about the codex: this applies to the Grey Warden cards. First, a general view:

All cards in different styles made by different artists. But my eyes caught on this:

It's the same art that was run through AI-editor. And these cards are nearby, in the same game! Maximum disrespect for the work of the artist who made the original art (for DAI btw). The codex contains a few cards with this style, let's see the details:

They all was AI generated. I thought I couldn't be more disappointed in BioWare, but it turns out there is.

[creature-wizard]

What an AI generated website can look like

Hey folks! I just encountered a website that's obviously AI generated, so I figured I'd use it as an example to help you spot websites that might be AI generated content farms!

First, the website is called faunafacts.com. And one of the first things that sticks out to me is how low-effort the logo is:

Regardless of whether a website is AI generated or not, a lazy and low-quality logo is a big clue that the website's content will also be lazy and low-quality.

If we click on Browse Animals, we see four options: Cows, Wolves, Bears, and Snakes.

Let's click Wolves.

The first thing I want you to notice is the lack of topical focus. Sure, it's all about wolves, but the content on them is all over the place. We have content on wolf hunting, a page on animals that resemble wolves, pages that explain the alleged social structure of wolves, and pages on wolf symbolism.

A website with content created by real people isn't going to be all over the place like this. It would be created with more of a focus in mind, like animal biology and behavior. The whole spiritual symbol thing here mixed with supposed biological and behavioral information is weird.

The next thing I want you to notice are the links to pages on topics that are quite frankly bizarre: "Wolf vs Mastiff: Things You Need To Know" and "Can You Ride A Wolf? (No, Because...)" Who is even looking for this kind of information in large enough numbers that it needs a dedicated page?

Then of course, there's the fact that they're repeating the debunked wolf hierarchy stuff, which anyone who actually knew anything about wolves at this point wouldn't post.

Now let's look at what's on one of the actual pages. We'll check out the wolves vs. mastiff page, and we can soon find a telltale sign of AI: rambling off topic to talk about something completely unrelated.

Both animals are carnivores. In the wild, wolves hunt large animals like bison, deer, and even elk. Sometimes, they may also hunt small mammals like the beaver.

Mastiffs, on the other hand, are mainly fed with dog food. As a dog, a mastiff left in the wild will eat anything. However, it will have difficulties hunting, as this instinct may have already departed the dogs of today.

A mastiff is not an obligate carnivore. Dogs can eat plant matter. Some say that dogs can survive on a vegetable diet.

Dogs being made vegetarians is a contentious issue. Scientifically, dogs belong to the order Carnivora. There is a movement today to convert dogs to a vegan diet. While science has nothing against it, the fear of many is that when dog owners do this, a vegan diet will certainly have an impact on the species.

This page is supposed to be comparing mastiffs with wolves, but then it starts talking about the vegan pet food movement. This happened because large language models generate text based on on what's statistically likely to follow the last text it just generated.

Finally, the website's images are AI generated:

If you know what to look for, this is a very obviously AI generated image. There's no graininess to the image, and the details are both unnaturally smooth and unnaturally crisp. It also has that high color saturation that many AI generated images have.

So there you go, this is one example of what an AI generated website can look like! Be careful out there!

[woman-eternal]

no you didn't??? what the fuck???

[depochu]

Yesterday I posted artwork using 3D image elements and I decided to share it in different formats. Mind you, this image was generated from AI which I re-edited in terms of design.

It's not perfect, but I hope you like it 🐹🤌💞

Check my Mouthwashing Playlist : https://spoti.fi/3EE98as

[musclefanboy]

If Travis becomes a bodybuilder

[solidaritygaming-fanblog]

I don't have time to swap accounts but I just watched an ai generated coca cola advertisement and I want to kill myself

[tanadrin]

the things which i think are harmful about social media are not specific apps, and short of banning all social media including tumblr (functionally impossible at this juncture), select bans of individual apps are not going to substantially mitigate those harms.

the only way to really deal with the problems of social media is for the culture to develop better discursive antibodies to the kinds of bullshit which spreads on social media--very old forms of bullshit, as a rule, which have diversified a little bit in the new environment, but which are not deeply ontologically different from the kinds of bullshit that led people to do believe stupid nonsense in any other era of human history--and for people to get better at messaging through and around social media. this is essentially the same thing that has had to happen in response to the rise of every form of mass media throughout history, though, and i think it's easy, when confronted by the challenges of a particular form of mass media, to overstate the ways in which this particular form is uniquely toxic or uniquely dangerous.

and fifty or a hundred years on from the introduction of many new forms of media and entertainment, more sober reflection has often lead us to realize that the arch terms in which we framed the dangers of those new forms were wildly overblown, and were as much the product of older generations' small-c conservatism and disdain for anything young people liked as they were for any actual social ills those new forms produced. there are a lot of millennials, on bluesky and elsewhere, who seem to think that they could never turn into grouchy old farts who reflexively disdain anything new that doesn't appeal to them, and i think they are probably incorrect in that assumption.

[helluvabossrewrite45]

Please Help the Animation Guild Members get a Proper Contract

Unrelated to HB but I will use it's tags in order for it to gain attention. Since HB is an indie animated show, I think it would fit at least somewhat. Say what you will about HB, but at least a human being made the entire thing.

The animation industry is currently struggling right now and it is predominantly at the animators due to corporations wanting to replace them with AI. There has been a contract to impose peace however, the contract doesn't indicate that AI will be banned. Animation is a significant passion and career for people and it shouldn't be robbed of it by machines just because corporations do not care for art like we do. Please help these animators by signing the petition.

[nyaa]

the thought of people tattooing themselves with designs made from bottlenecked ai freeware is so funny to me

[gunpowdercarousel]

Ngl, the fact that the Daggerheart website uses AI art is a pretty big turn off.

It just looks cheap.

[cyber-harpie]

I am well on my way to having read Every Single Prey Fanwork posted to Ao3