#Generative AI
Explore tagged Tumblr posts
victusinveritas · 11 months ago
Text
Tumblr media Tumblr media Tumblr media
179K notes · View notes
udoncurrybowl · 11 hours ago
Text
yeah!! and that's why hallucinations happen, because gen ai is essentially probablistic-based output with randomisation. there are loads of great resources teaching you about this online, so i won't go into too much detail! but essentially, the randomisation with its output is what allows gen ai to give more variety in its answers, however this also means that there's not even a fundamental guarantee that it's giving you the most frequent information.
using gen ai really doesn't beat good old fact-checking by searching the web. remember if it's something important you need to find evidence of, you can always search sites that offer free access to academic journals like jstor!
people are really fucking clueless about generative ai huh? you should absolutely not be using it for any sort of fact checking no matter how convenient. it does not operate in a way that guarantees factual information. its goal is not to deliver you the truth but deliver something coherent based on a given data set which may or may not include factual information. both the idolization of ai and fearmongering of it seem lost on what it is actually capable of doing
49K notes · View notes
luccorvus · 2 days ago
Text
Tumblr media
The Monster Wears Your Face
Written by @joles.bsky.social‬ on Bluesky
65 notes · View notes
elexuscal · 1 month ago
Text
"what did students do before chatgpt?" well one time i forgot i had a history essay due at my 10am class the morning of so over the course of my 30 minute bus ride to school i awkwardly used by backpack as a desk, sped wrote the essay, and got an A on it.
six months later i re-read the essay prior to the final exam, went 'ohhhh yeah i remember this', got a question on that topic, and aced it.
point being that actually doing the work is how you learn the material and internalize it. ChatGPT can give you a short cut but it won't build you the the muscles.
100K notes · View notes
lynati · 23 hours ago
Text
I admire art in part because I admire the effort and skill I can tell went into making it.
That admiration does not exist for a piece that is the result of someone clicking a remix button on a trove of other people's stolen effort and skill.
Honestly I can tell you finding out art was made by AI really does immediately, legitimately sour it for me, like people will trot this out as a Gotcha for anti-AI people but it's just making it clear they don't consider art to be the conversation that it is lol. It's similar to the way Harry Potter immediately soured for me because engaging with it while knowing the kind of heart Rowling is writing from changes the way the work feels; there isn't any moralizing or whatever that I have to do, it's easy to drop it because it's rotted in my hands.
"Oh but you LIKED this song before, nothing changed!" The conversational partner did. A very large portion of what is interesting to me about art is thinking of why the creator chose that instrumentation, or what made them want to make the thing in the first place. Finding out I've been talking to a wall completely removes an entire third of the force that art is to me, and I can't argue that anything about art or its consumption is Objectively Correct but I can argue it's fucking boring lmao
19K notes · View notes
prokopetz · 21 days ago
Text
A couple of years ago we were all terribly concerned about the fact that a lot of American high schools are assigning such crushing homework loads that some kids literally don't have enough time to eat or sleep (and all this in spite of the fact that there's no good evidence that assigning homework actually improves academic outcomes at the pre-university level), but now we're hearing stories about those same schools struggling to stop kids from using ChatGPT to write their essays and suddenly It's The Children Who Are Wrong. Like, do you think maybe there's a certain level of cause and effect in play here?
19K notes · View notes
leidensygdom · 4 months ago
Text
So, Discord has added a feature that lets other people "enhance" or "edit" your images with different AI apps. It looks like this:
Tumblr media
Currently, you can't opt out from this at all. But here's few things you can do as a protest.
FOR SERVERS YOU ARE AN ADMIN IN
Go to Roles -> @/everyone roles -> Scroll all the way down to External Apps, and disable it. This won't delete the option, but it will make people receive a private message instead when they use it, protecting your users:
Tumblr media
You should also make it a bannable offense to edit other user's images with AI. Here's how I worded it in my server, feel free to copypaste:
Do not modify other people's images with AI under ANY circumstances, such as with the Discord "enhancement" features, amidst others. This is a bannable offense.
COMPLAIN TO DISCORD
There's few ways to go around this. First, you can go to https://support.discord.com/hc/en-us/requests/new , select Help and Support -> Feedback/New Feature Request, and write your message, as seen in the screenshot below.
Tumblr media
For the message, here's some points you can bring up:
Concerns about harassment (such as people using this feature to bully others)
Concerns about privacy (concerns on how External Apps may break privacy or handle the data in the images, and how it may break some legislations, such as GDPR)
Concerns about how this may impact minors (these features could be used with pictures of irl minors shared in servers, for deeply nefarious purposes)
BE VERY CLEAR about "I will refuse to buy Nitro and will cancel my subscription if this feature remains as it is", since they only care about fucking money
Word them as you'd like, add onto them as you need. They sometimes filter messages that are copypasted templates, so finding ways to word them on your own is helpful.
ADDING: You WILL NEED to reply to the mail you receive afterwards for the message to get sent to an actual human! Otherwise it won't reach anyone
UNSUSCRIBE FROM NITRO
This is what they care about the most. Unsuscribe from Nitro. Tell them why you unsuscribed on the way out. DO NOT GIVE THEM MONEY. They're a company. They take actions for profit. If these actions do not get them profit, they will need to backtrack. Mass-unsuscribing from WOTC's DnD beyond forced them to back down with the OGL, this works.
LEAVE A ONE-STAR REVIEW ON THE APP
This impacts their visibility on the App store. Write why are you leaving the one-star review too.
_
Regardless of your stance on AI, I think we can agree that having no way for users to opt out of these pictures is deeply concerning, specially when Discord is often used to share selfies. It's also a good time to remember internet privacy and safety- Maybe don't post your photos in big open public servers, if you don't want to risk people doing edits or modifications of them with AI (or any other way). Once it's posted, it's out of your control.
Anyways, please reblog for visibility- This is a deeply concerning topic!
19K notes · View notes
rubyjones · 6 months ago
Text
Too many writers are using generative 'AI' to make their book covers, so I've written a guide on how to make your own cover for free or cheap without turning to a machine.
If you can't afford to pay an artist, you CAN make your own!
I hope this is a helpful overview that covers the basics and points to some free resources.
[Edit:] this is getting well outside this account's usual reach and a few people in the notes are making not so great comments about the fact I write weird erotica. So, for full disclosure: yes, I am an erotica writer. And the book covers include those for tentacle erotica and robot erotic romance. But you cannot see anything NSFW. All of these covers were approved by Amazon. Frankly, the cover of the most NSFW story, Oviposition, is downright tame! And to those complaining about the word 'Oviposition', that is a scientific word for a naturally occurring phenomenon - it's only naughty if you know the content of the story, which is only alluded to as 'sexy'.
I also do not believe that shirtless men require a warning, but yeah, two of the covers have shirtless men. The cover with a tasteful nude drawing (showing a bum, but not a weenor) has already been censored.
I can't do anything about the version of this post that got rb'd by a popular person, but for those who are concerned, consider this your content warning.
24K notes · View notes
coolmaycroft · 2 days ago
Text
Some people call it theft, and get very technical with the definition. Even if there was no theft per se, the models are trained on images/audio/video from existing artists, with no respect to consent on their part. That´s at least taking advantage.
And like, not to cheer on it, but at some point, some big corpo was going to put a lawsuit against generative models. Not to mention there were already several demands in process from other individuals and groups.
The scary part is that this is fucking disney, they have shaped copyright and trademark to their favor. So we're yet to see how this changes the landscape
i deploy this phrase quite often but i cannot emphasize enough how much i truly mean this: anyone who is cheerleading the universal and disney lawsuit is an enemy of art and everything that's valuable about it
4K notes · View notes
prismagic · 4 months ago
Text
Discord has added generative AI features to servers.
Here are THREE THINGS you need to do today.
1️⃣ YOUR SERVER
Go into server settings -> roles -> Default permissions
Search for "external apps" and disable it
Tumblr media
2️⃣ YOUR FEEDBACK
Access Discord support -> https://support.discord.com/hc/en-us/requests/new
Go into Help & Support > Feedback/New Feature Request
Denounce generative AI and request the feature be removed. Bring up privacy concerns for yourself, your data, and for minors.
You will get an automatic reply to this request- PLEASE REPLY TO THE EMAIL TO PUSH IT TO THE TEAM or else it will get discarded.
3️⃣ ON MOBILE
Access your app store
Leave a 1 star review and publicly denounce generative AI and Discord's use of it
Reblog this to spread the word, I'm tired of the enshittification of the things I use in my day to day life.
20K notes · View notes
victusinveritas · 9 months ago
Text
Tumblr media
24K notes · View notes
dicklessthewonderclown · 5 months ago
Text
ai has no fucking place in fandom. none whatsoever. even if you’re “just using it for feedback” or “just using it for prompts” or “just using it to help with writer’s block.” and if you’re using it for research? why on god’s green earth would you use the Lying Machine to do research???
all art, whether written work, paintings, drawings, anything, is a fundamentally human creation, and by using ai to create “art,” you are stripping the humanity from something inherently, intrinsically human
fanfiction and fanart and any kind of fan creation are borne out of a love of the original media and a passion for creation. there is no reason or need to supplement that passion and creation with gibberish from the Plagiarizing Machine
nobody expects perfection. it is the imperfections that make it beautiful, that make it raw and rich with humanity. this is a community that exists because of a shared love for something.it is built upon our own creations. ai generated content threatens its very foundations
9K notes · View notes
papercranesong · 18 hours ago
Text
Hey, I get where you’re coming from 👍 But I was trying to think of a more apt analogy, because autofill and ChatGPT are fundamentally different in the way they’re built and how they work…
Ok, how about this -
So autofill is like walking into a library, handing over the first few words of your sentence and asking the librarian to finish it using the most common phrase in the most-read book.
(Basically asking: “How do people usually finish this thought?”)
ChatGPT is more like walking into the library and saying: “Write me an article on how to build an aeroplane.”
And a very well-read librarian - one who’s read all the books but doesn’t actually understand flight - puts something together. It sounds plausible, and maybe even really helpful. But here’s the thing to be aware of:
They might fabricate details.
They might cite diagrams that don’t exist.
They might give you a plane that looks fine - until it doesn’t fly
I think asking the generative AI for things - basically, prompt engineering - is a real skill in itself, and something I want to get better at.
Whilst also being aware that the plane it builds might not fly all that well… so it’s best to supervise the AI!!
(This is the first time of me road-testing this analogy, feel free to push back on it!)
"I asked Grok.""I asked Chat gpt." ok, well, i asked Sam winchester, and he said,"So get this...
Tumblr media
4K notes · View notes
animentality · 1 year ago
Text
Tumblr media
26K notes · View notes
incognitopolls · 3 months ago
Text
This poll is asking about voluntary use of generative AI. For the purposes of this poll, do not count instances where you were required to use generative AI (e.g. for a school assignment, job required it, etc).
We ask your questions anonymously so you don’t have to! Submissions are open on the 1st and 15th of the month.
3K notes · View notes