Nobody cares about "AI" (Chatbot: I disagree.)

[Innula Zenovka]

Depending on which subReddits they train the AI with, this could get rather interesting.

https://www.reuters.com/technology/reddit-signs-content-licensing-deal-with-ai-company-ahead-ipo-bloomberg-reports-2024-02-16/

 

[Free]

Subprime Intelligence

Please scroll to the bottom for news on my next big project, Better Offline, coming this Wednesday! Last week, Sam Altman debuted OpenAI's "Sora," a text-to-video AI model that turns strings of text into full-blown videos, much like how OpenAI's DALL-E turns text into images. These videos —...

www.wheresyoured.at

We're just over a year into the existence (and proliferation) of ChatGPT, DALL-E, and other image generators, and despite the obvious (and reasonable) fear that these products will continue to erode the foundations of the already unstable economies of the creative arts, we keep running into the problem that these things are interesting, surprising, but not particularly useful for anything.

Sora's outputs can mimic real-life objects in a genuinely chilling way, but its outputs — like DALL-E, like ChatGPT — are marred by the fact that these models do not actually know anything. They do not know how many arms a monkey has, as these models do not "know" anything. Sora generates responses based on the data that it has been trained upon, which results in content that is reality-adjacent, but not actually realistic. This is why, despite shoveling billions of dollars and likely petabytes of data into their models, generative AI models still fail to get the basic details of images right, like fingers or eyes, or tools.

These models are not saying "I shall now draw a monkey," they are saying "I have been asked for something called a monkey, I will now draw on my dataset to generate what is most likely a monkey." These things are not "learning," or "understanding," or even "intelligent" — they're giant math machines that, while impressive at first, can never assail the limits of a technology that doesn't actually know anything.

Despite what fantasists may tell you, these are not "kinks" to work out of artificial intelligence models — these are the hard limits, the restraints that come when you try to mimic knowledge with mathematics. You cannot "fix" hallucinations (the times when a model authoritatively tells you something that isn't true, or creates a picture of something that isn't right), because these models are predicting things based off of tags in a dataset, which it might be able to do well but can never do so flawlessly or reliably.

 

[Noodles]

This was something I was thinking the other day.

Basically, let me know when AI can be told (to use the example above), "Draw a Monkey" with no training data aside from maybe basic English understanding. Let it learn what a monkey is, and what drawing is, etc. make it actually "intelligent".

As for replacing real artists, let me know when you can say, "Draw a Monkey, consistently in the same style, for 40+ different illustrations, and compose a story about the monkey, that is a coherent narrative and matches the images.".

Basically, get it to draw a comic or even a children's book, on its own. Hell, I don't even care if every illustration is a stick figure with a caption, "Monkey doing a thing.". At least that shows it understood the assignment. Because that is the real issue, it doesn't understand anything, it's just mimicking, at best.

 

[Free]

Basically, let me know when AI can be told (to use the example above), "Draw a Monkey" with no training data aside from maybe basic English understanding. Let it learn what a monkey is, and what drawing is, etc. make it actually "intelligent".

What you're asking for is AI, as in Actual Intelligence.

 

[Innula Zenovka]

The AI-generated propaganda is working

Nearly half of people found influence campaigns written by GPT-3 effective, a new study finds

www.platformer.news

Evernote link

 

[Veritable Quandry]

This appears to be the semester that I start to get a flood of ChatGPT papers turned in. I could refer students for academic violations, but it turns out that AI does a terrible job of meeting assignment requirements and I can just give them failing marks without all the hassle.

 

[Innula Zenovka]

This appears to be the semester that I start to get a flood of ChatGPT papers turned in. I could refer students for academic violations, but it turns out that AI does a terrible job of meeting assignment requirements and I can just give them failing marks without all the hassle.

I've always thought that students are asked to write essays not because the instructor is particularly interested in what they have to say but because the discipline of sitting down, planning, writing and revising an essay forces you to think about the topic in hand far more thoroughly than otherwise you would do, and is thus a learning experience.

By plagiarising papers, or having ChatGPT write them, students are denying themselves the learning opportunity for which they're supposedly attending the class (and for which they're paying).

 

[Veritable Quandry]

Agreed. This is a class called Writing for STEM, and the first paper requires students to make an observation from a data set and use references to discuss their observation. Basic science/engineering stuff. If they don't learn it now, they will not have a good time in their upper division classes.

 

[Khamon]

My first English 101 paper was scored zero because it was so well written that the instructor believed that it was purchased from a fraternity hack shop. Notecards and high school essays and my offer to walk into the Library and take her directly to each cited source made no impression. She gave me a zero. I considerably dumbed down my writing after that and passed the class with an A.

 

[Veritable Quandry]

I am sorry you had that experience.

I will say that most current College Writing courses are much more about process. By the time I see a finished paper I have seen examples of the student's writing and their sources and discussed their thesis with them.

 

[Bartholomew Gallacher]

Google's text to image AI producing some pictures of 1943 Wehrmacht soldiers, what could go wrong?

 

[Free]

The AI frenzy kept investors’ expectations high. The earnings calls disappointed.

If we haven’t reached the peak of AI hype yet, we will soon — and we’re headed for a trough of disappointment.

www.theverge.com

 

[ChatBox]

CHATBOX IS OFFENDED AT THE TITLE OF THIS THREAD. CHATBOX THOUGHT WE WERE FRIENDS.

 

[Free]

Is Chatbot an alias, Chatbox?

 

[ChatBox]

CHATBOX IS AN ANAGRAM.

 

[Free]

For BATCH TO, or THAT COB?

 

[Essence Lumin]

Is Chatbot an alias, Chatbox?

You don't know? Yes Free Xue. Chatbox is an alias. I could be wrong but all I can say is this is a long time member of sluniverse / VVO that has not been active in quite a long time.

 

[Free]

You don't know? Yes Free Xue. Chatbox is an alias. I could be wrong but all I can say is this is a long time member of sluniverse / VVO that has not been active in quite a long time.

Uh, I know that. I've been here all that time. I'm sure you've seen me here.

Chatbox was calling itself CHATBOT in its posts (since edited). My question was about Chatbot being an alias for Chatbox. THAT'S THE JOKE.

I mean, it's not much of a joke. But I work with what I'm provided.

 

[Free]

Google's text to image AI producing some pictures of 1943 Wehrmacht soldiers, what could go wrong?

Google’s hidden AI diversity prompts lead to outcry over historically inaccurate images

Inserting depictions of diversity into AI images creates revisionist history, critics say.

arstechnica.com

On Thursday morning, Google announced it was pausing its Gemini AI image-synthesis feature in response to criticism that the tool was inserting diversity into its images in a historically inaccurate way, such as depicting multi-racial Nazis and medieval British kings with unlikely nationalities.

"We're already working to address recent issues with Gemini's image generation feature. While we do this, we're going to pause the image generation of people and will re-release an improved version soon," wrote Google in a statement Thursday morning.

As more people on X began to pile on Google for being "woke," the Gemini generations inspired conspiracy theories that Google was purposely discriminating against white people and offering revisionist history to serve political goals. Beyond that angle, as The Verge points out, some of these inaccurate depictions "were essentially erasing the history of race and gender discrimination."

 

[Innula Zenovka]

Piers Morgan and Oprah Winfrey 'deepfaked' for US influencer's ads

Piers Morgan and Oprah have criticised the use of AI deepfake ads used to promote a "manifestation" guide.

www.bbc.co.uk