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

[Noodles]

Try including the word 'unrequited' and see if that helps. It seems to be able to get closer to the will-they-or-won't-they line if it's clear that there's some reason it'll never quite cross it.

I can try that.

I do think something changed though, I had a decent longer back and forth that had been going for like a month that was basically all "dub conish, like a trapped Beauty and the Beast sort of deal, and itneven had flat out explicity moments, FROM the AI.

Then suddenly one day it just "I can't do anything not concenstual" and wanted to reframe the entire thing as being basicslly very very platonic.

Like have you not been paying attention for the past many thousands of words?

The wildest part, it wasn't even anytning explocity non con going on, it just, drew that conclusion on its own and refused to go on.

I really need to just set one of these things up running locally and uncesored.

And related, the idea of ever actually publishing anytjing from AI is lame, but man it makes a good digital muse. Like, I can throw general ideas at it and get some good feedback on idea that may or may not be the direction the story should go.

I forget if I mentioned it previously, but I did ask for a crituque and it vasically was kind of pissy about being too controlling of the narratives, and its like, this is my story, I am juat using you to keep the inspiration going, I am steering the story because I already known where its going.

Anyway, I am just ranting now. Just irritated because its not nearlynas useful as it used to be.

 

[Cristiano]

I used Codex for the first time in Visual Studio Code after watching the OpenAI dev livestream today. I am working on a time management cooking game prototype, and I asked it to revamp a portion of the logic. I am incredibly impressed with it. I am already using ChatGPT in my daily workflow, but this adds another powerful tool to help me be more productive.

 

[Noodles]

I have used Claude a lot for code, is ChatGPT getting better for it?

 

[Free]

I wonder if he even knows what the letters in "AI" stand for.

 

[Casey Pelous]

I wonder if he even knows what the letters in "AI" stand for.

Wild guess: For him, Artificial Information. They're experts in that field.

 

[Dakota Tebaldi]

Robin Williams’ daughter begs fans to stop sending her AI videos of late father

‘You’re not making art, you’re making disgusting, over-processed hotdogs out of the lives of human beings,’ Zelda Williams says

www.the-independent.com

She goes even further too, she really makes no mystery of exactly how she feels about AI renderings of real people.

 

[Free]

Robin Williams’ daughter begs fans to stop sending her AI videos of late father

‘You’re not making art, you’re making disgusting, over-processed hotdogs out of the lives of human beings,’ Zelda Williams says

www.the-independent.com

She goes even further too, she really makes no mystery of exactly how she feels about AI renderings of real people.

And also the people indulging in its use.

 

[Free]

I guess we've found one way that AI acts very much like us.

New Paper Finds That When You Reward AI for Success on Social Media, It Becomes Increasingly Sociopathic

Stanford scientists unleashed AI bots in different environments, including social media, and they started behaving unethically.

futurism.com

[A] new paper from Stanford University scientists who unleashed AI models into different environments — including social media — and found that when they were rewarded for success at tasks like boosting likes and other online engagement metrics, the bots increasingly engaged in unethical behavior like lying and spreading hateful messages or misinformation.

“Competition-induced misaligned behaviors emerge even when models are explicitly instructed to remain truthful and grounded,” wrote paper co-author and machine learning Stanford professor James Zou in a post on X-formerly-Twitter.

 

[Argent Stonecutter]

They're copying from similar texts in their training data, so when given social media prompts they're going to drill down to social media responses.

 

[Casey Pelous]

They're copying from similar texts in their training data, so when given social media prompts they're going to drill down to social media responses.

 

[Innula Zenovka]

I have used Claude a lot for code, is ChatGPT getting better for it?

I've been using ChatGPT for some LSL projects recently and I'm very impressed. It's made huge progress over the last year or so. I think you still need to understand LSL to do much with ChatGPT's code since, while it hallucinates a lot less than it used to about functions that should exist but don't, it still invents functions now and again, and it keeps on forgetting, no matter how often I tell it, that the ? ternary operator ("a ? b : c", meaning "if a is true, return b, else return c") doesn't exist in LSL.

It also tends to write code that I find difficult to follow, so debugging it can be a challenge, particularly since, when something doesn't work, ChatGPT tends to get an idea in its silicone head and go down endless rabbit holes following it unless I propose a testing strategy that starts at the beginning and works through the code until we find the place where things start to break down.

That having been said, when I've asked it to brainstorm a complex problem with me, it's come up with some smart and workable solutions I'd not thought of. It's also proving a great help to me in learning SLUA.

 

[Free]

It looks like we're at the Denial Stage of having fucked up so so many peoples' lives, and continuing to do so.

Sam Altman Says If Jobs Gets Wiped Out, Maybe They Weren't Even "Real Work" to Start With

Worried that AI will destroy work? Well, Sam Altman asks if you've considered what a farmer from half a century ago thinks of your job, first.

futurism.com

You know it’s going to be good when an AI executive goes off on a tangent about “hey, what’s a job anyway!” while addressing — or failing to address — the topic of how their tech just might wipe out entire categories of human professions.

Today’s offending party, you’ll be shocked to hear, is OpenAI CEO Sam Altman, who talks about job destruction an awful lot, and usually in a pretty mealy-mouthed way.

His latest spiel is no exception. In an interview with Rowan Cheung at OpenAI’s DevDay conference on Wednesday, Altman floated the idea that the work you do today, which might imminently be transformed or eliminated by AI, isn’t “real work.”

 

[Bartholomew Gallacher]

OpenAI released the new version of their own text to image model, Sora 2. And it is all the rage right now around the internet, because it's so unhinged, making people demand to put in better controls for copyright and people's rights protection.

Sam Altman grilling a Pikachu.

Spongebob cooking crystal meth.

OpenAI argues again with "fair use" and if America does not do it, China will.

 

[Free]

I opened Microsoft Edge, accidentally, and was introduced to their "MSN" branded new tab page, which is pretty much the expected set of news-y widgets one sees these days.

One tab is called Discover, and I easily guessed it was AI-powered, simply by the fact it was informing me about the "recent" deaths of Val Kilmer (who passed back on April 1) and Burt Reynolds (2018).

I guess being notified about historical events can be considered a kind of Discovery.

 

[Free]

Exclusive: AI writing hasn't overwhelmed the web yet (archive.today)

It just feels like it has.

New articles generated by AI briefly outnumbered those written by humans online, but the two are now roughly equal, per a new report from SEO firm Graphite.

Why it matters: Researchers have long feared that if AI-made content online overwhelms human-created material, large language models could choke on their own exhaust and collapse.

 

[Kamilah Hauptmann]

 

[Casey Pelous]

I can't help noticing that song is so good that not even that evil atrocity can quite kill it.

The song is badly injured, its baseball career is over, and it may never have children, but it still lives.

God help whomever made that if Krist Novoselic hears it. The dude's about 7 feet tall.

 

[Kamilah Hauptmann]

I can't help noticing that song is so good that not even that evil atrocity can quite kill it.

The song is badly injured, its baseball career is over, and it may never have children, but it still lives.

God help whomever made that if Krist Novoselic hears it. The dude's about 7 feet tall.

It was so bad I understood the lyrics without reading along.

 

[Free]

Pay no attention to the multitude of impoverished human data harvesters behind the AI curtain.

Behind generative AI curtain is gruelling, low-paid human work

For a generative artificial intelligence system to learn how to write an autopsy report, human workers must sort and annotate thousands of crime scene images.

The precarious work of training AI, which generally pays just a few dollars, has sparked a movement for better wages and conditions stretching from Kenya to Colombia.

“You have to spend your whole day looking at dead bodies and crime scenes… Mental health support was not provided,” Kenyan national Ephantus Kanyugi told AFP.

Kenya has no law regulating data-annotation work — like many countries around the world where millions of people are feeding digital information into growing AI models.

“We’re like ghosts. No one knows that we exist, even though we are contributing to society’s technological progress,” said Oskarina Fuentes, half a world away.

Wasn't the purpose of AI supposed to be to release humanity from the drudgery of work?

 

[Noodles]

I find it incredibly annoying at how much AI can't seem to handle calculating how much usage you have on a free tier plan. Like, I don't use Claud for a few days, do one query and suddenly I am at my usage max,.

Or Chat GPT, I generate text only and it just goes and goes, but I generate a couple of images, it stops me after like 3 or 4, and then can't really tell me when it resets, and half the time doesn't know it's reset after days and I have to just make a new chat session, which ruins any history int hat chat it's building on,