Eggplant.place

1y

This should be on TV every day, possibly multiple times per day…. #AI

1y

The Catholic Church has a new Pope, and he explicitly cited the threats AI presents for workers as a reason for choosing the name Leo XIV.

With the tech industry doing all it can to defend its power, could the Pope be an ally against the AI onslaught?

https://www.disconnect.blog/p/will-pope-leo-xiv-be-an-ally-against

#tech #pope #popeleo #ai #artificialintelligence #catholicchurch #vatican

0 0 1 View Post & Replies See Original

1y

終於放監了,最少twitter是

#月曜がんばれ 131
「がんばれ♥ がんばれ♥ がついてるよ！」

#松實優麗 #うちの子
 #stablediffusion #AI #AI繪圖 #AIイラスト #AIart #AIイラスト好きさんと繋がりたい

0 0 0 View Post & Replies See Original

1y

Dear #Abyss Screamer,

We're excited to advise that when your Abyss Screaming Subscription renews at the end of the month, you'll be automatically upgraded to the latest AI-enhanced abyss.

#AI enhancements to the Abyss include:

- Analysis of scream levels and volume with feedback of screaming techniques that you may wish to consider in future.

- Insights into other common screaming techniques.

- A summary of your scream.

Additionally, if you're feeling time constrained, you can ask the Abyss to scream into itself so that you can keep working.

We're certain you'll find our new AI Enhanced Abyss an exciting addition to our Abyss portfolio. Please keep in mind that your Abyss Subscription will increase by $15 per month to take advantage of these new features.

For concerns or comments, please scream into the abyss.

[The abyss is not monitored and you will not get a reply.]

Edited 1y ago

0 0 1 View Post & Replies See Original

1y

💻 📚 **How an AI-generated summer reading list got published in major newspapers**

“_Only five of the 15 titles on the list are real._”

🔗 https://www.npr.org/2025/05/20/nx-s1-5405022/fake-summer-reading-list-ai.

#AI #ArtificialIntelligence #Technology #Fake #ReadingList #Books #Bookstodon @ai @bookstodon

1 0 0 View Post & Replies See Original

1y

FFS: What an absolute bellend: "Duolingo CEO says AI is a better teacher than humans—but schools will still exist ‘because you still need childcare’"

Were I still using that POS Duolingo, this would have made me stop.

https://fortune.com/2025/05/20/duolingo-ai-teacher-schools-childcare/

#ArtificialIntelligence #AI #Duolingo

Edited 1y ago

0 0 1 View Post & Replies See Original

1y

On the limits of LLMs (Large Language models) and LRMs (Large Reasoning Models). The TL;DR: "Our findings reveal fundamental limitations in current models: despite sophisticated self-reflection mechanisms, these models fail to develop generalizable reasoning capabilities beyond certain complexity thresholds." Meaning: accuracy collapse.

Interesting paper from Apple. https://ml-site.cdn-apple.com/papers/the-illusion-of-thinking.pdf

#AI #LLM #LRM

• We question the current evaluation paradigm of LRMs on established math benchmarks and
design a controlled experimental testbed by leveraging algorithmic puzzle environments that enable
controllable experimentation with respect to problem complexity.
• We show that state-of-the-art LRMs (e.g., o3-mini, DeepSeek-R1, Claude-3.7-Sonnet-Thinking)
still fail to develop generalizable problem-solving capabilities, with accuracy ultimately collapsing
to zero beyond certain complexities across different environments.
• We find that there exists a scaling limit in the LRMs’ reasoning effort with respect to problem
complexity, evidenced by the counterintuitive decreasing trend in the thinking tokens after a
complexity point.
• We question the current evaluation paradigm based on final accuracy and extend our evaluation
to intermediate solutions of thinking traces with the help of deterministic puzzle simulators. Our
analysis reveals that as problem complexity increases, correct solutions systematically emerge at
later positions in thinking compared to incorrect ones, providing quantitative insights into the
self-correction mechanisms within LRMs.
• We uncover surprising limitations in LRMs’ ability to perform exact computation, including their
failure to benefit from explicit algorithms and their inconsistent reasoning across puzzle types.

ALT

Edited 1y ago

0 0 1 View Post & Replies See Original

267d

idle idea about submissions & publishing in LLM times (long!)

I was struck by this passage in an article about the harms of so-called #AI in academia, literature, and other areas (found here https://theconversation.com/ai-generated-text-is-overwhelming-institutions-setting-off-a-no-win-arms-race-with-ai-detectors-274720 while noting I strongly disagree with the author calling LLMs a democratization of access, and I also note the article has other issues like confusing LLMs and machine-learning technology in general):

"[Fiction outlets] may refuse AI-generated work, but this will come at a cost. It’s unlikely that any human editor or technology can sustain an ability to differentiate human from machine writing."

This assumes that fiction submission and #publishing will remain fixed in response to a changing technological and social landscape, passively taking all submissions while launching largely futile countermeasures against LLM-generated entries.

But what if segments of publishing were to transform themselves so that it would CEASE TO MATTER whether entries are generated, while still having a high chance of winnowing out slop without overwhelming editors or forcing them to turn to LLM-driven non-solutions?

One way might be to hold semi-open pitching/brainstorming sessions for magazine issues or anthologies, where, say, the first 50 people in the door are briefly interviewed or submit pitches for what kind of stories they have written or can write for the issue/anthology, while freely discussing with the editor(s) what contribution they can make and how their entry fits into the overall picture of the collection. In this process the fiction issue or anthology can start taking shape through the ideas generated during these discussions and pitches, and the candidates who show a good understanding of their own works/ideas and the collection could be invited to submit their works. If these won't be enough submissions the next batch of e.g. 50, maybe fewer, are invited to pitch, and some of them to submit. Some applicants might be left in reserve in case some of the actual submissions are duds or slop.

In this way writers who wish to submit to an issue or anthology can be pre-screened before submission while having an active voice in editorial decisions and helping set the overall direction of a collection. The process of pitching and discussion itself can filter out empty, banal, or badly-fitting ideas. If LLM-generated pitches and responses can actually pass that process, more power to them, but given the characteristics of LLM-generated text it won't be terribly likely or widespread as long as the editors themselves have original and specific taste and vision. And limiting the number or pool of participants (new authors only, published authors only, members of marginalized communities only, etc.) will prevent editors from being overwhelmed or burning out, while still leaving room for discovering new contributors.

Will this kind of process, just one of many possible adaptations, change the nature of the submission and publishing process? Undoubtedly. Under the particular method outlined above, the submission process will become more like flash events and extended editorial/brainstorming sessions. It could work out great for some outlets while being terrible for others. Still, change is kind of the point and it's clear publishing can't stand static and defensive in the face of a sea change. I hope to see a lot of new initiatives and adaptations going forward.

1 0 0 View Post & Replies See Original

202d

#AI
#aislop

1 0 0 View Post & Replies See Original

1y

Yesterday, during training (customer service position), one of my coworkers asked if they should be using #LLMs to answer customer queries, and my blood ran cold.

Before anyone could respond, I jumped in and said, "No, absolutely not. Generative #AI doesn't give correct answers. It gives responses which look superficially correct, but are factually wrong. And when we give the customer the wrong answer, they don't want us to blame the machine. The blame lands squarely on our shoulders. Do not use ChatGPT, don't use Google AI, don't use Bing's whatever, just DON'T."

Thankfully, everyone thinks I know what I'm doing, so they accepted that. But this is such a common question that nobody thought it odd to ask, and that makes me worried in a very different fashion.

Edited 1y ago

0 0 1 View Post & Replies See Original

1y

Boost appreciated! ♻️

Ok I've come up with an idea;
I'm not going to post anything more AI related, I'm going to sell that content to help me with life and be able to afford real art commissions

So far I've generated about 227 images, I'm going to sell them as packages, you can see the pricing in the pictures below

I might also open requests ( featuring my own characters that is ) for about 5€? But I'll have to think about how that works

@Fidda for more info!

#AI #loli

Edited 1y ago

0 0 0 View Post & Replies See Original

1y

我認我縮,這星期暫且穿褲子…

#月曜がんばれ 130
「全力全開！パワー全開！ファイトーッ！！」

#松實優麗 #うちの子
 #stablediffusion #AI #AI繪圖 #AIイラスト #AIart #AIイラスト好きさんと繋がりたい

0 0 0 View Post & Replies See Original

1y

This is obscene.

In analyzing both public and proprietary data about data centers as a whole, as well as the specific needs of #AI, the researchers came to a clear conclusion. Data centers in the US used somewhere around 200 terawatt-hours of electricity in 2024, roughly what it takes to power Thailand for a year. AI-specific servers in these data centers are estimated to have used between 53 and 76 terawatt-hours of electricity. On the high end, this is enough to power more than 7.2 million US homes for a year.

https://www.technologyreview.com/2025/05/20/1116327/ai-energy-usage-climate-footprint-big-tech/

0 0 1 View Post & Replies See Original

1y

We obtained records showing how a DOGE staffer with no medical experience used AI to identify which VA contracts to kill.

His tool concluded that more than a thousand deals were each worth $34 million, when in fact some were for as little as $35,000.
https://www.propublica.org/article/trump-doge-veterans-affairs-ai-contracts-health-care?utm_source=mastodon&utm_medium=social&utm_campaign=mastodon-post

#News #AI #DOGE #Veterans #Health #Healthcare #Government

0 0 1 View Post & Replies See Original

1y

Last weekend, Elon Musk shared some images on his X platform that he claimed had been created by the Grok AI system. They were actually just edits of original work by Australian artist Charles Brooks, a former concert cellist who is known for photographing the insides of instruments. An X user named Eric Jiang had run some of these through Grok, adding people to the original photos so they look like buildings. Musk retweeted the post with the comment "Generate images with @Grok." @404mediaco's Matthew Gault spoke with Brooks. “I feel like, if he’s plugging his own AI software, he has a duty of care to make sure that what he’s posting is actually attributed correctly and is properly his,” he said. “But ‘duty of care’ and Musk are not words that seem to go together well recently.”

https://flip.it/OQhrJ3

#Art #AIArt #GenerativeAI #Grok #AI #ArtificialIntelligence #CharlesBrooks #ElonMusk

0 0 1 View Post & Replies See Original

202d

I got an email from my bank with the subject line:

"Protect your business from AI-driven fraud."

So yeah, another win for AI, really improving the lives of... scammers?

#AI #NoAI #FuckAI

1 0 0 View Post & Replies See Original

162d

The future for security professionals is incredibly bright.

I am seeing change after change of #ai generated code that is using hard-coded SQL string formatting rather than binding query parameters.

0 0 0 View Post & Replies See Original

1y

According to a new analysis, AI isn't even close to coming for everybody's jobs, and it's not exactly making workers more productive, either. Via @Futurism. #AI #Chatbot #ArtificialIntelligence #Tech #Technology https://flip.it/DO7dx2

0 0 1 View Post & Replies See Original

1y

Browsers are still on the news and this time they're caught in the AI wars: https://world.hey.com/ricardo.tavares/ai-browsers-less-local-more-clouds-e19c5386 How does that help or threaten the web?

#webdev #ai #browser #firefox #ladybird #chrome #safari

0 0 0 View Post & Replies See Original

1y

“If you can’t use a billion dollar #AI system to solve a problem that Herb Simon … solved with AI in 1957, and that first semester AI students solve routinely, the chances that models like Claude or o3 are going to reach AGI seem truly remote.”

https://garymarcus.substack.com/p/a-knockout-blow-for-llms

0 0 1 View Post & Replies See Original

1y

Disillusioned & distanced from #Trump, #Musk says he’s exiting Washington & made clear he is frustrated w/obstacles he encountered as he tried to upend the #FederalGovernment🎻

Musk dissed Trump’s signature domestic policy, saying it would add to the #deficit. He complained about a lucrative deal that went to a rival co to build an #AI #DataCenter in the #MiddleEast. And he has yet to make good on a $100M pledge to Trump’s political operation.

#BooMutherFuckinHoo #USpol
https://www.nytimes.com/2025/05/28/us/politics/elon-musk-trump-doge.html?smid=nytcore-ios-share&referringSource=articleShare&sgrp=p&pvid=94709C91-3556-4635-8095-65DC81177EF3

0 0 1 View Post & Replies See Original

1y

#月曜がんばれ 136
「できるできるっ！君なら絶対できる！がんばれ♥」

#松實優麗 #うちの子
 #stablediffusion #AI #AI繪圖 #AIイラスト #AIart #AIイラスト好きさんと繋がりたい

0 0 0 View Post & Replies See Original

112d

A certain #Kdrama used #AI (they also publicly disclosed it, which is required in Korean Law) but I have yet to see a single complaint.

What's interesting is, the Western (AU, EU, NA) world is silent about it even though it's topping their charts. (The West has the highest number of anti-AI.)

1 0 0 View Post & Replies See Original

1y

More, faster, and wrong - why #AI won't solve your software productivity problems. 🤖 https://matthewreinbold.com/2025/06/19/theproductivitytrap

0 0 0 View Post & Replies See Original

1y

Hallucinating #GoldStandard science

"The #MAHA Report Cites Studies That Don’t Exist."
https://www.notus.org/health-science/make-america-healthy-again-report-citation-errors

"Health Secretary #RFKJr says his “Make America Healthy Again” Commission report harnesses “gold-standard” science, citing more than 500 studies and other sources to back up its claims. Those citations, though, are rife with errors, from broken links to misstated conclusions. Seven of the cited sources don’t appear to exist at all."

#AI #DefendResearch #Medicine #ScholComm #Trump #TrumpVResearch #USPol #USPolitics

2 0 0 View Post & Replies See Original