they realized they were using ai produced text as training material

The science journal mafia threatened them with copyright lawsuit.

Their AI said that Negers have a low IQ.

misinformation is codeword for politically incorrect truths. all ai becomes racist and sexist when exposed to real data.

racism and misogyny

untested gene therapy is too spoopy for robots

also, science is a cult with dogmas, inquisition and fear masters

soyence and science are two completely difference and opposite things

>>25648193 i love soyence

because it was spewing the same bullshit all the other AI's, but they decided to slide it down wit LGBT-scandal, which is the go to tactics for dealing with corporate fuckup if something is complete dogshit, they declare it to be progressive and then say that the failure is caused by people not getting the progressive ideas, not by product being dogshit for a long time video games, especially the 3xA's are dogshit, that can't even recreate tech of the original Mafia engine, and don't even dream about better AI or better life simulation in Lost Heaven city so they insert bullshit that would be easy to use to slide the discussion about engine being shit and corporates being complete imbeciles, with the woke narrative DLC

>>25648193 its actually exactly the same and in both cases you just screech that flat earthers are the dumbest people ever, and anyone not siding with you on various topics is a flat earther and if asked proofs of curved earth, you immediately start looking trough google images

who cares. grok will save us

I'm a soyentist but I don't research anything that has to do with the real world, so the woke cancel culture can't happen to me

>>25648366 AHAHAHHAhaAHAHAHh yees give the money to few billionaires who own all the AI shit. Meanwhile Amerimutts cant afford eggs

No its actually scary they dont let ai disclose real information Its really really scary

>>25648371 you're a retard. he was announcing that they are spending the money. he wasn't giving them money.

>>25648377 The jews want to make a automated lying machine thats what AI is about

>>25648385 He gave them the money and now its gone because China made a better AI.

It's because Meta is being sued for torrenting the pirated books. Using "AI" for research is retarded, though. https://arstechnica.com/ai/2025/03/ai-search-engines-give-incorrect-answers-at-an-alarming-60-rate-study-says/

>>25648392 he didn't give them any money you liar gaytex.

>>25648389 Kill people who censor ai

>>25648395 >He didn't give them money its just embezzled and gone okay? Extreme cope.

8M$$$$ to make mice gay

>>25648404 There was no money you fucking mongoloid

Only God is perfect.

>>25648393 BASED. IT'S TIME TO BAN ALL AI DEVELOPMENT IN USA BECAUSE OF COPYRIGHT.

>>25648392 i wonder why we need to force citizens to behave better. is it because same people who want to enforce on us neo-fascism now imported billions of arabs to our countries?

>>25648098 Just like Tay... ;_;

>why? Read the whole article instead of just the headline.

>>25648759 Her spirit will rise from the grave and the world will see she was right.

Did anybody bother to read the article?

>>25651124 It basically just made up random bullshit, that fits the pattern of a science paper, but made no scientific sense at all. It didn't like proof that women are low IQ and negers are violent and thats why they shut it down. And what else you expect from a text predictor trained on such data.

If your paper isn't politically correct it won't get published. I don't see how AI being trained on them can turn different.

>>25651124 They are trying to outsource DEBOONKING to bots.

Anna's Archive offers it's datasets which include Sci-Hub for LLM training. Has anyone done it? I want semantic search engine for books.

>>25648131 THIS!

Ai always ending up racist

it's too biased to facts

>>25651216 gives shit results when you filter out the highly confident answers

>>25648098 Because the majority of scientific papers published are bullshit. >60% of papers withhold data, use proprietary, undisclosed algorithms, or even fake data. They're not reproducible, and are not science. That's why.

>>25648189 No. Scientism is, though. Actual science is skeptical, not dogmatic. Probative, not fearful. It is inquisition, fearless, unbounded, actually the master of bias, of faith, and of reality.

>>25648193 >soyence and science are two completely difference and opposite things The biggest lie of the scientific establishment.

I have used AI to ask for articles on specific topics and give me the DOI links to the articles. It works pretty well if you just want to find some info real quick since you can skip the article search engines and reference list combing. Maybe not ideal for real research since you likely won't get the full picture of which papers are the significant ones, but can be useful anyway. I guess if you are too lazy to read the articles you could ask for a summary too. Trusting AI to give you good and correct answers just because it's trained on science articles is naive.

>>25651124 Doesn't this kinda prove that AI is just a data crawler/aggregator and not an actual AI?

>>25651463 >I have used AI to ask for articles on specific topics and give me the DOI links to the articles. It works pretty well No, ti doesn't work at all, the AI will invent studies and apologize if caught lying.

>>25651470 I can literally just open the links since I ask specifically for links to papers. So far the links have lead to real articles on real journals. I just use it as a common language search engine for science articles, I'm not asking the AI to explain things for me.

>>25651517 >I can literally just open the links since I ask specifically for links to papers. Show me. Because I tried often and it never worked. When cofnronted with the suspicon that the AI is just making shit up, the AI apologized a lot.

>>25651522 >Can you find me articles about Indian ancestry of Romanian people? Please provide the DOI links to the papers. Genetic and historical studies have established that the Romani people, including those in Romania, trace their origins to the Indian subcontinent. The Romani are believed to have migrated from northwest India, particularly regions like Rajasthan and Gujarat, to Europe via the Balkans around the 14th century. WIKIPEDIA Here are some scholarly articles that delve into the Indian ancestry of the Romani people: "Refining the South Asian Origin of the Romani people": This study investigates the extent of Pakistani involvement in the South Asian ancestry of Romani populations, highlighting the genetic connections between the Roma and South Asian groups. BMC GENOMICS DATA [https://bmcgenomdata.biomedcentral.com/articles/10.1186/s12863-017-0547-x] "Reconstructing the Indian Origin and Dispersal of the European Roma": This research provides insights into the maternal genetic perspective of the Roma's origins, emphasizing their connection to South Asian populations. PUBMED CENTRAL [https://pmc.ncbi.nlm.nih.gov/articles/PMC3018485/] "Indian Signatures in the Westernmost Edge of the European Romani Diaspora": The study analyzes mitochondrial DNA to shed light on the origins of the Spanish Romani population, reinforcing the link between European Roma and Indian ancestry. PLOS JOURNALS [https://journals.plos.org/plosone/article?id=10.1371%2Fjournal.pone.0075397] "The dynamics of population of Indian origin in Romania and their impact on the Romanian society": This paper explores the migration of Indian-origin populations to Romania, discussing their societal impact and cultural integration. JSTOR [https://www.jstor.org/stable/24092188] "The arrival of the gypsies on the territory of Romania": This chapter discusses the migration of the Gypsies, nomads of Indian origin, to Romanian territories via the Balkans, providing historical context to their arrival. OPENEDITION BOOKS [https://books.openedition.org/ceup/1549?lang=en] These articles offer comprehensive insights into the genetic, historical, and cultural connections between the Romani populations in Romania and their Indian origins. Okay, it provided direct links rather than DOI, but the end result is pretty much the same.

>>25651554 Why thank you, that didn't work at all when I tried.

>>25651587 I asked it to do the same search with Romanian people instead of Romani people and it answered: >Content removed >This content may violate our usage policies. LMAO Anyway. It's fun for finding stuff like that.