@realMaalouf We need a new map with the most recent invasion of Western Europe of the last 50 years.
English
S
12.5K posts
S
@NakramR
Software engineer, Machine Learning, AI, ... Snark and sarcasm. Good faith arguments. Always include the word "salient" in your messages to me.
Katılım Haziran 2011
345 Takip Edilen320 Takipçiler
How did two billion Muslims, who had some of the biggest empires in history and wiped out entire civilizations, manage to convince the world that they are a persecuted minority and victims of ‘colonization’?
English
Azure customers hosting OpenAI models should now be seeing a 10x improvement in latency and throughput
You're welcome :)
English
S retweetledi
Do we know what caused this increase in rapes?
Geiger Capital@Geiger_Capital
New study from EuroStat… Rapes in the EU have surged by +150% over the last decade, more than doubling. The experts and feminists are baffled.
English
@ryancduff @taylorburrowes Even under the old rules, while permissible, it would have been expected from the guy to pursue other interests if he had any self worth.
English
@taylorburrowes She seems to be romanticizing a culture that hasn’t existed for a decade or more.
Unfortunately when society tears up the contracts and establishes new rules, you can’t keep playing by the old rules and expect others to go along with it.
English
@cremieuxrecueil Same as everything else that relies on authenticity of data: identity cards, passports, lease documents, contracts.
Web of trust, in one form of another, is going to become absolutely necessary for everything in the near future. From contracts to the press to science, etc.
English
Science is about to get absolutely nuked.
Unless we get extremely strict about providing and opening up code and data and documenting lab experiments rigorously, a torrent of credible-looking but fraudulent papers is upon us.
Ruslan Rust@rust_ruslan
Can water intake prevent Alzheimer’s disease? No. This is fully AI-generated… but the data below could easily pass as real. The new ChatGPT image model is truly impressive, but I think it poses a real risk for scientific integrity in future. For example, I could just generate a dataset with a single prompt that appears to show something like water preventing Alzheimer’s disease. Ironically, we used to laugh at obvious “AI slop” (like those weird generated mice), but that’s changing pretty fast. If I were reviewing this fake figure today, I’m not sure I could reliably tell whether this figure is real or AI-generated? The bigger issue is that the usual signals we rely on e.g., how realistic or plausible something looks are no longer enough. I think we really need more comprehensive AI detection and, more importantly, stronger verification standards for scientific submissions going forward. We’ll probably also need better ways to digitize lab notebooks and ensure access to raw data, something closer to how code and version history are tracked...
English
@RealAstropulse What happens if you feed it as input an image that is its default/max resolution output (1024x1024) and pre-fill everything with content already that it can't touch, except the 256x256 part that it has to fill?
Then it has no other choice but to be pixel perfect, right?
English
Retro Diffusion Pro vs. GPT Image 2.0
Same pipeline, same prompt, same size target (256x256)
Look at how insanely noisy the gpt 2.0 image is, and how the outlines and details don't maintain a consistent size. I even passed it through my own pixel art de-noising filter.
English
@jsors2020 @charliesmirkley You still don't understand my point.
The data doesn't support the argument.
The conclusion is one I agree with, but you need go data to support that conclusion. This isn't it.
Bye.
English
@NakramR @charliesmirkley Good for you.
Again, in this thread you played a different game until called out.
You responded to the OP dishonestly.
Bye
English
@jsors2020 @charliesmirkley It's quite funny because the last thing I posted (or reposted): "A teaching credential has no impact on student achievement. "
English
@NakramR @charliesmirkley Also, I do love how you're clearly fine with people lying about the intelligence of home school teachers. Your silence says it all.
English
@jsors2020 @charliesmirkley It's for the one with the claim to make their case, not for the one who questions to give contrary evidence.
You claim homeschool is better, prove it.
I make no claim at all other than saying "what you show doesn't prove what you think it does". Did you go to public school? ;-)
English
@NakramR @charliesmirkley Prove it. I know you love stats.
Demonstrate that the kids who were home schooled the last 20 years would fare the same or better in public schools.
English
@jsors2020 @charliesmirkley I think the data is that on average it doesn't matter, and in some cases homeschooled fare better.
And there's many arguments in favour of homeschooling, this statistics isn't one. Plenty of others to chose from though.
English
@NakramR @charliesmirkley So, what you're really saying is the anti home school crowd lies about home school teachers .
Whew, finally I understand. My apologies.
English
@jsors2020 @charliesmirkley I don't disagree with the intent or the fact that people should be free to chose how they school their children, but this statistics is incorrect if you want to decide whether to homeschool or not.
English
@NakramR @charliesmirkley Is that it?
What a lazy tired excuse.
Anyways, let families choose and take THEIR funding with them "for the kids".
English
@davidbessis You're confusing citizenship with culture.
You can legally have a french passport, but if you're eat primarily curry, worship hindu gods and dress in robes, you're not culturally french.
They used to be one and the same, but French culture is what's being talked about.
English
What is he talking about? My grandparents literally came to France and became French citizens.
Vivek Ramaswamy@VivekGRamaswamy
You can travel to Italy, but you’ll never be an Italian. You can travel to France but you’ll never be a Frenchman. You can live in Germany but you’ll never be a German. You can pack your bags and live the rest of your life in China or Japan, but you’ll never be Chinese or Japanese. Yet you can come from any one of those countries to the United States of America, and you can still be an American – so long as you work hard, you play by the rules, you make your contributions, wait your turn, pledge allegiance to the flag, and obtain your citizenship.
English
S retweetledi
A teaching credential has no impact on student achievement.
English
English
@bradwmorris @jeffreyhuber @OpenAI Sorry, this wasn’t intentional, text-embedding-3-small isn’t going away!
x.com/romainhuet/sta…
Romain Huet@romainhuet
Thank you for flagging this, Jeff. This was a mistake: we are not deprecating text-embedding-3-small. We’re looking into where this came from now, and we’ll also email users to clarify. Sorry for the confusion!
English
OpenAI is shutting down text-embedding-3-small?!?
I strongly believe that if you shut down a closed-source embedding model that you should open-source. Imaging the trillions of tokens that will no longer be queryable.
cc @romainhuet
English
@EpistemicHope @RokoMijic @bryan_caplan Following someone doesn't mean agreeing with all they have to say.
In this case, he hasn't argued as to why this suggestion is a bad one. Selective immigration is a good idea. The question is what you're selecting for. Criminality, intelligence, genetic health are all good imo.
English
@RokoMijic @bryan_caplan Your attitude to the first statement is very atypical for followers of Bryan Caplan. Most people reading this feel very negatively about it. For them, this whole tweet is therefore a pro-immigration argument.
That's what you're missing.
English
Think how you react to arguments of the form "Black Americans are worse than average on measure X, so we should ship them back to Africa."
That's how you should react to arguments of the form "Immigrants are worse than average on measure X, so we should deport them."
English
S retweetledi
S retweetledi
New paper! LLMs Corrupt Your Documents When You Delegate
LLMs are enabling a new way of working: delegated work, where users supervise an LLM as it edits documents on their behalf.
Delegation requires trust: does the LLM complete tasks without introducing errors?
We simulate delegation across 52 professional domains and find that LLMs Corrupt Your Documents When You Delegate. 🧵1/N
English