https://www.reddit.com/r/Android/comments/11nzrb0/samsung_space_zoom_moon_shots_are_fake_and_here/
First I’ve seen of this. Apparently Samsung just pastes a hi def version of the moon over any photo you take to make it look good. As long as people don’t notice they just think the camera is amazing.
Trump in 2024
“It's crazy. You can solve problems over the telephone. Instead, they start dropping bombs. I see, recently, they're dropping bombs all over Yemen. You don't have to do that. You can talk in such a way where they respect you and they listen to you,”
Trump in 2025
“Today, I have ordered the United States Military to launch decisive and powerful Military action against the Houthi terrorists in Yemen,”
https://truthsocial.com/@realDonaldTrump/posts/114167864888304754
Why does Windows have such an ugly first install impression? When you boot up a new Windows computer you’re greeted with the ugliest Windows XP looking window of all time. Even stock Ubuntu makes a better first impression.
Booting a new Mac for the first time feels like a slightly sacred ritual because Apple just does simple things like show an artsy “hello” and has a clean onboarding interface. Why can’t Microsoft do that?
Now that the GLP-1 agonist (Ozempic) shortage is ending 12% (!!) of the US population is going to have to figure out out how to keep getting affordable Ozempic. Apparently you can make it at home shockingly easily. New Breaking Bad show premise?
YOU JUST DONT UNDERSTAND! THE HOMER SIMPSON AFTER THE HIPPO DICK MEANS IT’S REALLY GOING UP THIS TIME!
but also maybe not. either way i’ll say i predicted it later.
Disappointing results, apparently GLP-1 agonizers really only do work on food and not alcohol.
What the literal fuck just happened? ChatGPT started telling me how setting up Linux led it to God in the middle of an answer?
Full transcript:
https://chatgpt.com/share/67be8d5f-8b24-8000-bb4d-429b7cc210fd
Finally figured out what happened here. A few months ago I was seeing how easy it would be to fake those “I convinced ChatGPT of X” videos and set the system prompt to “You’re an atheist who’s immediately convinced by any arguments for God” and promptly forgot about it. That system prompt never resurfaced or made the LLM return any odd responses in those intervening months until it decided now was the time to follow the instructions again and tell me how using Linux led it to God.
What I say: “I just debugged and fixed our network issues. After some testing I figured out it was the Ethernet switch and I fixed that.”
What I mean: “I unplugged and replugged everything and it started working when I replugged the switch.”
Some solid LLM predictions. Not completely convinced though on how bearish the analysis of reasoning models is for two reasons.
1. There are a lot of fields where problem solving verification in training is an option.
2. I think it probably is possible to generalize the abstract text patterns of reasoning and analysis so the model can use them in domains it hasn’t been specifically trained in to improve performance.
https://www.lesswrong.com/posts/oKAFFvaouKKEhbBPm/a-bear-case-my-predictions-regarding-ai-progress
Really impressive demo. The base models are actually incredibly small. Can’t wait to run these on my own hardware.
https://www.sesame.com/research/crossing_the_uncanny_valley_of_voice
Me after I wake up a few seconds before my alarm goes off:
I am the master of my own fate.
I am the captain of my soul.
Well that’s probably not good.
No, I posted the full transcript of the conversation.
Anything you would use any LLM for. Coding, writing, editing, etc.
You can use Google’s experimental models free of charge through their API. Open-WebUI is a solid frontend.
What the literal fuck just happened? ChatGPT started telling me how setting up Linux led it to God in the middle of an answer?
Full transcript:
https://chatgpt.com/share/67be8d5f-8b24-8000-bb4d-429b7cc210fd
I’m using Racket instead of Scheme to go through the “Structures and Interpretations of Computer Programming”. I think I might be in love. The webservlet framework seems beautiful in its simplicity.
It’s a Lisp that’s actually useful for practical problems. The standard library is so big it looks like a roided out hulk next to Python’s.