“Announcing GPT-4, a large multimodal model, with our best-ever results on capabilities and alignment.

“So I like the idea of being stingy with our goals for the new year. Make that New Year’s resolution to listen to fewer podcasts, to travel less, to never go camping. Pick things you explicitly won’t do. Your time is finite. A new year is not the time to pursue infinite goals; it’s a time to pick the important ones. And not everything can be important.

According to a 2021 study of 60 boxing and MMA fights, 78% of the fights led to a concussion, and the fighter that sustained the first concussion ultimately lost 98% of the time.

Seems like strike-based martial arts amount to "last one to get a concussion wins".

No good evidence that tailoring teaching to students’ preferred learning styles has any effect on objective measures of attainment. The effect of “nudges” (clever design of defaults) may be exaggerated in general. No good evidence of anything from the Stanford prison ‘experiment’. No good evidence from the famous Milgram experiments that 65% of people will inflict pain if ordered to. Lots of screen-time is not strongly associated with low wellbeing; it explains about as much of teen sadness as eating potatoes, 0.35%. The Pygmalion effect, that a teacher’s expectations about a student affects their performance, is at most small, temporary, and inconsistent, r<0.1 with a reset after weeks. At most weak evidence for stereotype threat suppressing girls’ maths scores. i.e. the interaction between gender and stereotyping. Good evidence against ego depletion, that willpower is limited in a muscle-like fashion. Questionable evidence for the “hungry judge” effect, of massively reduced acquittals (d=2) just before lunch. Questionable evidence for a tiny “depressive realism” effect, of increased predictive accuracy or decreased cognitive bias among the clinically depressed. No good evidence for multiple intelligences (in the sense of statistically independent components of cognition). Mixed evidence for a small marshmallow effect. No good evidence that brains contain one mind per hemisphere. At most extremely weak evidence that psychiatric hospitals (of the 1970s) could not detect sane patients in the absence of deception. Questionable evidence for the menstrual cycle version of the dual-mating-strategy hypothesis. No good evidence that humans are cognitively or emotionally limited to 150 relationships with other people (Dunbar’s number).

“The mixture of cold fresh water from the glaciers and warmer seawater drives ocean circulation near the glaciers and farther out in the ocean, meaning the melt has far-reaching implications. The Greenland ice sheet is also important for sea-level rise; Greenland ice holds enough water to raise sea levels by 20 feet (6 meters).

“We’ve trained a model called ChatGPT which interacts in a conversational way. The dialogue format makes it possible for ChatGPT to answer followup questions, admit its mistakes, challenge incorrect premises, and reject inappropriate requests.

Among the sample conversations included are topics like "help me write a short note to introduce myself to my neighbor", "this code is not working like i expect — how do i fix it?", "what is fermat’s little theorem?" and "how to protect my home from burglary".

“The Mystery Flesh Pit is an extremely elaborate fantasy worldbuilding exercise created by writer and illustrator Trevor Roberts.

Imagine a Permian Basin-sized eldritch organism hibernating under the soil of Texas just off of Abilene, that will end the world should it ever awake. And instead of having the discovery of such a thing drive America insane in a bout of cosmic horror, we decided to turn it into a National Park Service-operated tourist trap in the vein of the Carlsbad Caverns. So we wired its jaw open, stuck an elevator and a Burger King franchise down its gullet, and sold guided tours and camping trips into the living interior of the beastie. Oh, there's also an all-inclusive resort by its digestive system and an aphrodisiac spa in its amniotic sacs.

The joke, as it were, is this really incongruous mashup of Lovecraftian horror subject matter, but described in the breathlessly enthusiastic tone of a tourist brochure, and all illustrated by Robert's incredibly realistic parody of the NPS's informational panel art.

 I just learned that these creatures exist.

“eDiffi is a new generation of generative AI content creation tool that offers unprecedented text-to-image synthesis with instant style transfer and intuitive painting with words capabilities.

“In 1981, Gilbert and Jaime Hernandez wrote "Love and Rockets #1," a self-published comic book edited by their brother Mario. They sold that first issue at conventions for a dollar apiece and submitted it to be reviewed at The Comics Journal. Instead, Gary Groth, offered to republish it through Fantagaphics Books. The brothers accepted and made graphic novel publishing history.

“In our paper, published today in Nature, we introduce AlphaTensor, the first artificial intelligence (AI) system for discovering novel, efficient, and provably correct algorithms for fundamental tasks such as matrix multiplication.

Our system, AlphaTensor, builds upon AlphaZero, an agent that has shown superhuman performance on board games, like chess, Go and shogi, and this work shows the journey of AlphaZero from playing games to tackling unsolved mathematical problems for the first time.

“This website uses artificial intelligence to automate this task by translating back and forth between English and RegEx

It turns out that not only humans are interested in Stray, the new game where you play as a cat.

“Rivalry between stars of a unique accordion-based style of music in the southern African kingdom of Lesotho has sparked years of deadly gang warfare that has turned the tiny country into the murder capital of the continent. 

Famo artists and their fans, both in South Africa and Lesotho, have divided into warring groups, distinguished by wearing traditional blankets in a particular colour. "When artists start making a name for themselves, they record songs that are full of insults."

“OpenAI’s GPT-3 and other neural nets can now write original prose with mind-boggling fluency — a development that could have profound implications for the future.

DALL·E 2 is a new AI system that can create realistic images and art from a description in natural language.

(The above image was generated in a couple of seconds from the prompt "A photo of a robot hand drawing, digital art".)