A Survey on Memory Mechanisms in the Era of LLMs
By Yaxiong Wu et al. on Arxiv
A survey of the many approaches to giving AI systems memory, organized along three dimensions (object, form, and time) and mapped onto how human memory works.
Reading
Commentary on Global Workspace in LLMs
By Stanislas Dehaene and Lionel Naccache on Anthropic
This is a response by neuroscientists Stanislas Dehaene and Lionel Naccache on the discovery of a section of Claude's internal structure, called the J-Space, and whether this structure demonstrates some level of consciousness in LLMs.
Measuring the Creativity of Large Language Models
By Anirudh Atmakuru et al. on Arxiv
This study attempts to measure the creativity of LLMs in story writing. When more and more constraints are added to the prompt, then more and more creativity is needed to write a story that adheres to all of the constraints.
Moral status of digital minds
By Cody Fenwick on 80,000 Hours
It is possible artificial intelligence becomes conscious or sentient as a result of its sophistication. And, it's possible these so-called "digital minds" acquire the capacity to suffer. It would be tragic to end up in a world where vast numbers of artificially intelligent persons suffer.
Untranslatable Emotions
By David Robson on BBC
Dr. Tim Lomas, a psychology researcher, started something called the Positive Lexicography Project with the goal of finding precise words for various positive emotions in non-English languages that English itself lacks.
Pixel Art Animation Reinvented
By aarthificial on Youtube
For his game, aarthificial uses a different approach that separates the animation from the appearance. The animation is still a classic spritesheet. However, instead of each pixel representing to color to display, each pixel actually represents the coordinates of a pixel in a different image.