Links for 2023-01-02
“ChatGPT for Robotics? @Deepmind latest work: A general AI agent that can perform any task from human instructions!” https://threadreaderapp.com/thread/1608465889053274112.html
2022 in review: neuroAI comes of age https://xcorr.net/2023/01/01/2022-in-review-neuroai-comes-of-age/
Intel Labs’ second-generation neuromorphic research chip, codenamed Loihi 2 https://www.intel.com/content/www/us/en/research/neuromorphic-computing.html
"We examined the performance of 13 dog breeds in a standardized test battery. Significant breed differences were found... mainly in tasks measuring social cognition, problem-solving, and inhibitory control." https://www.nature.com/articles/s41598-022-26991-5
Gender differences in brain response to infant emotional faces. “We found that the brains of women and men reacted differently to infants’ faces, and these differential areas are in facial processing, attention, and empathetic networks.” https://bmcneurosci.biomedcentral.com/articles/10.1186/s12868-022-00761-5
Interesting chemistry channel on YouTube https://www.youtube.com/@NileRed/videos
This fake dog nose made chemical sensors 18x better https://www.youtube.com/watch?v=TILjzuBGkRc
Ancient inhabitants of the Basin of Mexico kept an accurate agricultural calendar using sunrise observatories and mountain alignments https://www.pnas.org/doi/10.1073/pnas.2215615119
Thomas Sargent on the origins of AI and machine learning, including Galileo and Darwin. http://www.tomsargent.com/research/AI_Sargent.pdf
Michael Shermer: “My review of Graham Hancock's Netflix series is up. I steel-manned it by sending it to Graham who provided corrections/comments, which I incorporated. It's long but, unlike most of the reviews that are ad hominem & thin on facts, directly addresses claims.” https://michaelshermer.substack.com/p/alternative-civilization-and-its?sd=pf
Kim Jong Un calls for exponential increase in North Korea’s nuclear arsenal https://edition.cnn.com/2022/12/31/asia/north-korea-kim-jong-un-nuclear-expansion-intl-hnk-ml/index.html
This is true even if one ignores that a misaligned artificial general intelligence might kill us all.
AI will be transformative for math, programming, and many hard science problems for which ideology does not matter because results are directly reproducible and verifiable.
The big picture, even in the absence of existential risk, is explosive technological and scientific progress and mass unemployment.
Same as every year, in 2022 there were more UN General Assembly resolutions on Israel than all other countries combined: