New neurotech eschews electricity for ultrasound(spectrum.ieee.org)

61 points by bookofjoe 2 years ago | 14 comments

TSMC bets on unorthodox optical tech(spectrum.ieee.org)

211 points by Rohitcss 1 year ago | 84 comments

What AMD Learned from Its Big Chiplet Push(spectrum.ieee.org)

78 points by mfiguiere 3 years ago | 46 comments

Human Brain Cells on Chip for Sale – First biocomputing platform hits the market(spectrum.ieee.org)

44 points by mdp2021 1 year ago | 22 comments

Germany’s Infineon opens major chip plant as EU seeks tech autonomy(rfi.fr)

169 points by giuliomagnifico 3 months ago | 49 comments

Cyborg Embryos Offer New Insights into Brain Growth(spectrum.ieee.org)

23 points by rbanffy 1 year ago | 2 comments

AI has access to a vastly larger working memory than the human brain(davidepiffer.com)

524 points by rzk 2 months ago | 452 comments

A Substack essay by Davide Piffer arguing that AI's edge over mathematicians comes from a vast symbolic working memory rather than superior reasoning.

AI models can provably hide arbitrary capabilities(arxiv.org)

252 points by sbulaev 13 days ago | 118 comments

An arXiv paper showing AI models can embed encrypted neural network backdoors that provably hide arbitrary capabilities from elicitation-based safety evaluations.

Language models can explain neurons in language models(openai.com)

659 points by mfiguiere 3 years ago | 416 comments

OpenAI research post describing using GPT-4 to automatically generate and score explanations of individual neuron behavior in GPT-2.

MIT Study Finds AI Use Reprograms the Brain, Leading to Cognitive Decline(publichealthpolicyjournal.com)

438 points by cainxinth 1 year ago | 432 comments

Article reporting on an MIT study claiming that using AI tools like ChatGPT alters brain activity and leads to cognitive decline.

The path to ubiquitous AI (17k tokens/sec)(taalas.com)

804 points by sidnarsipur 8 months ago | 436 comments

Taalas blog post by Ljubisa Bajic arguing that latency and cost limit AI adoption and presenting its approach to very fast (17k tokens/sec) inference hardware.

GPUs Go Brrr(hazyresearch.stanford.edu)

1056 points by nmstoker 2 years ago | 243 comments

Stanford Hazy Research blog post on what GPU hardware wants for efficient AI kernels, introducing the ThunderKittens CUDA kernel library.