Follow Users, not developers

Carlos E. Perez @IntuitMachine OpenAI just came out with their Prompt Engineering guide: https://platform.openai.com/docs/guides/prompt-engineering Replying to @IntuitMachine The AI groups are only trying to extend dependency on their solutions, and indirectly on their development tool suppliers. Those benefit from NOT creating true AIs. They make money by working at AI, not supporting AIs that work
Read More »

Quantum components cost for specific problems with specified engineering noise requirements

Chae-Yeun Park @chae_yeun_park I feel that our recent preprint https://arxiv.org/abs/2312.07658 has not gained enough interest, so I want to advertise it a bit. We proved that computing the output distribution is #P-hard for dynamics generated by a wide class of Hamiltonians, including the Ising, XX, Heisenberg models. Replying to @chae_yeun_park If you randomly sample your
Read More »

The usual lossy images from NASA

World and Science @WorldAndScience Astonishing: The James Webb Space Telescope captures the rare moment just before a star’s death in pin-sharp detail. (Credit: NASA, ESA, CSA, STScI, Webb ERO Production Team) https://pic.twitter.com/ofXtCVflZj Replying to @WorldAndScience The usual from “NASA” – lossy images, no meta data, no provenance, no links, very little scientific or educational value.
Read More »

3 Billion humans not on the Internet might be a good thing

Interesting Engineering @IntEngineering Approximately 3 billion people globally lack internet access, a gap companies like Starlink and Astranis aim to fill. Starlink uses numerous Low Earth Orbit satellites to provide internet, but it’s expensive and faces connectivity issues due to the rapid movement of satellites. https://pic.twitter.com/nj5nXlNJQg Replying to @IntEngineering Thank you for sharing this. I
Read More »

Age discrimination in education

Stéphane Hacquard @hacquard_st Hello world! A 2-year postdoctoral position is available in my group @mpipz_cologne. If you like microbe-microbe interactions, microbial genetics, and single-cell omics, this position is for you! Apply here -> https://jobs.mpipz.mpg.de/jobposting/d034a979b2dd0d25ef7ba30601631d9df31a20600?ref=homepage Deadline -> Jan 15th Please share 🙏 https://pic.twitter.com/YiD9xLbkjX Replying to @hacquard_st and @mpipz_cologne When I got past age 65, I found
Read More »

The AI is just a proxy for its owners, developers and community.

Aran Komatsuzaki @arankomatsuzaki TinyGSM: achieving >80% on GSM8k with small language models A duo of a 1.3B generation model and a 1.3B verifier model can achieve 81.5% accuracy, outperforming existing models that are orders of magnitude larger https://arxiv.org/abs/2312.09241 https://pic.twitter.com/a3gpu36hFw Replying to @arankomatsuzaki On life-critical global issues, 99.999% is required. That it possible with care and
Read More »

Pre-compile instructions so all pre-requisites are identified and resolved

Kirk Borne @KirkDBorne #infographic — Explaining supervised and unsupervised #MachineLearning algorithms Source: https://linkedin.com/posts/opensourcecommunity_confused-between-supervised-and-unsupervised-activity-7100765879373619200-MCzj https://pic.twitter.com/CYCRriGK3e Replying to @KirkDBorne Convert this into a form so that it can be absorbed by any AI and used immediately without error. They are going to have to be given the meanings and parameters of the many acronyms you used. You
Read More »

Solar eclipse oscillations dataset for open sharing and updates, active surfaces slide or pull through air

Stephen Wolfram @stephen_wolfram Preparing for 2024 … analyzing video I took at the 2017 eclipse. Here’s the light curve I got. The dip is totality. Anyone know what the glitches up and down are? iPad exposure control? (They’re too slow for shadow bands) https://pic.twitter.com/E9RhIEgMv5 Replying to @stephen_wolfram Stephen, have your favorite AI search ( “solar
Read More »

Gravity can be linear, but also nonlinear with flows and fluctuations

Science News @ScienceNews Jonathan Oppenheim thinks that gravity might be fundamentally classical, meaning it isn’t quantum at all. https://sciencenews.org/article/gravity-quantum-mechanics-physics-theory Replying to @ScienceNews Yes. 20 years ago I used the small network of superconducting gravimeters to measure the speed of gravity. At low sampling rates, the field is nearly perfectly Newtonian. At micro nano pico femto
Read More »