Small experiments with outstanding noise – shared globally for learning and development
Julian Stirling @stirling_julian I’m delighted to still be publishing papers almost 2 years after leaving academia to become freelance. This article is more of a “rant with references”, I try to tackle the problem of perverse incentives that prevent open science from flourishing.
https://royalsocietypublishing.org/doi/10.1098/rsta.2023.0215
Replying to @stirling_julian
There are fundamental experiments that can examine fundamental assumptions and require massive machine learning, but also reveal practical solutions. They are information rich by design. The Franck-Hertz is one. The notion I have tried to encourage is to have sites around the world run the experiments (and a few variations) 24/7 with lossless data storage and sharing — so new machine learning and algorithms can come from anywhere. Radiation monitors, software defined radio networks.
Even simple quantum devices can be useful — if you focus on the quantum noise – since it is infinite in variation,and linked to all global networks at the parts per trillion level, and many at parts per billion. So global open experiments. The goal is not any one device or group, but statistics, machine learning, real world applications, efficiency prizes. It is hard to learn some things (atomic and nuclear processes) if you do not have good experiments that can generate deep “noise”. LIGO earth noise, CERN nanovolt noise, radio telescope noise below 1 MHz, gravitational noise from the complex gravitational potential changes in the atmosphere and oceans at pico and femto levels, social noise.
Noise and experiment, information and true noise are linked. The noise of some experiments and devices is more valuable than the normal use. A “bad” amplifier might teach more than a canned product. Franck-Hertz is an open 3D transistor model of many things. A moving magnet and wire is not constant, if you measure the noise carefully and deeply. Radiation noise has deep model potential. A flame has infinite variations, when you look close and fast and fine enough.
Little simple things, combined globally, and teaching “fast and effective” – rather than giant monopoly projects that do not share anything useful.
