Heavy Wifi interference Solved – Kitchen magnet on the usb port on a raspberry pi. So simple it hurts. Magnetism is a reset field. Wow.
Super Important – Pro active Water Shortage rules for the summer. Same as earlier water shortages only better and in time. // Per #water
Super important – Action of Peace – Heatwave relief – Used boiled seasalt+tablesalt+water. Let it cooled. Sprayed in the air. Created clouds that will ?rain some time later. // Per
Machine Learning – Convolutional Model created in colab. Serialized for chainer cpu on a rpi4 2GB 32GB zram Speedup using multiprocessing 2.5 times faster using 4 cores. // Per
Peace Idea – Russia Siberian Rain – You need local siberian salt and local seasalt to stop cloud splitting. Boil the salt in water. Let cool. Spray the saltwater in the air from some elevated position in the solar rays so that selected salt can go up. #forestfire#Siberia // Per
Rain Success – I think it was the cooled boiled local swedish falk_salt+water that did it. Wow. Rain. Its still pretty hot indoors but outdoor its 18C. Perfect. Images from 27, 29 june
Livelihood after covid19 – Idea – IKEA could sell livelihood foodstand equipment. Improve small livehoods. Better hygien better protection against food/water shortages and finnaly covid resistance. // Per
Summer Heat – Too hot weather. Personal cooling with a dishwashing cloth and a bottle of cold water. Place the wet dishwashing cloth directly on the skin. Good to have with you outside. Stay safe stay cool. // Per #heatwave
Major Rain and cloud Discovery – Used two kinds of salt. Sea salt and course salt. Filled a bottle with warm water and sprayed a little outside. Also placed a thermos with salt and hot water outside. It created two kinds of clouds? I think ordinary caurse salt is better for rain.
Super Intelligence – Solving Word Crisis by secondary loss functions – Using chloroquine as a starting medicin researchers get ideas and money to invent a better cure. But you need the first treatments to measure against. The loss function. So for all crisis there is a first loss
Electric generator – A hard drive has a gas in the housing which enables better performance. Thus, gases have rotational velocity phases? Here perhaps one can extract energy by going from one phase to another. Any influx of single molecules from another gas would provide heat.Per
Idea – El generator – Could you stop a rotating disc with a laser. Like a CD player disc. If so what effect would you generate. // Per
Corona Virus Innovation – garlic + salt+water + paper to blow your nose in.
Bussiness opportunity. Optimized tissue paper.
Machine Learning Physics – What if water holds the key to forces. Two Entangled hydrogen with obstructing oxygen atom to keep the connection. So forces are conservation of uncertainty. Which hydrogen holds which opposite quantum state. // Per Wow
Education – Computer math literate – Game – Which function plot belongs to which function data. Learn simple classification by looking. // Per
Corona – plus 26 days from 13 of june ? Sweden.
Machine Learning Math Revolution? – High frequency has something to do with overcoming determinism obstacles. That’s why things iterate. It can see somehow. // Per
Machine Learning Universe API – Record mathematical events such as changing the target value of an equation. Train a model(). Run the model() on ?quantum noise for real world predictions. // Per
Scratching skin remedy – Spray salt+water on your hands and rub them over the problem area until its dry. Repeat. Just water with salt dissolved. Friction heat activates I guess the salt to deactivate the virus.
Machine Learning Idea – Noise level experiment for data. Made a machine learned experiment. Graph showing noise level in hidden layer. // Per
Corona Virus Idea – Blow your nose in a saltwater wet paper towel. This way the virus does not have time to escape. The nose snot is salty for a reason. Use table salt for your advantage. // Per
Machine Learning – Create a Fourier experiment. Set two random vectors. Smooth the other. Calculate the loss between them when adding a sinus function. (yt-(x-sin(x)))^2->0. Vary the sin with different frequency…
Machine Learning – Action Loss Manipulation – A painting contains initial errors far away from square_error. So change the loss value so it fits a pencil. So loss[i] = (x-x.mean())[i] for a long x vector. // Per testing
