> I also wonder how effective this would be as a compression algorithm My guess is it would be the opposite of compression in many cases, due to the number of

> I also wonder how effective this would be as a compression algorithm My guess is it would be the opposite of compression in many cases, due to the number of required cells you need to store. Using weird, creative ways of encoding data tends to be bad for compression, and in the very best case you'd have something that competes with JPEG but is much slower. It's an enticing idea, just like that concept of encoding arbitrary data by looking for the exact same bit pattern in decimal places of Pi and then storing that position. But reality is often disappointing, and it doesn't really work because more often than not you'll need more space to store the decimal position than the original information.