3 ms·
The idea is interesting and probably not unique. If you try to paint a diagram of it, how would that look? It's not time that is accelerating, so you don't just
by posterboy 8y ago
The idea is interesting and probably not unique. If you try to paint a diagram of it, how would that look? It's not time that is accelerating, so you don't just draw the time axis just longer with wider gaps between the minute marks. Which leaves me guessing and remembering my old, discrete notion that time is count of change of things, so that, for sake of argument, in one unit of time, one state change happens. In that setting, it doesn't make sense to say a change was further apart than another. Actually, information decreases, thermodynamically speaking. So, tracking back infinitely, you'd have to start with an infinite amount of entropy, proportional to energy. But the theory assumes a fixed amount of energy, as far as I know.
And it's not that the lightspeed increases with time, either. Rather, so they say, space expands - which is imaged trivially by stretching the space axis. I suppose that's normalizing for time and you could just as well fix space and compress time, invariably, to arrive at your notion, with a total lack of rigour at that.
So normalizing for