Why Time Flies By As You Get Older 252
Ant notes a piece up on WBUR Boston addressing theories to explain the universal human experience that time seems to pass faster as you get older. Here's the 9-minute audio (MP3). Several explanations are tried out: that brains lay down more information for novel experiences; that the "clock" for nerve impulses in aging brains runs slower; and that each interval of time represents a diminishing fraction of life as we age.
Re:1 Day Expressed as a Percentage of Your Life (Score:3, Informative)
So my conclusion is to go with Einstein in that time is relative.
Except that Einstein's special theory of relativity is talking about time _really_ being relative, not perception of absolute-time being relative.
Re: Relative memory versus time (Score:3, Informative)
At age 57, time doesn't "pass faster" for me than it did when I was 23 or 24, but each day adds a lower percentage of new experiences and memories than it did back then.
Well, duh. Near the level cap, it takes more XP to advance.
Porcupine Tree (Score:2, Informative)
Re:Michio Kaku (Score:5, Informative)
Picard would probably use his 100-billion neurons firing 1,000 times per second = 100-trillion operations/second to ask "Why is Data so slow? Can't he get an upgrade?"
Full attention != Full processing capacity (Score:3, Informative)
Full attention != Full processing capacity
He obviously has tons of background daemons running and was in a situation of "some degree of peril" and physical change (the skin graft thing) which clearly would have triggered several others. A more useful, relevant, pertinent (and I predict...) reliable benchmark would be something like "thoughts per second" or "operations per thought" (since different thoughts would have different operations and a different number of operations). "Thought operations" (or "thoughts") could be a standard for thinking-speed.
So - how an AI thinks is more important than how fast it thinks since "operation speed" can changed via hardware. Thinking-speed is a result of underlying algorithms that actually make up the "I" in AI. Thinking-speed is also affected by the AIs own growth and ability to change itself. IQ of an AI would also result from that.
Which also implies that AIs can be distracted from a task - simply by causing it to a) spend more operations about input received or b) making it think about something else simultaneously. An AI being able to manage that/reduce the effect of the distraction, again, is dependent on design, self-growth and self-modification.
It would be have sounded even more worrying if in those 0.68 seconds, he had as many "thoughts" as an intelligent person would in a year...or two. (Endless loops, non-breaks, etc. adds to the worry.)
Ofcourse, I do agree with your statement. It IS an eternity for any AI.
Sidenote with math-conjectures:
60 trillion OPS = 60 x 10^12 OPS (or in MIPS = 60 * 10^6 MIPS, since we are not assuming only-FLOPS) - and yes, many operations make up 1 instruction, so assume best case is 1 instruction = 2 operations. So 30 * 10^6 MIPS? An X9100 is 32472 MIPS ~= 32 * 10^3 MIPS. So only slightly-less-than 1000 times slower than Data.