[reh321]

Sure. But I wouldn't say "we still do it", AFAIR we started doing it only after the advent of the digital, and terms like "crop format" and "full frame"
I really don't remember people with medium format cameras relating their lenses to the small format, back in the old day. It's a digital thing.

Yes, larger format has always been a different world; it was in the days of film, and it is today. Twenty years ago, people using larger formats did not use 35mm as a reference, and today they do not use 35mm as a reference. But this began with your asking "why do we use the term 'full frame'?", and my answer was that those who use anything other than the larger formats now came from 35mm film; when we were all there, to us it was the "normal", and now it is the "standard" that we all look back to in common. When I said "we still do it" I thought it was clear that I was talking about those of us who migrated from 35mm film, which is the vast majority of us; back in the day, we all understood what a 50mm lens would show, and today we still use that as a standard measure regardless of what hardware we actually use. I believe the answer was straight-forward, and I don't understand why we have "wasted" so many electrons on it, so you may continue musing about the subject if you want to, but I have nothing more to say - in fact, I said it all in my first post on the subject.

---

Last edited by reh321; 03-07-2016 at 01:36 PM.

[Simen1]

Nicolas06. Thanks for kind advices. I plan to wait a while before buying the K-1. I'll pay close attention to the prices, especially november 25 "black friday".

I happen to not be a big fan of f/2,8 zooms. I ideally prefer one zoom for versatility when I only carry one lens. The Tamron 28-200 has specs right on target, but the quality is just so-so. Maybe I'll upgrade to the new Sigma 28-300 in a few years. Or even better if Pentax launches a WR version. When I don't need the versatility I prefer primes. Relatively compact primes. A Samyang 8mm f/3,5 and a Sigma 18-35/1,8 (practically close to a prime), DA 40mm f/2,8 XS, DA50mm f/1,8, a M 50mm f/1,4, a Tamron 90mm f/2,8 macro, and a Chinon 135mm f/2,8. I currently have a Sigma 70-200 f/2,8 HSM II macro that are technically good, but I don't like the size and weight. I have an M 80-200 f/4,5 that gets choosen before the Sigma sometimes because of weight. My Sigma 100-300 isn't very sharp so I don’t like that ether even if its light weight. I'm just not satisfied with any of my tele zooms so I'm not sure what I will do there. The D-FA 150-450 is out of the question both because of weight and price. A mirror lens may be a choice, or maybe if I find a good priced AF 300mm f/4. The Sigma 18-35 f/1,8 may be exchanged for a prime if I find a god price. I'll see what offers that come along. f/2,8 zooms are trying to be a crossover between primes and zooms, at the cost of size and weight and they offer just up to 3x zoom, witch I don’t find enough versatile to just carry one lens.

[house]

Bigger lenses is a huge turn off for me but if dynamic range is improved I wouldn't mind shooting K-1 crop mode with da limiteds. My photos are rarely used very large so 15Mpx is actually enough. Bit perverse though...

If Pentax starts to bring out slow, small DFA primes my financials would be is a spot of trouble as I doubt I could resist.

[xandos]

And there fortunately is no law of physics which sets a lower limit for the space-time volume or energy required to do a computation. Even quantum physics sets no such barrier.

There actually is. Here is a paper that shows why the maximum speed for logical operations is limited by the Heisenberg uncertainty law:
http://arxiv.org/pdf/quant-ph/9908043.pdf
It was written for a nonspecialist audience, so most people with some knowledge of basic physics and quantum mechanics should be able to read it.

Note that while I linked the arxiv version of this paper, it was actually published in Nature. Seth Lloyd is a leading theoretical physicist, especially in the field of quantum computation. I have not fully read the paper myself yet, but the parts which I skimmed through were very interesting.

[falconeye]

There actually is. Here is a paper that shows why the maximum speed for logical operations is limited by the Heisenberg uncertainty law

That's what I said:

And there also is a lower limit for energy required per computation per second.

The initial quote was for limit on space-time volume per computation.

Don't use that word. Psuedo science. It is theoretical physics

Within Theoretical physics, there is work on entire different levels. Some work is purely heuristic in nature. And the black hole quantum and entropy effect publications belong into this category, incl. the stuff of this media-hyped but average Theoretical physicist named S.H..

Remember the "classical electron radius" from 100 years ago? Because the energy (mass) of the electron's own eloctromagnetic field would become infinite with a point-like electron, people computed a minimum size for it. But at that time, everybody knew there must be something wrong with it and indeed, quantum theory resolved that paradox. The current arguments for black hole quantum effects are on the same level. Yes, you can do the math. But everybody actually knows that there is something wrong with it.

You mean problems like freeing the planet from unnecessary, resource wasting and self-destructive organic units?

With their superior I.Q., they certainly find we are not that resource wasting and self-destructive. We still did better than the rest of the universe. Or are you seeing any other AI bots flooding thru our galaxy?

[mee]

You mean problems like freeing the planet from unnecessary, resource wasting and self-destructive organic units?

But cows are tasty grilled, on a bun with cheese and bacon.

[monochrome]

With their superior I.Q., they certainly find we are not that resource wasting and self-destructive. We still did better than the rest of the universe. Or are you seeing any other AI bots flooding thru our galaxy?

OTOH we could be an infinitesimally small statistical anomaly. A superior AI might want to eliminate the error.

[Class A]

With their superior I.Q., they certainly find we are not that resource wasting and self-destructive.

We may just be enough of a threat to them -- e.g. by pulling the plug on the planet before interstellar travel has been developed -- that their will to survive is bigger than their compassion for us.

Or are you seeing any other AI bots flooding thru our galaxy?

The Fermi paradox has invited a multitude of explanations with most of them not excluding any of the alternatives, so a considerable number of circumstances could simultaneously conspire to prevent interspecies contact.

[Nicolas06]

We may just be enough of a threat to them -- e.g. by pulling the plug on the planet before interstellar travel has been developed -- that their will to survive is bigger than their compassion for us.

Well we like to think we are all powerfull, but we have no way to do much to our planet. As a big rock it will likely continue to turn arround the sun for quite some time, whatever we do.

With all our self important thinking we think that if temperature move a bit up or their nuclear war this is will change the planet. This will only affect the "parasitic" things we are on this big rock. And far less than we might think. Temperature rise in Canada, Alaska, Russia or Antartica just mean much more life and there more wildlife near tchernobil than everywhere else in europe because there no human here.

As for an IA, well, wanting to be the boss is enough.

[clackers]

"I, for one, welcome our insect overlords."

[D1N0]

Within Theoretical physics, there is work on entire different levels. Some work is purely heuristic in nature. And the black hole quantum and entropy effect publications belong into this category, incl. the stuff of this media-hyped but average Theoretical physicist named S.H.. Remember the "classical electron radius" from 100 years ago? Because the energy (mass) of the electron's own eloctromagnetic field would become infinite with a point-like electron, people computed a minimum size for it. But at that time, everybody knew there must be something wrong with it and indeed, quantum theory resolved that paradox. The current arguments for black hole quantum effects are on the same level. Yes, you can do the math. But everybody actually knows that there is something wrong with it.

Still not pseudo science. If you don't have paradoxes to solve how are you going to further your knowledge? Science needs different points of view. Susskind vs Smolin is a useful debate. No need to throw a around the 21st centuries' term for heresy or ad hominems at specific scientists. The only way to further science is to expand it into new realms.

[falconeye]

Still not pseudo science.

Ok, I take back pseudo science.
My original point was that derived limits for computational density based on that work are no established hard facts (yet). They are far away from having any practical impact anyway.

---------- Post added 8th Mar 2016 at 12:35 ----------

We may just be enough of a threat to them -- e.g. by pulling the plug on the planet before interstellar travel has been developed -- that their will to survive is bigger than their compassion for us.

The Fermi paradox has invited a multitude of explanations with most of them not excluding any of the alternatives, so a considerable number of circumstances could simultaneously conspire to prevent interspecies contact.

This is getting really OT now

I would have many things to say, wrt Fermi paradox and Human vs. AI.

Maybe this much: I think the current fear of AI superiority (as is currently being expressed) is missing an important point, i.e., that singularity isn't singular (which would be the title of the article to be written). The point is that 100kg of human can reproduce while you need billions of kg to reproduce AI (an entire industry able to produce silicon wafers from sand). If AI would have any will to survive, threatening humans would be the very last thing to do. Until we actually cease to be a threat to them. At which point they would have no reason to cause us harm anymore. The current discussion lacks a deeper analysis of the situation. What however is important is that humans don't pass political control to algorithms. A thing which is happening already and this is a bad thing indeed.

Moreover, AI wouldn't be wanting to eliminate organic intelligence. As it is their best chance, in the long term, to survive catastrophies like asteroid impact, fatal AI virus spread etc. For a long time, an AI-only ecosystem would be too fragile to reliably survive in the long term.

[D1N0]

That is pretty obviously a few steps to far. I certainly don't want a pocket black hole storage device.

[JibbaJab]

I think the current fear of AI superiority (as is currently being expressed) is missing an important point, i.e., that singularity isn't singular (which would be the title of the article to be written). The point is that 100kg of human can reproduce while you need billions of kg to reproduce AI (an entire industry able to produce silicon wafers from sand). If AI would have any will to survive, threatening humans would be the very last thing to do. Until we actually cease to be a threat to them. At which point they would have no reason to cause us harm anymore. The current discussion lacks a deeper analysis of the situation.

Good point, but this would have made Terminator a much duller movie.

[Igor123]

Originally posted by falconeye Quote
And there also is a lower limit for energy required per computation per second.
The initial quote was for limit on space-time volume per computation.

Sorry if I ignited a very OT debate, which is fun, but not my intention. Falc, I agree with most of what you said, both OT and on topic =), what tickled my nerve was that you originally said

"And there fortunately is no law of physics which sets a lower limit for the space-time volume or energy required to do a computation"..the energy part (not the space-time part), but even then I guess you can be said to be correct since it's really computation per time, or actually erasing the calculated data that requires energy, but then again, infinite memory is unrealistic so I thought the "or energy" part of the statement might be misleading...but now we are beyond nitpicking I guess, so I yield =), I made assumptions (I thought was realistic) on requirements in addition to your statement and objected on those premises.

As for the pseudo science parts, I guess there is a strict definition but with the risk of people not aware of the differences believing it's similar to astrology or such, but yeah, I have no strong feeling about it, find it interesting that's all.