41

Just Chat! / Re: Can someone assist me with some scientific terms?

« on: 14/05/2023 16:48:24 »

No disagreement.But how would the world look if F = GMm/r^2.1?

The "laws" of physics are all circumscribed by boundary conditions such as incompressibility and the like, which actually makes them very like good statute law, where "normally" and "reasonable" make life tolerable. But whilst hard cases make bad law, they make interesting physics!

42

Physics, Astronomy & Cosmology / Re: Can Light Experience 'Time'

« on: 14/05/2023 08:49:12 »

Time is what separates sequential events*. A photon cannot experience sequence as it has no memory.


*an original statement, but apparently my predecessor Prof A Einstein said "time is what prevents everything from happening at once", an arguably anthroponormative definition but having the same import.

43

Just Chat! / Re: Can someone assist me with some scientific terms?

« on: 14/05/2023 00:25:41 »

For example, Newton's laws of motion are the axioms of Newtonian Mechanics and it doesn't matter if Newton's laws were found to be wrong

Spoken like a true mathematician.

It is indeed the case that the whole of newtonian mechanics can be rigorously derived from Newton's laws, but the world would be a very different place if they were wrong!

"Just Six Numbers" (Martin Rees, 1999) explains the profound significance of the six critical ratios that determine the observed size shape and history of the entire universe provided that Newton was right.

44

Physics, Astronomy & Cosmology / Re: Talking about Physics

« on: 14/05/2023 00:09:34 »

New readers:

Note the difference between physicists ("consider a spherical cow in a vacuum.......") and mathematicians ("an instantaneous sample of all the photons in the universe......").

Or as others have put it, "Rocket science is two equations. Rocket engineering is a lot more complicated."

45

Just Chat! / Re: Is there a universal moral standard?

« on: 13/05/2023 11:40:13 »

In the next century, they will have completely different genes from current humans.

So they won't be humans.

The objection is that all living things modify their environment up to the point at which they poison themselves with their own excrement and detritus. Having screwed up one planet, what right do we have to design an animal to modify Mars?

46

New Theories / Re: Universal Utopia? What's The Universal Terminal Goal?

« on: 13/05/2023 11:34:55 »

There would be some optimal and balanced compositions for different cases in different conditions and importance.

A well-known problem with databases. The customer always asks for an accurate, up-to-date, database. The supplier asks "which do you really want?"  Big problem with medical research: we know exactly who entered the trial 5 years ago, but we don't know who died yesterday. So how many people do we need to recruit in order to decide whether the procedure actually extends life? The quicker we get the answer, the more lives we can extend (or not harm) by our proposed intervention, but the less confidence we have in that decision. So we recruit a bigger sample, but the trial may end up doing more harm if the intervention turns out to be harmful, so we gradually expand the numbers if it looks promising, but at some point the cost of the trial will exceed any profit we might make by putting the procedure on the market, or if we expand too slowly someone will come up with a better solution.....

47

Physics, Astronomy & Cosmology / Re: Talking about Physics

« on: 13/05/2023 11:18:48 »

I think you have answered your own question quite elegantly.

Einstein said that repeating an action and hoping for a different result is madness, but two random samples of the same thing is not a repeat - definition of randomness!  So the question is what level of confidence you require to assert that they are samples of the same distribution.

The power of the χ² test is it can tell you not only the extent to which your samples may be said to be representative, but also if the fit is "too good" - evidence of a failure of the mechanism, such as a bit of the primary beam getting through your supposed filter.

48

Just Chat! / Re: Can someone assist me with some scientific terms?

« on: 13/05/2023 06:24:33 »

We start with observations and develop hypotheses that model them and predict the next observations. When we have gone through the loop a few times we can identify theories (a guess at the mechanism behind the obs) and laws (robust mathematical projections of what will happen in the wider universe, regardless of the mechanism).

Thus we have a law of gravitation F = GMm/r²  that is good enough to put a satellite into orbit, but no idea of how bodies actually attract each other apart from the theory that mass bends spacetime.

Underlying all this is the common language of mathematics which is based on the mutual acceptance of certain axioms such as those of Euclidean geometry (similarity, identity, parallelism, etc) with useful theorems (statements that can be rigorously proved as long as the axioms are true) such as Pythagoras.

As far as I can see, a conjecture is a mathematician's hypothesis: a  theorem that looks reasonable but hasn't been proved.

The difference is that mathematical proof is singular and positive (if A then B, QED) whereas scientific proof  is continuous and negative (we haven't found a system that doesn't obey Newton's laws).

49

Physics, Astronomy & Cosmology / Re: Talking about Physics

« on: 13/05/2023 05:52:04 »

If the laws of physics are constant (which seems to be the case) you would always expect to see the same pattern because the probability distribution will be the same each time. The question is how many dots are required to recognise that a sampled distribution is consistent with your theoretical continuous distribution. 

I'm sure ES has a better grasp of formal statistics than I, but the χ² "goodness of fit" test is rattling about in the recesses of my memory. Essentially, you partition your expected distribution E into "cells", calculate the fraction of the total area in each  cell, and count the fraction of actual events A in each cell. Then the sum of the squares
∑_x (E_x/E - A_x/A)²
tells you how close your actual distribution is to the hypothetical continuum.

The point at which you announce the result is up to you! My favorite story concerns the Indian Queens Bypass on the A30 through Cornwall. Not sure how accurate it is but this is my recollection from the local School of Mines about 20 years ago: The engineers had asked for 1000 trial boreholes along the proposed route, to determine the subsoil structure. The first two tests struck granite at a couple of meters below the surface, in the middle of the route. Politicians and accountants announced that  the subsoil was granite and no more tests were required so land was purchased and construction commenced immediately on that basis. It turned out that these were the only two granite boulders in 30 miles of peat bog. The project ran umpteen years late and God knows how many times over budget.

50

Physics, Astronomy & Cosmology / Re: Talking about Physics

« on: 12/05/2023 22:47:04 »

How many dots are needed to see the interference pattern from a distance? Is it the same number for any kind of particle?
And so on. We know that up close, each particle that leaves a dot actually leaves a lot more than a pointlike mark, but we ignore it.

Er, no. If you are thinking about single-photon two-slit experiments, we know that each receiver event involves a single photon with the same energy as the one that left the transmitter: the photon clearly doesn't split and interfere with itself because that would give you two red dots from each blue photon, but what we observe is a pattern of blue dots.

So how many events constitute a pattern? That is pretty much the same question as how long is a piece of string. The more you know about the cause, the less you need to know about the effect to calculate the entire pattern - a case of fully encoded lossless compression. If I know you have a blue light source and two slits with a defined geometric relation, I can tell you what the interference pattern of an infinity of photons will look like as soon as I have detected just one.

If I know nothing about the slit geometry I will need enough receiver events to indicate where I might find maxima and minima, and the more information (events) I have,  the more confident I can be in describing the slit geometry. But some of the events will be redundant repeats, not contributing any more data to my calculation, so it looks as though the error bounds decrease as 1/√N.

51

Physiology & Medicine / Re: At what age did you (or plan to) talk to your kids about sex?

« on: 12/05/2023 22:24:26 »

Plus one for your waist.

52

Physiology & Medicine / Re: At what age did you (or plan to) talk to your kids about sex?

« on: 12/05/2023 18:33:23 »

When you see a scene in a movie where a character is engaging in the sex act, the actor's penis is often covered for a variety of reasons.

Yes, dear, that's how it works.

And in the UK, (a) briefs generally don't have a door - that's boxers and (b) ladies (you need one to make a baby) don't usually have a penis.

53

Physics, Astronomy & Cosmology / Re: Talking about Physics

« on: 11/05/2023 12:14:45 »

I think you need to distinguish between mapping (where you lose dimensionality that can be inferred), minimisation (losing genuinely redundant information) and lossless compression (coding recognised sequences into shorter sequences that fully identify the original). Not sure how any of that can apply to a fundamental particle which by definition does not contain or represent anything redundant or synthesisable from data known to the "recipient".

54

Just Chat! / Re: Great British Bake Off

« on: 10/05/2023 23:02:09 »

So the Parliamentary Conservative Party is not entirely devoid of talent after all. (Other parties are available).

55

Physiology & Medicine / Re: At what age did you (or plan to) talk to your kids about sex?

« on: 10/05/2023 22:50:40 »

I knew the sex act had something to do with naked people in bed

Not according to Hollywood. I've often wondered how Americans manage to reproduce without removing their underwear.

Apropos the question: my wife was a nurse, and we always talked "shop" at mealtimes, having first dissected whatever fish fowl or animal we were going to eat, so the kids were immersed in anatomy and physiology from birth, and helped out on local farms and smallholdings at the weekends. No problem: that's how your siblings, friends and dinner are made.

56

Physiology & Medicine / Re: What usually causes chemo to fail?

« on: 10/05/2023 16:49:59 »

Apropos the OP, chemotherapy, radiotherapy and even surgery are mostly damage limitation and delaying responses to a system that is out of control. "Failure" isn't an appropriate term since "success" would imply the indefinite prolonging of life, which has never been demonstrated. The best we can do is delay the enemy's advance until he establishes another salient.

57

Physiology & Medicine / Re: What usually causes chemo to fail?

« on: 10/05/2023 16:45:16 »

This is the argument that springs to mind every time I hear the advertising slogan "Domestos kills 99.9% of bacteria dead!".

Fact is, it might do even better, depending on the bacterium, but the slogan was apparently advised by the company's lawyers who pointed out the legal bomb that would explode if they claimed 100% and one bug survived from a billion.

58

Just Chat! / Re: Where is the "Spiel Chicken"?

« on: 10/05/2023 10:48:08 »

No voluntary spell check via Microsoft Edge. I get flags for unrecognised words but nothing more, and I don't recall ever being offered it.

59

New Theories / Re: How close are we from building a virtual universe?

« on: 10/05/2023 10:44:07 »

Our ancestors survived by hunting and gathering. Why should they develop agriculture and undergo industrial revolutions?

You can still survive as a hunter-gatherer if the local population density is small and the environment can sustain it without significant intervention. Problem is that politics, greed and overpopulation are displacing those who know how.

60

Just Chat! / Re: Is there a universal moral standard?

« on: 10/05/2023 10:37:39 »

They can agree on one thing while disagree on other things.

Funny, that. Lots of people think that consensus equals correctness, but we scientists know better. The important question is which hypothesis stands up to test. If an idea is not testable, it is of no consequence.

What on earth is a "newly meaningful way to be human?" The only meaningful way is to have human DNA.