Big Bang Bullets II

Page 7 - Seeking answers about space? Join the Space community: the premier source of space exploration, innovation, and astronomy news, chronicling (and celebrating) humanity's ongoing expansion across the final frontier.
This month's All About Space contains an interesting piece about Artificial Intelligence being used on the BB. It does not give an actual reference, but I have managed to find it:

Neural networks have made a breakthrough in modeling the Universe (universemagazine.com)

How does this fit the bullets?
(I don't know - I am just interested).
It looks like they are using it for predicting the spectral results found in clouds. The ability to predict what the spectral lines should look like, especially with respect to the Lyman-Alpha Forest, should add both support and possible refinement to the model.

Only if there is a hiccup between predictions and observations in, say, the CMBR, then we maybe headlines will come our way that question some features of the BBT. I can't imagine anything that would warrant a major change to BBT, however.

Apparently, nothing in this AI effort suggests any issue, else that would be the headline. :)
 
  • Like
Reactions: Catastrophe

Catastrophe

"Science begets knowledge, opinion ignorance.
Helio, may I ask your esteemed opinion please?

I have been taking your BBT as being after the hiatus surrounding a supposed "singularity". In other words, in my statements, I have been taking the BB as starting after the problem of division by zero. After this non-scientific, metaphysical/philosophical hiatus, the BBT works extremely well, as you have pointed out. :) :) :)

However, I have recently been reading several books, which seem to state that the BB includes (division by zero) and the once accepted singularity. Have I been wrong in stating that BBT begins after the discredited singularity? I have referred to (t = 0) as much as possible (to avoid referring to a singularity. Perhaps I should check the dates on the books more carefully (they seemed otherwise OK) but I am pretty sure they were not all out of date (as much as any book is). What is your opinion of the official stance please?

Cat :) :) :)
 
However, I have recently been reading several books, which seem to state that the BB includes (division by zero) and the once accepted singularity. Have I been wrong in stating that BBT begins after the discredited singularity? I have referred to (t = 0) as much as possible (to avoid referring to a singularity. Perhaps I should check the dates on the books more carefully (they seemed otherwise OK) but I am pretty sure they were not all out of date (as much as any book is). What is your opinion of the official stance please?
I wonder if there is an "official" position. But science, by definition, cannot go to places that are beyond any hope of falsification. If it is deemed impossible to test a t=0 event for the entire mass/energy of the universe, then it is metaphysics.

Some scientists may be convinced the t=0 singularity is valid simply by extrapolation from t=10^-43 sec, or by extrapolating from perhaps the 10^-12 sec (one trillionth) of second that comes with observational tests (CERN) as we have discussed elsewhere.

The multiverse idea is arguably not a true scientific theory since no one can possibly imagine a way to test it, excluding a big "bruise" in the universe due to a brane concussion, though none has been found.

So, similarly, if the multiverse is metaphysics then so too is pushing the BBT to where it cannot be tested, thus it can't be part of a scientific theory, but it is suppositional science (metaphysics). It's difficult to suggest that scientists and journalists should not mention a possible singularity scenario because doing so, admittedly, adds great sizzle to the BBT story. But the real story, as you know, is that the theory began starting from today and winding the clock backwards to see what fits the predictions (Lemaitre, 1927). They fit down to as far as we can go, and that is to the first trillionth of a second only.

I just think it doesn't hurt to cut folks slack if they like the thought of a singularity. Look at how few times I criticize those who color our Sun that awful color - yellow, knowing demonstrably that it has not a hint of a tint of yellow. :)
 
Last edited:
  • Like
Reactions: Catastrophe

Catastrophe

"Science begets knowledge, opinion ignorance.
View: https://imgur.com/a/o3kgocO


Helio, as you know, any function divided into 1, gives a graph essentially based on the above. I chose this to avoid "mirror image" graphs such as those you get from 1/x or 1/x^2 et cetera. My point is that I see a "hare and tortoise" paradox here. It is my conjecture (pure metaphysical imagination, I admit) that there may be a function here which produces the t = 0 problem. Just look at the 1/e^x graph. Wouldn't it be fun to find that this backward extrapolation invokes an impossibility which disguises the "real" shape of the line at the smallest time values? And all these minute time intervals and soaring temperatures are the result of this unwarranted extrapolation.

I am, of course, mentioning this purely as a little metaphysical fun.

Cat :) :) :)


P.S. I do know that 1/e^x has mirror images, but these are not shown in the selected graph ;)

P.P.S. Note how the y value is flat from x = -5 to x = -3, so that a straight line would fit the facts over this range. Only approaching certain values does the "extrapolation" break down, and the function shoots to "infinity". :)
 
Last edited:
Getting back to using "AI" to "model the universe", I see a problem with "training" the AI algorithm. If it is "trained" to match a theory, rather than actual observations, then it will, of course, "find" that theory. Where we have no observations, it can only find the theory. Where we have some observations that require interpretation, then it seems to me that we need to be very careful to not train it to only see our current interpretations.

So, I see AI only as a tool for looking at a multitude of observations and finding patterns that we may have missed. I don't see AI as being "smarter than humans" and being able to independently verify our theories that are mostly conjecture.
 
  • Like
Reactions: Catastrophe
Meanwhile, what do you think of #156? ;)

I have the same misgivings about extrapolating the BBT back to extreme densities and temperatures, and especially to extremely fast expansions.

As an aside, on the mathematics. division by zero at a point on a graph is not necessarily going to give a "blow-up" result. Yes, it is "undefined", but there are mathematical methods for defining the result, and some come out as reasonable and realistic numbers.

To explain a bit, if you have a theory that give a functional relationship as f(x)/g(x), and g(x) = 0 at x = 0, we really need to think about what f(x) is doing as it approaches zero. In some cases, the limiting value of f(0)/g(0) is a finite number. The easy way to think about that is if the two functions are simply multiples of one another. If f(x)= 2 x g(x), then the relationship can be algebraically simplified to 2g(x)/g(x) =2 for all of x, including x=0.

For more complicated functions, think about their Taylor's Expansions as they approach zero. For instance tan(x)/x approaches 1 as x approaches zero, and the expansion terms divided by x are 1 + terms with x in the numerator, not the denominator.

There are more complicated limit calculations for more complex functions.

And, there are also cases where the limiting result is "infinity". A distinction is made for functions that go to positive or negative infinity on both sides of the zero point, compared to those that go from negative infinity to positive infinity across the zero point.
 
  • Like
Reactions: Catastrophe
I'm sorry I failed to get back to this post, Cat.

UNC addresses this question nicely.
View: https://imgur.com/a/o3kgocO


Helio, as you know, any function divided into 1, gives a graph essentially based on the above. I chose this to avoid "mirror image" graphs such as those you get from 1/x or 1/x^2 et cetera. My point is that I see a "hare and tortoise" paradox here. It is my conjecture (pure metaphysical imagination, I admit) that there may be a function here which produces the t = 0 problem. Just look at the 1/e^x graph.
Any 1/x^n will get you, eventually, to a zero. The tortoise always crosses the finish line because the time has an inverse effect on halfing the distance each time, so it's not a paradox.

But like your tortoise example, we must also consider the other functions as t--> 0, like temperature and density. Those don't go to zero or any other number since they keep climbing towards infinity. Some may say otherwise, which is fine for math, but science claims require testing. The closest science has reached seems to be where t ~ 1E-12 sec. thanks to CERN efforts.

Wouldn't it be fun to find that this backward extrapolation invokes an impossibility which disguises the "real" shape of the line at the smallest time values? And all these minute time intervals and soaring temperatures are the result of this unwarranted extrapolation.
I'm unclear what your idea is here.
 
  • Like
Reactions: Catastrophe
For more complicated functions, think about their Taylor's Expansions as they approach zero.
Yes. And one doesn't need too go far in crunching those extended terms in a Taylor series to realize a finite answer is the result.

And, there are also cases where the limiting result is "infinity".
Yes. F(x) = x alone does this, though exponents will get there quicker, of course. :)
 
Jun 14, 2022
4
1
15
Visit site
OOPS… I left a hole in. Sumittal:

“Third, today, the universe is comprised of Dark Photons (may explainThird, today, the universe is comprised of Dark Photons (may explain A BLACK HOLE)
 

Latest posts