Thank you to brilliant.org for supporting PBS.

What if every single black hole that formed in our universe sparked the big bang of a

new universe?

Cosmological natural selection proposes exactly this - but even better, it claims to be able

to test the hypothesis.

Physicists have been struggling for some time to figure out why our universe is so comfy.

Why, for example, are the fundamental constants - like the mass of the electron or the strength

of the forces - just right for the emergence of life?

Tweak them too much and life, stars, galaxies, the universe as we know it wouldn’t exist.

In recent episodes we explored one possible explanation for this - the anthropic principle

and the idea of the multiverse.

If there are countless universes with different fundamental constants, then it’s not surprising

that a few exist with the right numbers for life - and certainly not surprising that we

find ourselves in one of those good ones.

But if you don’t like the anthropic principle - and many scientists don’t - then rest

assured, there’s an alternative.

You only need to accept two things: that our universe formed inside a black hole, and that

universes can evolve.

Our universe appears, in some sense, designed.

It has finely tuned parameters that seem deliberately set for a particular outcome - life.

There’s another example in nature where the illusion of design has a perfectly natural

explanation - and that’s life itself.

We now know that the fantastic complexity of living organisms is an inevitable consequence

of evolution by natural selection.

Inspired by biological evolution, theoretical physicist Lee Smolin came up with Cosmological

Natural Selection.

It goes like this: the formation of a black hole triggers the formation of a new universe

“on the other side” in a new big bang.

Those daughter universes go on to expand and make their own black holes and hence their own

daughter universes.

But in their formation the fundamental constants of the daughter universes are shifted slightly

and randomly from their parent - mutations are introduced.

Some of those shifts improve the daughter universe’s ability to form new black holes.

Those universes have an advantage in propagating their cosmic genetics, and so gradually the

ensemble of all universes get better and better at making black holes, just as biological

organisms with helpful mutations can get better at surviving and reproducing.

Now by happy chance there’s a correlation between making lots of black holes and making

life - both require stars.

The universe that is better at making stars is better at making planetary systems is better

at making us.

Seems fair enough.

But is it any more than a cool story?

Bro?

Let’s take this apart to ask two questions: is it plausible, and is it testable?

First up, for any of this to make sense black holes need to create universes.

This is by far the most speculative part.

In fact we have no idea, and only very tentative reasons to think so.

The idea originated with one of Lee Smolin’s mentors, Bryce deWitt, who postulated that

when a black hole collapses, its mass doesn’t all end up stuck in the central, infinitely

dense singularity.

Rather it sort of bounces - but unable to exit the event horizon of the black hole,

it forms a new region of spacetime, effectively creating a new universe.

The details of how this happens is presumably buried in the as-yet-unknown theory of quantum

gravity.

There are various proposals for how such a bounce might happen - all of which are massively

speculative, and perhaps we’ll cover another time.

John Archibald Wheeler expanded on the procreating black hole idea by suggesting that the fundamental

constants of these new universes could be different to their parents.

This seems plausible - if fundamental constants can change at all then surely it’s in the

highest possible energy environments, which is exactly the end of a black hole collapse.

Perhaps the configuration of the geometry string theory’s extra dimensions gets shifted

- this would do the job.

Inspired by this idea, Smolin added one thing: what if, when universes reproduces, the constants

aren’t randomly reconfigured but rather change only slightly - analogous to a small

number of genetic mutations.

If that were the case, a sort of evolution by natural selection would become as inevitable

as biological evolution.

We have no good reason to believe any of this procreating universe stuff - and Lee Smolin

has readily admitted that.

The point is to instead ask: what if it’s true?

What are the consequences?

And can we test them?

The exponential nature of the proposed process means that the ensemble of all universes should

very quickly be dominated by ones that are extremely good and making black holes.

Any given universe may not be totally optimal because its constants varied randomly from

its parent - in the same way that any given living organism isn’t the paragon of its

kind.

So there’s a prediction: the fundamental constants that define black hole production

should be close to optimal in a given universe, at least for a given mechanism for making

black holes.

In our modern universe, black holes are made when the most massive stars explode as supernovae.

There are other ways to make black holes, and we’ll come back to them.

So we should expect our universe to be optimized for producing as many of the most massive

stars as possible.

Well is it?

It’s actually very hard to say, but it does seem like there’s some fine-tuning there.

Stars are formed when giant clouds of gas collapse under their own gravity.

But in order for that to happen the gas needs to cool to just a few degrees above absolute

zero, rather than the typical 200-kelvin temperature of the typical interstellar nebula.

That cooling is extremely slow if the gas only contains the hydrogen and helium produced

in the big bang.

Heavier elements and molecules allow clouds to cool and stars to form much more quickly,

and of these,

carbon monoxide is by far the most important coolant.

In addition, gas needs to be shielded from the heating effect of other stars - and that

seems to require the presence of tiny particles of ice and hydrocarbon dust.

So without carbon, oxygen, water, and chemistry in general, far fewer stars and so far fewer

black holes would form - and of course these factors also seem to be essential for life.

But what about other sources of black holes?

Theoretical physicist and cosmologist Alexander Vilenkin proposed that if a universe lasts

forever then in the distant future, quantum fluctuations of that near vacuum will cause

black holes to spontaneously appear - and given infinite time these will eventually

outnumber those produced by stars or stellar black holes.

If all this is true then the most black holes would be produced by the biggest universes

- more space means more chances for these quantum fluctuations.

That favours lots of dark energy generating rapid expansion.

And that is definitely not our universe.

Lee Smolin has various arguments against this: for example, we don’t know that our physics

can really be extrapolated to the insanely long timescales required for these quantum

fluctuations to happen.

I would also add that even if Vilenkin’s argument holds, there are no doubt different

regions in the landscape of possible fundamental constants where different types of black hole

are optimized.

This would lead to multiple branches of the cosmic genetic tree - some of which correspond

to producing lots of stellar black holes.

And naturally we’d find ourselves on one of those branches because those also happen

to be the ones that favour life.

But, whoopsie, I just invoked the anthropic principle, which is exactly what were trying

to avoid with this whole idea.

As speculative as all of this is, Smolin claims there’s a concrete test for the idea.

If cosmological natural selection is true, then the fundamental parameters favouring

black hole production should be optimized completely independently to those that also

favour the appearance of life.

And he suggests there is one such parameter.

But first some background.

When massive stars die, they actually mostly produce neutron stars - planet sized balls

of neutrons so dense that they teeter on the edge of collapsing into a black hole.

Black holes only form when the neutron stars is above a certain mass limit.

Now it may be that in the cores of the most massive neutron stars, some particles can

convert into strange quarks.

The resulting material is even denser than the original neutron star, and so brings the

star closer to collapse.

And the lower the mass of the strange quark, the easier it is to convert lighter particles

into strange quarks.

That in turn means less massive neutron stars would be able to collapse into black holes.

Surely, then, if universes evolve to maximize the number of black holes, then the strange quark

mass should be optimized to make the cutoff between neutron stars and black holes as low

as possible.

Lee Smolin calculates that optimized cutoff at around 2 times the mass of the Sun.

So, if this universe is optimized for black hole production then there should be no neutron

stars more massive than 2 solar masses.

And?

Well, the most massive known neutron star is 2.17 solar masses, discovered just this year.

Now perhaps the extra .17 can be factored into the uncertainties of the theory...

Or perhaps this is the falsification we were looking for.

We await Smolin’s comments on this.

I’d like to add my own objection: cosmological natural selection is meant to explain the

fine tuning in the fundamental constants, which appear to be either set by design or by extreme luck.

It tries to avoid the anthropic principle by proposing a natural selection that favours

black hole production, and it’s just a happy coincidence that the same factors also favor life.

But then do we really gain anything?

It just so happens that carbon and oxygen are good for both black hole production and

organic molecules ... but what if it was, I dunno, beryllium and boron that helped stars

form - or other elements that were useless to life.

If we causally disconnect the selection process for cosmic reproduction from the emergence

of life then it seems we still have to invoke a good lot of good luck?

Overall, cosmological natural selection is an appealing idea because it seeks a natural

explanation for fine tuning, and one that parallels a known process in nature - biological

evolution by natural selection.

It also seems to give us predictions that we can try to test and falsify.

And even though this idea is probably not true, it’s really important to remember

that speculative ideas like this are exactly how we probe the edges of science.

No one of them is likely to be true, but they help us explore the vast space of all possible

realities - where somewhere is hidden the true nature of our reality.

Or, you know, our universe's momma might be a black hole, and we live in an endlessly evolving, proliferating

space time.

Thank you to brilliant.org for supporting PBS.

If you want to understand astrophysics you’re going to need to have a solid understanding

of relativity.

Brillaint.org has a course on special relativity that includes interactive challenges and problems

to solve.

A hands-on approach can guide you through thinking strategies for challenging subjects

like relativity. In this course you’ll begin by understanding Einstein’s postulates

and the Lorentz transformations, and as your knowledge advances you'll learn how faster

than light travel can break causality and you'll even solve the famous twin paradox.

To learn more about Brilliant, go to brilliant.org/Spacetime.

Hey everyone, thanks for watching - your support each week is what makes this show possible.

Now, totally optional, but one way to help out even more is to become a patreon contributor.

Even 2 bucks a month gets you access to our hopping discord channel.

Thanks a ton if you’ve already joined us, and today an extra special huge thanks to

Big Bang supporter Craig Stonaha. Craig, as a small token of our appreciation we’ve

contacted our friends at the large hadron collider - they’re going to make you a black

hole universe.

Please email us with the configuration of fundamental constants you’d prefer and we’ll

get it right out to you.

Oddly enough it’s the same shipping company as our Merch Store, which you can check out

at pbsspacetime.com

Last week we talked about the doomsday argument - the unnerving idea that, statistically speaking,

there are aren't likely to be vastly more generations of humans ahead of us than there have been in the past.

There were a lot of counter arguments - perhaps good ones because apparently we're still here.

Many people come up with a similar objection.

I'll quote Mr Fantastic, who articulated it well: A human born 2 million years ago would

come to the conclusion that the end of the world is nigh, and so would a human born 2

million years from now.

If we accept the reasoning of the doomsday argument, doesn't this just mean that everyone,

for all of history, would come to the conclusion that we're all going to die sooner rather

than later?

This point it totally valid - in fact a cro-magnon should reach the same conclusion and predict

doom long before the 21st century - and obviously they would be wrong.

But the doomsday argument isn't saying that every member of a species who employs this

reasoning to predict their species doom is going to be right.

It says that the majority will be right - if they predict that there will be a similar

number of future generations as past generations.

And by similar, I mean within a factor of a few.

So ancient philosophers would have got it wrong - and perhaps we'll eventually also

be wrong ancient philosophers to some very distant future generation.

The point is if any given individual assumes that we're randomly sample from all generations who thought about the

doomsday argument then chances are they didn't come near the start of their species.

Now, the real problem with the doomsday argument isn't that users of it in the distant past

would be wrong.

Rather, it's that it's not at all clear that it's reasonable to count ourselves as "randomly

selected" from all of the users of the doomsday argument.

Zahaqiel highlights this trickiness in defining reference class with a great example:

Step 1: Define reference class as "homo sapiens sapiens existing concurrently with the internet".

Step 2: Observe that the internet has existed for approximately 30 years.

Step 3: Assume self-sampling assumption makes sense and internet access is bell curved over

over the duration of its existence or skewed towards late-phase access...

Then the internet's going to cease to exist some time in the next 30 years guys.

No one highlights another misuse of this idea.

To quote: "What are the odds that I'm born as the prince of France?", asked the prince

of France.

Well, from the Prince of France's perspective, the answer to that question is a probability of 1.

From everyone else's perspective, the answer is a probability of 0.

And this really highlights the challenge in identifying our "reference class" for this sort of anthropic

reasoning. Now, the Prince of France knows that he's the Prince of France, so from his perspective

the likelihood that he's the Prince of France is indeed 1.

But imagine the Prince of France was raised secretly in a normal French family.

He doesn't know who he is - all he knows is that someone in the population is the Prince of France.

If you ask him for the probability that he's the Prince of France he should probably say

1 in 33 million, or whatever the male population of France is.

The point is that you need to take into account prior knowledge when you're defining your reference

class.

Nick Bostrom has another nice example of misusing the doomsday argument.

Adam and Eve really want to ... you know, hook up.

Except they're afraid of God's wrath if Eve gets pregnant.

The serpent comes along and explains that according to the doomsday argument the chances

of Eve getting pregnant are nearly zero.

After all, it's incredibly unlikely that Adam and Eve are the first two out of billions

or trillions of future humans - therefore, odds are, they can have all the fun they want without

risk of spawning an entire species.

But an even better doomsday absurdity was from thatisjustgreat who says "I was 30 seconds

in when I realized the video is probably almost over".

By that logic, this video is only half way through.