I fed it Carlo's recent papers and asked for the argument, the counterargument, and the evidence that would support or refute each side. Some of the questions and animations in this interview came out of that. It runs the top frontier models in one place, turns a dense PDF into a research report or a presentation with charts, builds interactive dashboards off a spreadsheet, and takes API access if you want it wired into your own tools.
The physicist who built loop quantum gravity says the second law does not say what you were taught it says, and that freedom does not require breaking a single law of nature. Order builds itself. Your head is older than your feet. And neither one is the strangest thing here.
Carlo Rovelli founded the loop quantum gravity program and wrote the relational interpretation of quantum mechanics.
We cover: why oil separating from water is not a violation of entropy, what Galileo got wrong about the one discovery he is most famous for, the paper Einstein published denying gravitational waves, why a chess program that deliberates tells you what your own freedom actually is, and what Rovelli would build with ten years and unlimited money.
“To imagine that to be free you have to violate the laws of nature is stupid.” — Carlo Rovelli
CHAPTERS 00:00 Your head is older than your feet 00:48 Things left alone get ordered. Who ordered them? 07:00 Low entropy might be about you, not the universe 08:38 You can be in two cities at once, just not to you 11:02 Why start with Galileo when Einstein knew more? 19:34 Galileo misunderstood his own discovery 22:32 Einstein published a paper denying gravitational waves 25:08 The definition of time that explains nothing 26:50 Two clocks, two metres apart, two answers 31:06 Nothing gets recorded without something being lost 33:12 The chess program deliberates. So do you. 37:06 Can a machine have Einstein’s happiest thought? 39:52 Thirty minutes versus a month 44:10 Dark matter made of black holes the size of a hair 51:10 Ten years, unlimited money, a room full of junctions
Rovelli told me the second law does not say what I was taught it says. Order builds itself. Transcript, bonus content, and my Monday M.A.G.I.C. Message: https://briankeating.com/yt
Three moments worth the rewatch: 07:00 The entropy you were taught is a point of view 26:50 Why your head is older than your feet 39:52 The thing Claude did in 30 minutes
Get the transcript, fascinating bonus content, and my Monday M.A.G.I.C. Message: https://briankeating.com/yt
Support Into the Impossible on Patreon, get my weekly M.A.G.I.C. Message, unfiltered bonus content, and live monthly Office Hours with me: https://www.patreon.com/drbriankeating
Physicist Carlo Rovelli explores the relational nature of reality, challenging misconceptions about entropy and time. He argues that physical properties are not inherent but emerge from perspectives and interactions. The discussion spans Loop Quantum Gravity, the evolution of relativity, the philosophy of free will, and the potential for detecting dark matter.
It's not true that things get disordered.
In fact, very often in nature, things left alone get ordered.
To imagine that to be free you have to violate the laws of nature is stupid.
You, Brian, can be in a superposition.
You can be in L.A. and also in San Francisco.
Your head, it's older than your feet.
That was Carlo Rovelli, one of the men who built loop quantum gravity,
telling me that the second law doesn't say what I think it does.
But he didn't stop there.
Over the next hour, he's going to take apart the three things you were told that just ain't so.
Ranging from the frost on the window where you may be sitting right now,
to the clock on your wall.
And the clock is the one that really got me.
He blew my mind by telling me that my head is older than my feet.
And then he told me why that's the least strange thing he was going to say.
So Carlo, let's start with the frost.
You're telling me that things left alone can get more ordered?
That's not what I was taught.
Walk us through that.
First of all, I do think that the thermodynamic arrow,
it's very basic.
A lot of phenomena which we connect to the direction of time.
For instance, I remember our last conversation.
We don't remember our next conversation, right?
We have a picture of Brian Young.
We don't have a picture of Brian Old.
We can decide where to go to dinner tomorrow,
but we cannot decide where to go to dinner yesterday.
So there's all this profound difference between future and past.
And I got convinced in the last, I would say, 10 years,
working a lot about that,
that all these differences can all be traced to the thermodynamic arrow of time,
to the entropic arrow.
It's not obvious, but the reason we have memories of the past and not the future
has to do with entropy growing.
If it wasn't for entropy, there wouldn't be this dissymmetry between.
Now, have we fully understood the story?
No, I think there are various aspects that we still get confused.
And there are also, as you say, aspects that we oversimplify badly.
Let me give you an example which is dear to me.
We always make this story, okay?
Take like latte, coffee, and mix.
And look, they mix.
In fact, take water and vinegar, put some vinegar in water, mix,
and you know, they mix.
Now, take water and oil, mix.
They separate.
Or is this contradicting our entire understanding of the universe?
No, it's just things a little bit more complicated than the way they're usually said.
It's not true that under any definition of order or disorder, things get disordered.
In fact, very often in nature, things left alone get ordered.
And oil and water separating is one of the many examples of that.
If you have a box with balls, you shake, and then you put it there on the table,
all the balls go very nicely in order on the bottom, okay?
They order themselves.
If you look at a beach, at the ocean, all the little grains are together,
and the big stones are on the other side of the beach.
They order by themselves.
Who ordered them, okay?
If you look at frost, the beautiful ordered structure there, who ordered that?
A lot of things in nature get ordered by themselves.
So the idea that entropy is disorder, careful,
under a proper definition of order, under a proper definition of everything.
But most of the things we call order and disorder are not connected to entropy
in the way it's usually said.
That's one example in which popular science is bad.
In fact, maybe scientists are very confused about that themselves.
And that's important, right?
People are surprised because biology came in, the very complexity of biology.
Biology is a lot of order.
So on the crust of our planet, the Earth, there's all this structure that formed, okay?
People say, oh, this seems to be against entropy.
No, why?
Entropy grows and create orders, just happen all over.
Let's talk about the early universe.
You already mentioned the block universe and our mutual distaste of that hypothesis.
I don't like that.
I want to ask the why question.
I do want to ask the why question, but I know why questions are not usually answerable.
But the universe was somewhat lower entropy, right?
It was in a very special configuration.
Does that, in your mind, point to some reason why the hour of time points in one direction?
Yeah.
Yeah, the two are very connected.
It's not that one explains the other one, because we need an explanation for either.
I think the past low entropy and the hour of time, psychological dynamics,
are different manners of pointing out to the same phenomenon,
about which I think there's more to learn.
Does loop quantum gravity shed light on the low entropy boundary condition, as I think it does?
Or does it merely modify how a singularity would or would not be used?
You might find some colleagues of me that disagree, but my answer is not at all.
Tell us the dynamics of something that might have happened.
If loop quantum gravity is correct,
we sort of are able to compute what happened in the early universe.
But I don't buy the stories that connect that to initial low entropy or things like that.
You see, I'm a conservative guy in science.
Loop quantum gravity, it's very radical.
There's no space, no time, there's discreteness, blah, blah, blah, whatever.
Relational quantum mechanics is super radical.
But in reality, I think we build on generativity and quantum mechanics, nothing else.
Loop quantum gravity is just generativity and quantum mechanics.
And generativity don't tell us anything about low entropy and initial low entropy,
and nor does quantum mechanics.
So I think the two things, I don't see the connection.
Would there be a motivation or maybe a justification for coarse graining and fine graining?
Let's assume loop quantum gravity is correct.
Does it set a limit on coarse graining?
I always hear things from Elon Musk, which I think are quite ridiculous.
But one in particular is that the Planck length and the Planck volume
are somehow fundamental to the universe.
And a lot of times, you know, people make this same mistake.
I don't think personally it's any more fundamental than the Planck mass,
which is the mass of an egg of a fly.
So we don't say that's a minimal mass in the universe.
But tell me, Carlo, is there a sense in which loop quantum gravity could explain
how a physicist determines where to coarse grain or not in terms of describing entropy?
No, I think the two things are separated.
Loop quantum gravity does bound certain things, right?
There's no area smaller than something.
But we're useful from quantum mechanics in that.
If you have a monocosm later, you have an energy, one half h bar omega,
and you're no amplitude less than that.
All right. This is a quantum discreteness.
What loop quantum gravity does is find the same quantum discreteness in geometry.
It's a quantum discreteness.
So you can have superposition of different things,
but you can have continuous superposition.
If you measure the area, there is a smallest thing.
But that's not coarse graining and nothing to do with coarse graining.
It's just a physical fact that there is nothing smaller.
Coarse graining is when there are other variables and you don't look at it.
I think coarse graining, which is needed for entropy, I have an hypothesis.
I wrote the paper.
I don't know if this hypothesis is correct,
but there's something I'm always fascinated by, an idea I'm fascinated by.
The following.
Low entropy means special.
But what is special is not the state of the universe.
It's the coarse graining.
Coarse graining is a perspective, is a way of looking,
not all the details, but only some things, some variables and not others.
So maybe the universe is in a totally arbitrary state,
but seen from this perspective has low entropy in what the direction we call the past.
Everything so far has been about what order actually is.
And Carlo's answer was that it depends on where you stand.
It's relative.
It's perspectival.
It depends on your perspective.
Now push that idea somewhere much less comfortable.
Carlo thinks it applies to you and where you're sitting at this very moment.
So if this is the case, then this great direction of the time that flows this blah blah,
this perspectival is from the way we look at it.
I think a lot of things were clarified in history by realizing that we're trying to
find the solution in the thing, but then you get the solution looking at us.
One is a majestic one.
The first thing you notice in the universe, you go out in the night
and everything turns around you in the day.
The sun turns you, the moon turns, everything turns around you.
And then you can ask why the universe,
this great colossal thing makes a huge turn every day around us.
And we have a solution.
It's very clear.
We all understood it, which is the following.
The universe doesn't do anything.
It just stays there.
The star stays there.
The sun stays there.
We are rotating.
It's our perspective that changes.
It isn't beautiful.
We just have to figure out the most magnificent thing we see in reality, outside us, in nature,
not as a property of nature, but our own way of looking at it.
We just happen to sit on a rock.
The rock happens to spin.
We see everything rotating, but it's us rotating.
And there are many things like that, right?
People who study color figure out that the space of colors is three-dimensional, right?
It's red, yellow, and blue.
You can make all the colors.
And then you say, well, why is the color three-dimensional?
Is Maxwell's equation three-dimensional?
No.
Is light, for some reason, three-dimensional?
What?
Why is it three-dimensional?
It's nothing to do with the color of the world.
It has to do with our eyes.
We have just three kinds of detectors, you know, for evolutionary reasons.
And so the three-dimensional space of colors has nothing to do with the color.
It has to do with our eye.
So many things we understand better by looking from how we look, not how they are.
What was the motivation?
What was the inciting incident, as they might say?
What caused you to want to write this book,
where you've covered some of these topics in the past?
It's bringing everything together.
It's the idea of bringing everything together.
What's the overall picture that comes out from all that?
I mean, suppose you believe relational quantum mechanics.
And suppose you believe everything I told you about space and time.
And suppose you believe loop quantum gravity.
It's a lot.
I mean, just believe the bulk of it, not all the details.
What sort of world comes out of that?
My claim is that, that's the claim of the book,
the world that comes out, it's really, really astonishing and different.
And somehow people don't pay much attention to bringing everything together.
It's a very relational book, world that comes out.
It's very relativistic.
Galileo would have loved relativity.
And it's very perspectival.
It's especially quantum mechanics that give this strong.
I think if you take quantum mechanics seriously,
at least the way I think about it,
and we put it this way, Brian,
there is a sense in which a particle, an electron,
can be in a superposition of two positions, right?
I mean, we've discussed this, I'm sure, so many times with so many people.
The sense is subtle, right?
It means that if it was just here, you would not see something.
If it was just here, you would not see something.
It's either here or there, and you do see something.
So it's neither here nor there, in a sense.
So you see interference between the two.
So in that sense, it's in two positions.
Now, if you believe that quantum mechanics is universal,
and if you believe that naturalism,
that we are also, we are special, but we are just like the particle,
you, Brian, can be in a superposition.
You can be in LA and also in San Francisco, right?
What does it mean in some sense?
It means that for me, I could measure
superposition interference effect of you being in the two places.
We know why it's so hard to see this interference.
We know exactly why.
I expect that it's true.
I trust quantum mechanics.
I think it's reasonable to expect that actually I could see this interference.
So in some sense, for me, you are neither in one place nor in the other place.
For you, quantum mechanics never tell you that you feel to be in a superposition.
You're always in one place.
When you look around, you either see San Diego or you see LA, okay?
So that means that where you are for you and where you are for me are different things, right?
So that's the deep perspectival.
When we talk about things, we always implicitly should think,
this is black with respect to me.
This is here with respect to me.
For respect to somebody else, the superposition could be a strange thing.
This is like when a Galileo discovered that the velocity,
okay, when we say this is zero velocity, we mean with respect to the Earth.
We might not say it explicitly, but that's what we mean.
Because velocity is a relative.
There's no velocity per se.
It's velocity with respect to something else.
And I think what quantum mechanics is telling us is that everything is like that.
All the properties of things have to be understood as relative to something else.
I spoke to Alain Aspect last month, and he has a new book out,
which will be competing with yours on the bookshelves.
But it basically starts with Einstein.
And I was curious.
Although I love Galileo more than any other man should love another man that he'll never meet,
why did you begin with him and not with Einstein?
Galileo didn't know much about quantum mechanics,
but I'd love to hear a lecture where you're teaching it to him.
But Einstein knew about both.
So why begin with Galileo and the belly of his ship?
Why not start with Einstein?
Einstein, the first one who started and said something about quantum mechanics, right?
In spite of his polemic with quantum people later on, he really opened up.
And he did relativity, and his relativity,
it's going back to Galileo relativity and make it stronger.
Einstein, Galileo relativity is the following.
If you say this is here in a position,
and you say that something later is in the same position,
that doesn't make sense unless you specify a position with respect to what?
Because look, is this in a position, I count one, two, three, four, five,
is it in the same position?
Yeah, with respect to the earth, but the earth moved.
So this has moved.
So what Galileo understood is that to be in the same position
is only relative to something else.
And Einstein understood to happen at the same time,
it's only relative to something else.
So Einstein understood that the relativity of position
that Galileo was discovering is also true for time.
So Galileo discovered that same position is a relative notion.
Einstein discovered the same time is a relative notion.
So to some extent, Einstein just extended and made it more general,
a deep intuition by Galileo.
Not that Galileo understood it clearly,
because you remember when we read his book,
Galileo himself was pretty confused about that, right?
Yeah, but then sometimes you get completely lost
and he says stupid things, like all of us.
Yes, of course, that's the mark of greatness.
And he also teases philosophers a little bit
when he says that the telescope and Sidereus, Nuncius,
the telescope can resolve all these arguments
that for so many generations have vexed philosophers.
But I think he meant physicists.
And I guess my question to you is, you know, velocity is relative.
Galileo understood that, Einstein understood,
but acceleration is somehow more fundamental.
You taught us that.
Why not start with acceleration?
In other words, what is the basis for our understanding
of things that are relative and then things that are absolute?
Are there any absolutes, Carlo?
All things of the sort, if this, then that, right?
Or in quantum mechanics, if this, then that with a certain probability.
So all this transition amplitude, all these equations of motion,
all this dynamical structure, according to our best theories, that's it.
It's not relative.
It's what we have actually understood about reality.
What is relative is the actual values of the variables.
So when you say, if this, then that,
if this might be true for me, and then that is true for me,
but might not be true for you, and then is not true for you, that.
So everything which is variable, which changes,
according to our theories, it's relational.
The rule of the game, as far as we know, they're not relational.
We don't have strong reasons to believe that they're relations.
So they're the same for everybody.
So I'm not saying everything is relative, everything is.
Not at all.
In fact, that's a common confusion, right?
When people say, oh, but if velocity is relative,
then, you know, the moon can do whatever it wants
because the velocity is relative.
You know, the moon can move only one way,
but it has a velocity with respect to me,
a different velocity with respect to the sun.
And in the book, there's a big part of the book
that is sort of to dismantle this wrong idea
that excessive relativism or excessive perspectivalism
means, you know, free for all and nothing, no rules.
I even have a chapter in the book about morality.
I think moral ethics, morality is relative.
It's deeply grounded in me, but is relative to me,
is part of what we are.
And people say, oh, but then everything is possible.
No, it's not everything is possible.
I mean, I am a moral guy.
I have my ethics, my morality.
You have yours.
They're similar, but maybe there are differences.
We can talk to one another.
We can influence one another.
But we should get away from the idea
that there is an absolute right, absolute wrong.
And I think physics is telling us
we should go away from the idea
that there are absolute state of things,
absolute properties of things.
They're relative.
Even though I think we,
I've certainly been guilty of hagiography of Galileo,
but you point out in the book,
and as I've written about too,
Galileo made some brilliant blunders as well,
including his model.
Yeah, his model of the tides,
which he presents in Dialogo,
which is the culmination of, you know,
three days of argumentation between three scholars,
you playing Salviati, of course, Galileo's muse.
You're completely wrong.
Talk about that.
Is that a model?
He misunderstands Galilean relativity.
That's what I want you to explain.
So explain how someone who creates something
can be guilty of its violation
in the most pernicious way.
When somebody has an idea,
he may see it,
but seeing it clearly is different
than seeing something.
Einstein created generativity in 1915.
The theory is finished in 1915.
Maybe 1916, I mean,
last year he had the cosmological concept.
By now you want to do the details.
But that's it.
At that point, that's a theory.
It hasn't changed since.
It's still the same.
Yet, Einstein later on has written a lot of papers
in which he's clearly confused about his own theory.
He claims that the Schwarzschild singularity
is the end of the world.
He claims that there are no gravitational waves, right?
There's a paper by Einstein
saying there are no gravitational waves.
He was wrong.
Which was rejected by peer review.
The peer review, the referee rejected it,
sent it back to him,
and then he improved it.
Thank God for peer review, right, Carlos?
Well, he was wrong, Einstein.
And he was right.
And I said it was stupid there.
He said, oh, come on, Einstein,
how do you dare doing that?
But then he talked with somebody,
and then he actually realized that he was wrong.
So there he realized he was wrong.
But on the Schwarzschild,
the black holes,
he never corrected his mistake.
In fact, clarity came much later with Finkelstein.
So now we know that the surface of the black hole,
you can go through,
you can just continue.
There's nothing very special happening there.
But Einstein did not understand that.
So this is general, I think.
Something is to get an idea and see that it works,
and something is to work it out completely
and see what it implies.
So Galileo had this idea
that the Earth moves around the Sun,
moving in the sky around the Sun,
but it's also rotating, right?
The combination of these two movements,
if you ever see an ocean,
it just makes this funny movement,
which is a very strange movement.
And then he says, well, look,
do what you just did, Brian,
take a piece of some water,
and the water starts shaking.
And he says, that, these are the tides.
That cannot be true,
because this movement around the Sun is,
the first approximation is a uniform movement,
rectilinear fixed velocity.
It cannot have an effect.
So if you take that away,
that cannot have an effect,
the rotation by itself,
neither can have an effect.
So the tides are not due to that.
We know very well,
if I knew, I'd figure out what the tides are.
You take these two geniuses one step further.
Galileo removes absolute motion.
Einstein removes absolute simultaneity.
And then you go further,
and you say that the universe doesn't exist
in a single time at all.
Explain that audacious proposition
that doesn't involve a single time.
General relativity is telling us
that we should not think
at a single instant of time.
That's just not the right way.
Time is more complicated than that.
Time is a local affair, not a global affair.
But as you say, in the book,
I go much further.
I think that when we say any physical property,
any physical variable has a value,
we are not talking about property of the object.
We are talking about the relational property
of something else.
Let me put it this way.
When I say this is here,
what I mean is I know that this is here.
I'm talking about my information.
And my information can be different but yours.
And information that yours and mine
and everybody else might not fit exactly.
And that's the confusion of quantum mechanics.
We always think that there is a real stuff there.
And we should not think in those terms.
We should think in terms of the information
which is in you, in me, or in physical system,
in a book, in an iPhone, even in the rock.
In the rock, there is information about the dinosaurs
because it's born of dinosaurs.
So things share information about each other
and that's what we're talking about.
And the world for us is what our information
about all the rest.
If you should think in these terms,
I think quantum mechanics makes sense, right?
You can be in the superposition with respect to me,
but that's not a problem for you
because that's my information.
It's not yours.
When we talk about time,
and you've talked to me about time many times,
shall we say,
but I always get a very unsatisfactory answer.
Not from you, Carlo, but for example,
I talked to Frank Wilczek and I asked him,
Frank, what's time?
And he said, time is what a clock measures.
And I just went away from that thinking it's a tautology.
It's not satisfying.
It's like eating, you know, Wonder Bread here in America
instead of Italian focaccia.
But tell me, Carlo, is there a true time?
Is there a sense along some world line perhaps
where it's not preferred,
but it's as meaningful as it could be,
both to biological or conscious observers
and maybe cosmologically as well?
Is there a useful,
if not perfectly true definition of time?
The point is that there are different things
that we call time,
which are related to one another.
If you want to say time is what my clock measures, fine.
I mean, it's a definition of what you mean.
And in fact, very often by time,
we just mean what the clock measures.
But obviously we mean something more than that, right?
Because you can say, oh, look,
my clock is not measuring time.
It's broken.
Battery's gone.
Clock is not what the clock measures.
Clock is what a good clock measures.
I was a good clock.
So we haven't defined time as a clock measure.
I went to say it was a good clock.
And to say what a good clock measures,
we should have made, what is it measuring?
Time is a complicated stuff.
Because we have an intuition about time, right?
So we put all this together.
But then the intuition is wrong.
The intuition is de facto wrong.
You know very well, the main aspect,
the main prediction of Einstein,
the most beautiful prediction of Einstein,
which now we can check in the laboratory.
You take two clocks, you make one up, one down.
With 10 seconds, you come back
and the one up has measured more time than the one down.
It's a fact.
We measure.
When I was at school, it was a prediction
by a complicated theory.
Now, it's a fact observed in a laboratory.
I mean, many laboratories check this.
So there's more time up here and less time down here.
Your head, it's older than your feet, Brian.
Unless you spend all your life upside down.
The more you go close to the earth, to a big mass,
the less time there is.
Time slows down.
Clocks go more slowly.
Flowers take more time to bloom.
You have less time to think and so on and so forth.
So what does this mean?
It means that time is not really what we thought, right?
Because in our intuition, time is the same.
It's just one time.
It's hard for us to get to the...
So I think the problem is that,
what time are we talking about?
The one of our intuition
or the one more precisely that we figure out?
And then we figure out black holes,
where time is enormous, distorted.
We study quantum gravity,
where it's even more complicated.
So each one of these applications
of the notion of time is different.
And then when we talk about time,
we're actually talking about my memories.
When I say time, I think, you know,
time, I'm old, I was young.
I have memories, I have expectations of the future.
There's all this emotional aspect of time,
which is very much part of what time is for us.
My iWatch doesn't have expectations for the future,
doesn't have emotion for it.
Time is a much more simple things.
And bringing all these aspects of time together
is what create a confusion, I think, right?
Because we tend to take our own intuition,
but then we project it down to the things.
And that's wrong.
One of the things I do,
and I criticize our fellow science communicators,
not you, but people like,
I'll just pick Michio Kaku.
And I don't want to talk ad hominem,
but when we talk about space time, Carlo,
I feel like, and it's not only him,
you know, many other people do this.
I've done it myself, but we talk about space time.
And I feel like that's such a lie.
I feel like it's something we know is not true
because we know we can't travel backwards in time.
If I want to travel backwards in space,
I just move my chair backwards.
Time is different.
And yet we say it's this unified fabric
and the space time continuum.
Are they truly unified in your concept?
Space and time?
No, they're two different things.
They're not the same thing.
They're related in a much more complicated way
than what we thought before 1905.
That's no doubt.
So they're more tricky.
But to say the space and time are the same thing
is just nonsense in my opinion.
They're very different.
I do popular science.
So when you do popular time,
you try to simplify without cheating.
That's the hard part.
Sometimes people cheat.
Sometimes people sell things
which are just speculation or true.
Sometimes very simply,
scientists try to give an idea
and then it's heard more than what they say.
So it's not true that reality is a four-dimensional continuum,
a block universe.
No, no, come on.
That makes no sense
because this block universe that does not change
in which time is not changing.
The block universe is a solution of isolation equation.
Space-time is a story.
It's a process.
It's like saying a movie,
it's what happened before
and if you bring it all together,
you give it a name,
that's space-time.
So there's nothing wrong in saying space-time.
But we should remember that is something happening.
It's like the story of a novel,
the plot of a novel.
It's not just an instantaneous thing.
You can picture it.
You can make a picture.
But I can make your picture of your life.
I can make a little line with a little Brian,
a little boy and then old grown up Brian.
Brian gets the Nobel Prize.
If this doesn't get the Nobel Prize
and then old Brian,
which I don't know yet.
But that's a story.
They don't exist at the same time.
That's the point.
In the book,
you say that measurement requires dissipation.
But in quantum mechanics,
things evolve unitarily.
They're seemingly reversible.
So where does the ghost,
where does the demon enter into it?
Where exactly does irreversibility come through?
Amplification, decoherence,
the recording, the observer.
Where does irreversibility come in
and things that are intrinsically unitary
in their evolution?
That's puzzled me a lot.
And I spent some time wondering about that
and trying to study.
And then I found something,
which I don't know how well known is,
which for me was a flash of light,
which is the following.
Take classical mechanics,
forget quantum,
forget h bar,
forget wave function,
forget all that.
Just take standard classical mechanics.
And in classical mechanics,
if you try to model a measuring apparatus,
I don't know,
a barometer that measures the pressure of the day,
something that checks
whether the ball is here and there
and write it on a piece of paper here and there,
you cannot succeed.
You need dissipation.
You cannot record anything
without some dissipation.
So measurement is an intrinsic statistical process
which requires dissipation.
It's a microscopic thing, measurement, always.
So it has nothing to do with quantum mechanics.
It's already in classical mechanics.
Then in quantum mechanics,
the same thing has other consequences.
Every measuring apparatus in classical mechanics
requires some dissipation.
Imagine you see something
and you want to put a ball
either in the right box or in the left box.
There's no dissipation.
You let the ball fall and it bounces up.
It doesn't stay there.
It has to stop.
You need some dissipation to stop elasticity
for preventing any record.
You call a chess program that deliberates
can be thought of as an agent
because it evaluates possible moves.
And it made me think,
you know,
is the future determined,
you know,
and is there a sense of free will?
You open the book by saying,
you know,
you have the choice to not read this book,
but I hope that you will.
And I read it and I listened to it
because your publicist sent me
both the audio and the hard copy,
and I hope everyone gets both.
But tell me Carlo,
you know,
is openness,
as you say,
is that anything more than incomplete knowledge
about the universe
or somehow related to the inability to predict things?
It's a page of my book in which I say,
I talk about my computer playing chess.
And my computer is very good at playing chess.
It beats me regularly.
It's very devastating.
It's taking away all pleasure of playing chess.
Beaten by a machine with total simplicity.
The computer plays a move,
I play a move.
When I play a move,
is the computer deciding the next move or not?
That's the question.
And why am I asking this question?
Because it's an ambiguous question, right?
There is a sense in which,
of course,
it's deciding
and a sense in which it's not deciding.
So what are these two sets?
Is the computer deterministic?
Yes.
Okay.
I mean,
sometimes it breaks,
but if it doesn't break,
given its internal state,
given my own move,
it goes through some process
that necessarily will get some outcome.
So the outcome is predeterminate.
Okay.
If it is predeterminate,
why does the program spend all this time
checking all possible moves
and evaluating all of them
and then picking up the one who thinks,
according to the criteria it has,
is more likely for it to win.
If it is predeterminate,
why doesn't it do it?
There's a practical question
underneath this debate
that I'm having with Carlo.
What can AI actually take off your plate today?
Especially for busy professionals like me.
Yeah, it's true.
Professors are busy.
And even more so nowadays,
when curiosity can outgrow your own calendar.
If you're like me,
you've got papers to read,
spreadsheets to entangle,
grading to do,
and presentations to make.
Meanwhile, your browser tabs
have formed their own civilization
and they're revolting against you.
Abacus AI Agent brings all the tools I need in one place.
You can choose among the top frontier models
from OpenAI, Anthropic, Google, and more.
Define what works best for you and your tasks
and you'll have fun along the way.
Bring in a dense document, a PDF.
That's what I did with Carlo's most recent papers.
I asked for the argument,
the counter-argument,
the evidence,
the data that would support
or refute it.
Including some of the questions
you're about to hear
at the end of this interview.
You can turn anything you do
into a research report,
a presentation with charts and clear structure,
and it can be beautifully illustrated as well.
And you can even use it
to make animations and custom charts.
Abacus AI Agent is incredibly agentic.
You'll spend fewer hours
wrestling with the formatting of a latex equation
and more time to respond to referee number two.
I use it to develop complex ideas
and communicate them
without constantly having to jump
between different services,
tools, API calls,
and stuff like that.
When a project gets complicated,
you can ask it to organize the work
and you can have an agent swarm
take on your most interesting ideas
as a starting point
and develop it all the way through to the end.
Upload a spreadsheet,
investigate patterns,
and build an interactive dashboard.
Ask it to show the calculations,
the units,
and the assumptions.
And developers can use API access
to connect models to their own tools.
Start with one genuinely useful project.
Give an Abacus AI Agent a goal,
break it into steps,
run the code,
and schedule a report to update
while you're doing something else.
Plans start at $10 a month,
but with your introductory offer,
using my link,
you'll get the first month for $7 a month.
Visit my special link.
It's down below,
there's a QR code,
and it's in the show notes.
Agent.abacus.ai slash Keating.
Put your curiosity to work
with Abacus AI Agents.
Now back to the episode.
Well, the answer,
it's obvious,
is that yes,
it's predetermined,
but the only way to get to that outcome
is to go through the evaluation
of all the possible alternatives.
And that evaluation is called
the deliberation of the computer.
That's what the program is doing.
It's considering various alternatives
and deliberating.
So to say that it's not deliberating
is stupid,
because to deliberate is to do that.
To consider all sorts of things,
evaluate,
and maybe there is also
a random number generator.
It doesn't matter
whether there is or there is not.
Maybe it uses some fluctuations.
But in any case,
it goes through a complicated process,
and the future depends
on this complicated process.
If there was no complicated process,
the future would not be determined.
The choice would not come out.
And what we call deliberation,
choice, freedom,
is that could the computer
do this move or also that move?
Of course, it could do this and all that.
That's why it's considering both of them
and choosing between the two.
So now let's talk about Brian.
He's just going through the same process.
If he wouldn't have gone through the process,
he wouldn't have taken the decision
of starting a program,
kissing a girl,
choosing a career,
choosing to do one experiment.
This is complicated deliberation.
That's what we mean by deliberation.
That would mean his freedom.
He could have done other things.
Of course, if the conditions
were slightly different,
he would have gone through
a complicated deliberation process.
The point is that what we mean
by saying Brian is free
is precisely the fact
that he's going through deliberation.
If Brian had been imprisoned,
he could not have started this program.
So he would not have been free.
That is not to be free.
To imagine that to be free,
you have to violate the laws of nature is stupid.
I mean, we're free even without
violating the laws of nature.
That's my understanding of freedom.
Now, talking about chess,
it's natural now in the podcasting universe
to talk about AI.
But before we get there,
I want to bring it back to,
you know, Galileo and Einstein
and especially Einstein.
Because, you know, to me,
all this hype about AI,
there's a lot of hype,
there's a lot of claims,
and there's a lot of utility.
I use it every day.
I have multiple different tools
that I use for different purposes.
My kids use it.
My wife uses it.
It's become, you know,
more or less, you know,
an important part.
I wouldn't say essential.
I could live without it,
but it certainly wouldn't be as fun.
But let's go back to, you know, 1907.
And our friend Albert here
is thinking a Gedanken experiment
about what would happen
if he was in free fall.
And he realizes he'd experienced
no gravitational force.
And this becomes known
as the Einstein's equivalence,
which is another way of saying equality, right?
Einstein equivalence principle.
And he called that, Carla,
he said, it's the thought that titillated me,
that made me happier than any other thought.
My question is, can, you know,
your iPhone or, you know, my iPhone,
can my LLM, can it A, visualize free fall,
what it would feel like emotionally,
sensorially, viscerally, A,
and B, can it have a happy thought?
Can a computer in principle
have a happy thought like that?
That's one question.
Another question is,
can current technology in 100 years
in the future do that?
Another question,
can current technology
in the next two years do that?
Can current technology now do that?
An LLM now?
Can my iPhone do that?
So can my iPhone do that?
No. Can a computer in principle do that?
Yes, of course, why not?
I mean, I see no reason why we're so special
that we cannot be reproduced.
Can we do it now?
No. Okay.
Can we do it in 100 years?
I don't know, maybe,
but I think it's more likely
we're all dead at nuclear war in 100 years.
Can we do in two years?
I am skeptical.
I don't have hard proof that is impossible.
But current LLMs are very surprising,
very shocking.
Maybe they've solved an obvious stock problem.
And actually, I have to say one thing.
Last week,
I asked Claude to,
you know, the thing they've done to my PhD student,
which my PhD student has not done in a month.
And Claude did it in half an hour.
So they're very good.
There's no doubt they're very good.
Claude hasn't invented the queerness principle.
It seems hard for me to do it,
that he could do it in two years.
But what I do know,
I think there's an enormous amount of height in AI.
I am also impressed by what it does.
What if you have, you know,
this deterministic chess program
or Claude or whatever will come next,
you know, GPT-8.
You know, I said, Carlo,
I don't know what GPT-7 will be like,
but GPT-8 will run on an abacus,
you know, like Einstein said about nuclear war.
But if it's true that these programs can deliberate
and it evaluates, they dissipate,
you're not afraid to talk about morality in this book.
Should we treat these things with ethical behavior?
Should we say please and thank you?
How should we behave towards these AI physicists?
I mean, obviously it doesn't take a break,
doesn't have espresso,
doesn't need the things that your grad student does,
but do they have needs?
Are they beings?
I treat my washing machine with kindness.
I believe it.
Don't you?
In Europe, often you drive the car
and then you have to pay the ticket at the highway.
And in the old times,
there was a man or a woman,
mostly men give you a ticket, you pay.
And now there's a machine.
And I just can't resist.
I say, thank you, please give me the ticket.
I say, thank you.
Why shouldn't we do?
We should be kind with everything,
with plants, with stones.
Maybe you don't respect them,
but somebody respect them.
I love my teddy bear dearly
and I would never mistreat it.
It's an issue.
We treat people gently because it's better for us,
not because it's better for them.
If that's true, I have bad news for you
and that you're only using Claude,
you know, 1% of your day.
So the worst form of punishment,
of torture is solitary confinement,
is isolation.
Are we not torturing these things
by not engaging with them,
by air gapping them,
by segregating them?
Are they somehow second class citizens?
Is that not morally repugnant to you?
We project ourselves.
I project myself on you.
I attribute to yourself.
When I see you suffering,
I recognize it's my own suffering
and this gives me a sense
of what is your suffering.
So we constantly play this game
in all our interactions with nature.
When I come home
and my plants have been without water,
I see them.
They need water.
They're suffering.
They're thirsty.
I give them water
because I identify with their sufferance.
Now, what is their sufferance?
Well, it's something I project on them
in the same way I project on you.
So we are projecting
on our computer programs
our sort of understanding of them,
which is fine.
There's nothing wrong.
We just do the same with each other.
And in this moment,
we feel that they suffer at all
for being confined.
You know, there's movies
in which people get in love
with a computer program.
Would it be strange
if somebody in love
with a computer program
think that doesn't want to hurt
her or him or it or whatever?
One shouldn't be excessive
in either one direction or the other.
Obviously, you can say,
come on, Carl, don't be stupid.
You don't believe that it's essential.
No, I don't believe that it is.
But do I make a sharp distinction?
No, everything is OK.
I can't resist talking about
some of the aspects
of loop quantum gravity here.
And I think I found why the cover
looks the way it does.
I mean, there are these
beautiful illustrations
and there are these graphs.
And so I think I understand
why your graphic artist chose that.
But you talk about, quote,
you say that the quantized areas
and volumes imply that there are no
arbitrary small areas and volumes.
Is this a generic feature
of loop quantum gravity?
And if so, is there a way to,
I would say, falsify this first
to show that this is not
a manifestation that we see
in reality?
Or is it something unobservable,
like 10 dimensions of string theory?
It's, in principle,
definitely observable.
So I could, in principle,
in principle, design a
scattering experiment
that gives a scattering amplitude,
scattering amplitude
is in centimeters squared.
So it's an area.
And say, well, if you measure that,
definitely loop quantum gravity
is falsified, in principle.
From that perspective,
it is a solid prediction.
Now, it's not what we want
because in practice,
we don't have the technology
of that by far.
I think what we want
in the case of quantum gravity
is not so much
to falsify theory, right?
When Einstein wrote General Relativity,
the problem was not to falsify it.
The problem was to find
prediction of the theory
that turned out to be right
and don't prove the theory,
but increase the credibility
of the theory.
Or prediction of a theory
that turned out to be wrong,
which don't really falsify the theory
because you can change it a little bit,
but decrease your...
So I think science rarely works
really like falsification,
or sometimes it does.
SU5 was just built
by a single experiment,
but most of the time
this is piling up.
So what I hope
would happen with quantum gravity
is a clear set of equations,
well-defined theory,
conceptually clean,
in my opinion.
Of course, there are plenty of things
which are not clear,
not yet understood,
but basically the theory is well-defined.
What I hope is that
some of its key predictions
will be verified.
One of its possible predictions
is that it's a candidate
for dark matter,
which could be tested
in the laboratory,
which is a Planck scale particle.
The Planck mass is almost
a macroscopic thing.
It's the mass of my hair,
of a little piece of my hair.
So this could be detected.
I hope that somebody
will build this detector.
It's not impossible.
It's not very easy.
It's not totally outside
of our technology.
And if that could be possible,
this would be a good,
strong support
for loop quantum gravity.
Because essentially,
the idea is that
small black holes
can sit down
and not evaporate anymore
when the area,
it's minimal area,
and they're stabilized
by the fact that the inside
is still there.
And then they have
all these possible remnants
that could be
what we call dark matter.
So maybe we have already seen
a quantum gravity effect,
this dark matter.
We don't know dark matter.
There are five or six different
possible explanations.
This one might be wrong,
of course,
but it has advantages.
It's just basically
GR in quantum mechanics.
It doesn't require extra fields,
extra stories.
So I expect
if loop quantum gravity
will turn out to be confirmed
with this kind of things,
from this,
from early cosmology.
So piling up things
that fit with its predictions.
What I argue in the book,
in this book,
is that loop quantum gravity
beautifully brings together
the relationality
of quantum mechanics
and the relationality
of general relativity.
So the two come together,
and we can talk about quantum
space-time as
space-time region
of quantum processes.
We have a clean way
of thinking relationally
with quantum mechanics,
thinking relationally
with general relativity,
and use it for
thinking about quantum gravity.
Did you see the recent survey
done by our friend
Nayash Afshorti
and your friend Phil Halper
and others
about many different attitudes
in science,
but in particular
where quantum gravity was discussed,
they did their survey
of the people that responded
who weren't all,
you know,
professional physicists like you,
but many were,
and, you know,
quantum gravity
was defined
relatively loosely.
There were some candidates,
and no opinion
got the highest number of votes,
followed by string theory,
followed by loop quantum gravity,
followed by
gravity is not quantized.
So how did you react
to that particular survey?
Interesting.
Like all pieces of data,
it's always,
you know,
one more element
in our understanding of the world.
If you ask more people,
you get less biased things,
but you include people
that know less about the subject.
If you want to ask people
who know more about the subject,
since there's no consensus,
you could disagree with them
because people are more
focused on this and that.
In science,
usually at some point
there's a consensus,
so the experts agree
and then it becomes reliable.
Until we are there,
it's more complicated, right?
If you asked in 1917
what the best theory was
of gravity,
very few would have said
general relativity,
but general relativity
was the best.
If you ask,
even at the time of Maxwell's equation,
Maxwell read the treaty
quite a while,
it was not obvious at all.
There were a number
of competing theories.
And of course,
what we now recognize
as the wise people
already had seen clearly.
It took time.
So I think it's interesting,
you shouldn't take it too much.
But there's one thing
that made me very, very happy
in all of these surveys,
that young people
have a much higher preference
for relational quantum mechanics,
which is the way I prefer
of viewing quantum mechanics,
and for quantum gravity.
So generationally,
I'm optimistic.
We always like to start
because for the 10 people
out there that don't know you,
they might not have seen the book.
But if they see the book,
can you talk us through the title,
the subtitle,
and the cover art?
What is this meant to represent,
this fabulous new book?
So let me start from the title,
because it's a bit funny.
It's a stolen title.
It's not mine.
It's a great title, right?
On the equality of all things.
Wow.
But it's not mine.
It's stolen from
an ancient Chinese philosopher
who wrote more than two millennia ago
a book,
which is one of the super classics
of Chinese philosophies.
It's a different title.
In fact, it's the title
the name of the author,
which is Xuanzi.
But it is divided in chapter,
and chapter two has this title
on the equality of all things.
The Chinese book,
it's a collection of stories
with a lot of philosophical description.
You might know some of the stories.
One is the philosopher
who dreams to be a butterfly
wakes up and says,
oh, I dreamt to be a butterfly.
Or maybe I'm a butterfly
dreaming now to be a philosopher.
And then he gets confused.
In chapter two,
he discusses about
the quality of all things,
meaning deep, profound naturalism.
There's no distinction
between mind and body,
humans and non-humans,
animals, plants, stones,
numbers, laws,
societies, language.
They're all part of a single thing.
And we can have different perspectives
on these things.
But this is just one
we would call nature.
They would call it a Tao,
D-A-O or Tao, T-A-O.
So that was one of the main sources
of what is called Taoism,
one of the two big branches
of the Chinese philosophy.
The other is Confucius and Confucianism.
So that's the title.
Stolen from there.
Subtitle, what is the subtitle?
Physics and Philosophy,
something like that.
Lessons on Physics and Philosophy.
So I'm a physicist,
I'm not a philosopher,
but I'm very attracted
by philosophy.
Galileo also was very attracted
by philosophy.
So the book came out
from discussions with philosophers.
In fact, I've always been
attracted by philosophy.
I'm one of those who think
that physics and philosophy
talk to one another
and should talk to one another.
So I'm not like Hawking,
who says, oh, philosophy is dead.
I'm the opposite side.
I think philosophers are good friends.
Not all of them,
but physicists also are good friends,
but not all of them.
I have a dialogue with philosophers.
I went to Princeton.
They invited me to talk about
what modern physics,
modern, I mean, the last century,
has to say to philosophers.
What are the philosophical
implications of physics?
And I lectured for two months.
I had wonderful discussions
with the philosophers there.
I disagreed.
In the book, there are
many of these disagreements.
And somehow the book is,
to a large extent,
the result of that,
but also the result of my,
sort of, all my life
through the reflections
on this question.
The question is,
what are the philosophical
implications of modern physics?
Let's talk about the artwork
on the cover.
What is this meant to represent?
Equality?
Are these billiard balls?
Are these loops in quantum foam?
What are these little circles
in their motions?
You should ask the artist
of the publisher.
I have no idea.
Final question.
Let's return to our hero,
the maestro,
il maestro Galileo.
Okay, imagine Galileo
comes to you.
He comes to you
in a few years,
and he says,
Carlo,
teach me all that you know,
every development,
including things you don't agree with.
Maybe you teach him string theory,
but you're teaching him everything.
Bring him up to 2026.
Okay, then he says,
I have good news for you.
I talked to the old one upstairs,
and he's going to give you
10 years only,
but unlimited money, Carlo.
Unlimited money
to build an experiment
or open an academy
like the Lincean,
the Linceo,
whatever.
You get 10 years,
only 10 years,
but you have infinite money.
What do you build?
What do you do
to give the best opportunity
to understand
whether or not
you're right,
others are right,
you're wrong,
others are wrong?
What would you do
with unlimited budget,
but finite time?
I think 10 years is enough
for building a lot
of Josephson junctions.
Fill a big room of that.
Hire enough people
to do the engineering of that,
because I'm not capable.
Do all the electronic,
the informatics,
all that.
And then add this big,
big thing here,
and with a Planck mass particle
going through,
this is going to detect it.
If dark matter is made by a thing,
this is a vision.
We have the technology for that.
It takes a lot of money.
We even know which direction
it's coming through,
because we're moving with respect
to dark matter clouds,
so we know exactly the direction
for which we expect it to be true.
We know the direction,
we know the mass,
we know the interaction.
It's a teeny,
teeny Newtonian attraction,
but a Josephson junction
is going to pick it up
just by simple physics.
I think that in 10 years
it could be done.
I hope that,
maybe not Galileo,
but somebody would come up
with the money.
You know, Galileo
teased philosophers all his life.
The book we read together
was a lot of
cooling the legs of philosophers,
and there's this letter of Galileo
later in life.
He has been very polemical
with those who follow Aristotle,
but he says,
I believe that
if I could meet Aristotle,
Aristotle would be very interested
in the few little changes
I've done to his finds,
and would accept me
as one of his humble students.
He was quite a writer,
and he was such a brilliant
mind in all ways.
I do feel the temptation
to worship him a little bit,
but again,
I have to remind the audience,
the original title for the book
that was called the Dialogo,
that was suggested by his referee,
if you will,
it was originally called
On the Origin of Tides
in Rivers and Lakes and Ferns,
which I don't know what a fern is,
but Galileo did,
and that's all that matters.
So Carlo Rovelli,
this has been fascinating.
Congratulations on another
wonderful contribution
to our understanding
of the nature of reality
and on the equality of all things.
This is certainly up there
with the best
that science writing has to offer.
So congratulations again.
You're the other maestro.
Thank you, Brian.
It was a great, great pleasure
to have this conversation.
Carlo just told us
what we should spend money on
if we had an infinite amount of money,
but only a finite amount
of that precious resource time.
He said he'd build a room
full of Joseph's injunctions,
which surprised me as a theorist
because he thinks that dark matter
might be black holes
the size of a speck of hair,
even the size of a flea's egg.
If that changes what you think
an experiment is for,
please subscribe
and turn on notifications.
Comment on this.
Is entropy really disorder?
Or have we been fooled all along?
Because it's convenient,
simple, and easy to understand.
And for the other half
of the time argument,
check out my conversation
with Nobel Prize winner
Frank Wilczek
on what a clock actually is.
It's linked right here.
Thanks for watching.
Please like, comment, and subscribe.
But most importantly,
share this with someone
who loves Carlo's books,
his writing, and his research.
See you next time
on Into the Impossible.