In this episode, I unpack why today’s AI may be trapped in a “prison of its own success,” why it’s far less advanced than many believe, and whether it truly poses an existential threat to humanity. I also take you inside the science behind the Simons Observatory, revisit the inflation controversy that cost my team a Nobel Prize, and explore everything from aliens and first contact to why the Nobel Prize no longer reflects how science really works.
Key Takeaways: 00:00 Intro 01:17 Why Brian Keating says AI is trapped by its own success 05:22 What could kill us first: AI, asteroid, or aliens? 11:46 What is the Simons Observatory? 21:17 How gravitational waves could reveal the universe’s first moments 33:47 Why Brian Keating wrote Losing the Nobel Prize 39:10 Why Brian Keating thinks about aliens all the time 44:10 Will we first meet aliens — or their information? 46:24 Would discovering aliens change our understanding of God? ----------------------------------------------------------------------------------------- Join this channel to get access to perks like monthly Office Hours: https://www.youtube.com/channel/UCmXH_moPhfkqCk6S3b9RWuw/join
📚 Get my books: Think Like a Nobel Prize Winner, with productivity tips from 9 Nobel Prize winners: https://a.co/d/03ezQFu
Focus Like a Nobel Prize Winner, with life-changing interviews with 9 Nobel Prizewinners: https://a.co/d/hi50U9U My tell-all cosmic memoir Losing the Nobel Prize: http://amzn.to/2sa5UpA
The first-ever audiobook from Galileo: Dialogue Concerning the Two Chief World Systems: Ptolemaic and Copernican https://a.co/d/iZPi9Un
AI isn't even like a cockroach yet.
So you have this danger now that the AI can only progress so much as it's training data.
LLM plus GPU.
It's fundamentally predictive.
Don't you really think that AI development is existential danger to humanity?
I do.
Brian Keating is one of the most influential astrophysicists of our time.
A world-class cosmologist, author, and public intellectual.
He has dedicated his career to unlocking the biggest mysteries of the universe.
If you said there's one galaxy, you're stupid.
If you said there's one planet, you're stupid.
If you said there's one...
So why say there's one universe?
He leads the Simons Observatory in Chile,
where scientists study the afterglow of the Big Bang
to uncover the universe's earliest secrets.
The multiverse concept seems to come along in most models of inflation.
What will kill us first, AGI or asteroid?
What about aliens, you didn't say?
I think about aliens all the time.
Why all the time?
If we detect these gravitational waves, then we win a Nobel Prize.
He said, Brian, I want you to build the experiment that detects it once and for all.
He didn't believe the universe began with the Big Bang.
Jim Simons called me up the day after this press conference.
It's on that phone right over there.
What's going on here?
Why are they claiming they detected this?
What goes through your mind when you look at the current stage of AI development?
I think AI is locked in a prison of its own success currently.
I've talked to some of the biggest leaders in AI,
and I certainly want to talk to many more of them.
The analogy I like to use is a strange one, and it's called the QWERTY keyboard.
The QWERTY keyboard was created to solve a problem for mechanical typewriters.
There used to be things called typewriters.
You're too young to remember them.
But they were mechanical things that you would hit hammers.
The hammer would hit a strip of ink-covered ribbon,
and it would have the shape of the type letter that you're trying to imprint on a piece of paper.
And it was a big hassle when you had these.
And the very first ones were purely mechanical.
The hammer would fire because you pressed hard enough on the paper.
Later, they had electronic motors that would hit, and things got better.
But in the beginning, the first typewriters I ever used were mechanical hammers.
And they would hit the paper, and they would hit the ink, rather.
That would make an imprint on the paper.
They found when they did that, if the letters were being typed,
sometimes the letters would be typed so fast
that the two hammers would go at the same time and stick together.
So the hammers for certain letters, like Q and U.
Q and U are almost always together.
So if you typed in Queen, the two Q and U would come out at the same time.
They would stick together.
Some guy developed this keyboard called the QWERTY keyboard.
The QWERTY for Q, W, E, R, T, Y.
Because the letters and that frequency, it's almost never the case.
In fact, it may never occur in English.
I'm sorry, in Polish, you guys have this so much more advanced language than we poor Americans.
But anyway, in English, the letters Q and W may never, ever fire together.
So there was no probability they would get stuck.
So this guy, whoever it was, or girl, I don't even know,
invented a keyboard to slow down progress of writing for a purely mechanical problem.
But it took off because it was the most effective thing and it was very popular.
Nowadays, we have the following.
We have a GPU and we have LLMs.
GPUs were invented to be graphical processing units
that are extremely efficient at matrix multiplication and matrix operations.
And those are the types of operations in what's called an abstract vector space
that LLMs like to use.
LLMs tokenize words.
And they're basically predicting the next letter that will come
after typing our next word or next fragment of a letter called a token.
They're extremely high probabilistic things.
They don't always get it right.
But with very high probability, they're parroting or predicting what the next word will be.
They do that extremely efficiently because the GPUs
were meant for doing matrix multiplications.
Is it just statistics?
It's extremely highly reliant on statistics.
So the original neural networks, the first ones that were invented,
were really just solving a very simple kind of vector-based dataset
for the U.S. Postal Service.
So it was called Perceptron.
It was the very first neural network.
And it was done because people write their addresses.
You probably have zip codes in Poland.
But in America, you know we have zip codes, right?
Ours here is 92093.
So there's only so many different ways you can write the number nine
So they had primitive optical character recognition.
They would digitize that in a 16 by 16 grid.
And then they would just have billions of people
after the U.S. Postal Service has processed a billion letters by the 1970s.
And they would scan it.
And they would just matrix and put it into a vector database.
And they would say, how often or how parallel are these two vectors?
My point is that this architecture, ChachiBT,
Claude, NotebookLM, all these things, Gemini,
all the top leading language foundation models
are all predictive statistical-based machines.
What will kill us first, AGI or maybe asteroid?
What about aliens, you didn't say?
No, asteroid is more realistic, I think.
Asteroids, thank God, we have a new telescope in Chile
in the southern Atacama Desert called the Vera Rubin Observatory.
Vera Rubin was a great astronomer.
She co-discovered dark matter.
She actually worked in this office too.
She was in UC San Diego working with my colleagues, the Burbidges.
And she was a tremendous astronomer who deserved the Nobel Prize.
Her telescope, named after her,
is detecting something like 1,000 asteroids an hour.
And we've detected most of the ones that have any probability
of crashing into the Earth in the next 30 years.
We don't see any.
And if they did detect one that's coming 31 years from now,
probably we could unite as a planet to deflect it slightly off course.
We have enough time to deflect it.
So I think I rank asteroid calamity,
Earth extinction-level event, very, very low, but not zero.
Thanks to tools like the Vera Rubin Observatory,
my colleagues that work on this, I think that's very low.
LLMs and GPUs, I also rank very low.
Because as leading to a disaster,
you're asking about which is more likely to cause disaster.
I rank them probably the same as the asteroid.
I don't rank AI that low.
But because the AI that we have now, LLM plus GPU,
it's fundamentally predictive.
That means it has to be pre-trained.
You know, GPT, general pre-trained transformer.
What is it trained on?
It's trained on stuff that humans have done,
and now increasingly on what other AIs have done.
Okay, so you have this danger now
that the AI can only progress so much as it's training data.
Nowadays, it's not the LLM, it's not the GPU,
how many GPUs you have.
Those are important, but it's your training data.
These groups and the weighting, what's called the model weights,
they fight over this, they keep it private.
That's why they have such power.
Don't you really think that AI development
is existential danger to humanity?
I do.
That's why I said I'm worried about AI, perhaps,
but because GPU plus LLM is so successful.
Right now, there's another type of keyboard,
I didn't mention, it's called Dvorak,
which sounds like somebody's name.
I think it was somebody's name.
So you can buy a Dvorak keyboard,
there's like 100 of them for sale.
How many QWERTY keyboard?
There's literally six or eight to 10 billion of them
because they're on every phone, right?
Every single keyboard.
That's because it got what's called technological lock-in.
When something is successful, it locks in.
Imagine you're building a new railway,
high-speed railway across Poland or across America.
We're trying to build one in California.
We spent billions of dollars.
We didn't lay any track, right?
But imagine you said,
I'm not only going to build a new railroad,
I'm going to make the railway,
I'm going to make the distance,
the gauge between the rails 10 feet wider
because I can have a massive...
That would be even harder, right?
Because we're so locked in.
All the train tracks that we have,
all the wood that we use,
all the actual locomotive car engines,
they only run on that distance, right?
So you'd have to rebuild everything.
So what I'm saying is,
the distance between the rails was locked in.
Do you know what caused the lock-in of that distance?
No.
A horse's butt.
The backside of a horse, okay?
So in Roman times,
a chariot was driven by two horses.
Your history on a little bit, right?
I love the Roman Empire, right?
I gotta research it.
No, I love learning.
I'm passionate.
I go to sleep.
My wife hates it, but I go to sleep.
I wear headphones.
I'm listening to a podcast.
Are you a rule breaker?
I think so.
I don't get off on it.
I don't look for reasons to break rules
like Richard Feynman or somebody like that.
He's not...
I mean, he's a hero in some sense,
but I don't look at him as an avatar.
I'm more about how much new knowledge can I generate
with the finite time I have
while still being a good husband and father,
friend, citizen, and so forth.
I have a very deep sense
of what it means for me to be a man.
That's my definition.
And part of what I do as a man
is gather as much resources and energy as I can.
And that's mostly intellectually.
I'm not the strongest, fastest.
I don't have sharp claws or tiger teeth,
but I have a rapacious desire to learn and to teach.
I view myself as a scientist first.
I view myself as a father and husband first,
but in work, I'm a scientist.
Second, I'm a teacher.
I love teaching.
I love learning.
You were here setting up.
Yeah, because you wear a lot of hats.
Yeah, I do.
I have to.
As a professor, there's not one job being a professor, right?
There's not one job being a podcaster or a scientist
or an interviewer or an editor.
There's millions of jobs, right?
But for me, since you didn't ask,
yes, these are important questions.
I'm not worried about these things happening.
I talked to Jan LeCun, who's a great...
Actually, you talked to him.
Yeah, you had an interview.
You actually had a post about him.
By the way, she decided to leave math.
Yeah, I know.
I saw your tweet about it.
Yeah, that was great.
And then congratulations to him.
He's a brilliant guy.
He told me, and he probably told you the same thing.
AI isn't even like a cockroach yet or a cat yet.
You and I are processing trillions of bits of information.
LLMs plus GPUs can process millions
or maybe millions per second, but not multimodal.
It's not smelling.
Right now, if we smelled something right now,
we would leave the building.
If I have a llama running on my Mac Pro over there,
they wouldn't run out of the building, couldn't do anything.
Here's what the Keating trigger is, though.
I have a trigger when I will start to worry about it.
And that's when AI, whatever it is, I know it won't be...
I won't say I know, because that's arrogant.
I'm never arrogant.
I never try to be arrogant.
I will never say that LLM plus GPU can't destroy the world.
It can, but I don't think it's likely,
because it has to be trained on things
that humans have given to it.
And we haven't given it the instructions for AGI.
It doesn't have...
Humans have not created by definition, right?
Artificial superintelligence does not exist
in the human corpus, which is what they're trained on.
And because Claude, ChatGPT, Gemini are so successful,
they squeezed out any other alternative.
There's almost no other alternative.
NVIDIA just brought this company called Grok.
What is the first tremor you are trying to catch?
So the Simons Observatory is $100 million...
What is the Simons Observatory?
Maybe that should be my first question to you.
So the Simons Observatory came about
as one of the positive consequences
of BICEP's wrong claim, retracted claim,
of the discovery of inflation.
Jim Simons is a longtime friend of my late father.
They both worked at the State University of New York.
And they were both mathematicians.
And Jim Simons was kind of like an uncle to me as a kid,
growing up, and I used to play with him.
And then he went on to make literally billions
and billions of dollars.
And then he established a charity
called the Simons Foundation.
In 2014, BICEP announced,
we detected waves of gravity
imprinting the cosmic microwave background
with beam-mode polarization.
Six months later, it was proven to be
not from the cosmos, but from the galaxy.
It was proven to be grains of dust
aligned magnetically with the galaxy's magnetic field
that perfectly mimicked and impersonated
the effect of the Big Bang producing
inflationary gravitational waves.
It was very, very unlucky in some sense.
And you lost your potential.
Not only me, but all of our colleagues.
But after that experiment was done,
we worked with the Planck Satellite,
which is a European, primarily European-led project,
ESA project, called Planck.
The BICEP team and the Planck-
Because of, I'm sorry, because of the interstellar dust, right?
Exactly, yeah.
They had data that we couldn't get.
We, for technical reasons, we operate at one channel,
one frequency, and that one frequency
can only see one signal.
It can't see the CMB, gravity waves,
gravitational waves, and dust.
But Planck could see gravitational waves,
dust, electrons.
They were less sensitive than us,
but they had more leverage on detecting the stuff
that we didn't want to see, the dust.
And so they wouldn't share that data with us
before we published our paper,
but they did share it after we published it,
and they helped us retract the claim
of detection of inflation.
So we worked together.
So now we know no one's detected
inflationary gravitational waves.
Doesn't mean inflation didn't happen.
Doesn't mean it did either.
It just means there's a lot of contamination
that you have to get rid of in your instrument,
in your observations.
Jim Simons called me up the day
after this press conference,
the day of this press conference,
on that phone right over there.
And I recognized his voice, and he said,
Brian, what's going on here?
Why didn't you detect this?
Why are they claiming they detected this?
He was already funding an experiment
that I had started building with my colleague
at Berkeley named Adrian Lee called the Simons Array.
It was a precursor to what became...
And he said, I want to talk to you about this.
Come out.
And by the time I got together
and talked with him about six months later,
BICEP had been falsified.
Our claim had to be retracted.
So no inflationary gravitational waves.
Again, could be there.
We don't know.
He said, Brian, I want you to build the experiment
that detects it once and for all,
that measures the dust,
measures the synchrotron,
measures all these contaminations,
and measures the gravitational waves,
if they're there.
Because he didn't believe the universe
began with a big bang.
But he was brilliant.
He said, if we build this experiment,
now called the Simons Observatory,
with my colleagues Mark Devlin and Suzanne Staggs
and Lyman Page and Adrian Lee and Cam Arnold
and all of my great colleagues
on the Simons Observatory back in 2015,
David Spurgill, who's now the president
of the Simons Foundation, a brilliant scientist.
If we detect these gravitational waves truly
after removing dust and everything else,
then we win a Nobel Prize.
If we don't detect it,
that gives more evidence, perhaps,
that the universe didn't have a big bang,
which is what he believed till his dying day.
He died in 2024.
I remember my interview with Keith Turner.
Yeah, so it's exactly, exactly the same thing,
except they detected it first.
Yeah, they detected gravitational waves,
but not from the big bang.
They detected them from massive black holes
in our galaxy and other galaxies.
But this is what I want to say.
The novel feature of the Simons Observatory
is it's in the Atacama Desert.
It has the ability to measure dust
and the cosmic background at the same time,
remove the dust,
and then what's left will only be the cosmic signal.
Why Chile?
So, Chile, these experiments
are very sensitive to microwaves.
This is what we're trying to measure.
The heat left over from the big bang
is now redshifted from X-rays and gamma rays
and ultraviolet.
Now it's redshifted by a factor of 1,000.
That means it occurred at a redshift of 1,000.
And that redshift means that the wavelength
is now in the microwave,
which is about a millimeter in wavelength.
So you can actually see the wavelength
with your fingers, right?
You can't see the optical waves that your eyes see.
So it's thousands of times smaller
in optical light,
so thousands of times lower frequency.
Microwaves, if you've ever used a microwave oven,
are absorbed in water molecules.
Water molecules gyrate, vibrate,
and start to heat up.
That's how you boil water in a microwave
or heat your food.
If you put in a cup like this cup here,
if I put it in the microwave with coffee in it,
or vodka, if I have vodka in here.
I do have some actually over there.
This cup is made of ceramic.
It's completely dry,
but the water inside of it will heat up.
But it can touch the cup because it's dry.
Water absorbs microwaves.
So what we want to do is go somewhere
where there's no water.
Where is a place where there's no water?
Outer space.
Outer space is incredibly expensive to get to.
Even with SpaceX, it's incredibly expensive.
Billions, billions of dollars.
Once Roger Penrose used the similar cap
to explain me something in a cosmology.
So that's why I'm smiling.
He was probably talking about rotations.
That's why I'm smiling.
Group theory, right?
Like you can do a two-fold rotation.
So getting back to the Atacama,
there's two places on Earth that are
what we call a poor scientist satellite.
You'd like to be in space.
This plant cost a billion euros.
Our experiment now costs 100 million dollars.
It's still expensive, but it's 10 times less.
We couldn't launch one of our telescopes.
It's six meters in diameter.
It won't fit in a telescope.
It's bigger than the Webb telescope.
The Webb telescope costs 12 billion dollars,
so it's 100 times more expensive.
I'm not saying it's cheap,
but the Simons Foundation,
our partner institutions and universities
have been very generous, and we funded it.
It goes to Atacama Desert.
We're at 17,200 feet.
We're at 5,000 meters above sea level.
There's almost no water in the atmosphere.
You don't want this photon traveling
for 13.8 billion years.
Right before it's about to hit
your telescope mirror or lens,
you don't want it to get absorbed
and heat up a water molecule.
So that's why we go to the Atacama Desert.
The other place is the South Pole, Antarctica.
Very dry.
I know that you are building your observatory
on land that is sacred to local society.
Does what you do justify changing
landscape wholly for the people?
I need to ask you about that.
It's a really wonderful question.
In 2019, there was a total eclipse
of the sun visible from Chile.
We actually went to the site in Chile
with Jim and Marilyn Simons and David Spergel.
We broke ground on the Simons Observatory
right before this eclipse.
Simultaneously, the Simons Foundation
hired a film crew.
The film crew was led by Deborah Kellner,
who's a famous Parisian filmmaker.
She's done many award-winning documentaries,
Syria and the Maldives and other places.
But also, she came on a four-year expedition
with us tracking the inception to design,
to building, to shipping, to installing,
to taking data.
She made a film about it.
It's called The Eternal Sky,
and it features me in some tiny role,
but mostly Jim and Marilyn Simons
and David Spergel and others.
It traces the human lives
of what it takes to build.
But we captured the eclipse
from the eyes of the Atacaminos.
These are the indigenous people
of the Atacama Desert.
They were so transparent.
They had never seen,
we gave them eclipse glasses.
We let them observe
through optical telescopes.
We participated in this.
We have a mission and a mandate
to allow Chilean astronomers
have access to our telescope.
They have to have it or we get shut down.
It doesn't matter how much money.
We have to allow education
for indigenous and Chilean astronomy,
and this has been a huge benefit
to their economy and even culturally.
Now they'll do tours,
so the Atacaminos and others
will come up to the telescopes.
We support them.
They support us.
We do education in their schools
in the local town and village
called San Pedro de Atacama
and throughout Chile.
So we try to give back.
We try to respect it.
We just decommissioned the Simons Array.
We had to restore it
to the pristine original condition.
A lot of the difficulty working there
is that there's a lot of resources in Chile.
A lot of minerals are located in Chile.
Lithium, one of the greatest stores
of lithium on Earth,
one of the biggest copper mines.
Remember, these miners
were trapped underground.
That was in the Atacama Desert
10 years ago, 13, 14 years ago.
These are, it's incredible
full of resources,
but it's also very much a modern country
that wants to improve their science,
technology, engineering, and math
while respecting the cultural traditions
of their ancestors.
How does searching for swirls in gravity
allow you to see something more?
So what we see is sort of a,
we're a space-time detector.
We're not, yes, we use light
to detect things, right?
There are many ways
you can detect something, right?
Imagine right now,
we're near a military base here.
There's a marine air station here.
25% of that city, San Diego, yeah.
Yes, is military.
So we have F-35s, like Poland has F-35,
the newest stealth fighter, right?
So when they fly over on a cloudy day,
yes, in San Diego,
some days are cloudy.
Unfortunately, a few percent per year.
You can't see the F-35,
but you can hear it.
So you build a detector for F-35s.
The stealth doesn't help you, right?
Imagine you were trying to detect it.
It's crude, but you could detect it
with your ear.
And then if you were an enemy,
you could try to hurt it
or do something, right?
I'm giving you an example.
You can detect things
in many different ways.
One of the ways that you detect the F-35
is by looking at it directly with your eye
or bouncing a radar wave off of it.
That'd be hard
because it has a very low cross-section.
Or hearing it, right?
It's just making an experiment.
But the same thing is true of the cosmos.
There are many ways to detect
many different phenomena.
And a good scientist should use
all of the things that are available to them.
Sight, sound, radiation, etc.
We're trying to see waves of gravity.
We can't actually see them optically.
We can't see them with radio telescopes.
But we can see their effect on things.
For example, LIGO.
You talked to Kip Thorne.
LIGO detects gravitational waves.
It doesn't see them.
It sees the compression and expansion
of what's called an interferometer
located on the Earth's surface.
There's two of them.
One's in Washington.
One's in Louisiana.
And they squish and squash
depending on the pressure that's exerted on them,
the gravitational force that's exerted on them,
by waves of gravity coming from some direction.
And let's say it's coming from a direction
that goes through Washington state first
and then goes to Louisiana.
We know the distance between those
a couple thousand kilometers.
We know how fast gravitational waves should travel.
That means if you see one here,
you can set off an alarm a millisecond
before it gets to the other one,
and then you can prove that it exists.
If you only see it in one
and you never see it in the other,
then you falsify that it exists.
What we're trying to do
is not use the Earth squishing and squashing
or these tubes.
We're using the light
from the formation of the elements called the CMB,
the cosmic microwave background,
the leftover heat.
That surface, there's a surface,
fictitious surface called the surface of last scattering,
that also will squish and squash
due to gravitational waves.
Not from black holes,
but from the background that was produced
from this incredibly violent expansion of space-time
called inflation.
So we use the entire photosphere
as a detector of the gravitational waves
that were produced,
presumably during inflation.
What if your observatory shows
the universe had no beginning?
So what's crucial, again,
is what we try to do
is to establish what is impossible.
My podcast is called Into the Impossible.
What is not possible
is to say right now
that the universe is static,
that the universe is not changing.
So that's one thing we can say right now,
the universe is expanding.
That's one fact about the universe
that's incontrovertible,
that people try to controvert,
and there are people that believe
that that's not true, by the way.
They're fringe.
They're outsiders of the mainstream.
Their opinions are not really very valid.
There's no data to back up.
There's evidence against them, right?
So we can say for sure
in mainstream cosmology
that the universe is expanding.
That means it's not static, right?
Even Einstein believed the universe
was static for some time,
and then he was convinced
by Edwin Hubble's observations
that the universe is expanding.
We now know for over almost 100 years now,
the universe is expanding,
and we know that it's expanding
at an increasing rate.
That expansion is accelerating, right?
We say the expansion is accelerating.
We believe that's caused
by a mysterious force
that is known as dark energy
or vacuum energy
that was originally put in the equations
to stabilize the cosmos against collapse
because they thought the universe was static.
They knew the universe,
things weren't moving,
but they knew if you had gravity,
the universe should collapse.
Or maybe dark energy is a mistake
of Einstein's equations.
No, he thought it was a mistake.
At the end,
he was convinced by Hubble, by data.
That's what I'm saying.
What Einstein did,
here's the fact pattern,
and again,
always think when you ask me,
do I believe?
I don't believe in anything.
I look for evidence for things, right?
So I want to see if there's evidence
for or against.
So the evidence at the time
was that all the stars,
the universe was our galaxy.
This is what Einstein believed.
The entire universe was our galaxy.
By the way, Copernicus believed
our solar system was effectively
the whole universe, right?
Because that's all they could see.
They didn't know how far away
the stars were at that point.
Bruno and Galileo made predictions about it,
but they didn't have any evidence either.
And the fun fact that people like to think about
that Galileo proved Copernicus right
is they're actually incorrect.
Galileo had evidence
that was completely wrong.
His idea was right.
The earth does go around the sun.
We know that's a fact,
but not from Galileo
and certainly not from Copernicus.
Copernicus had no way to prove
or disprove his idea
that we now call Copernicanism, right?
There was no data.
There were no telescopes,
Tycho Brahe,
they didn't have data.
No tools.
There were no tools, right?
I'm a tool builder.
That's what an instrument,
that's what an experimental scientist does.
So you said it could be this or that.
Okay, so Einstein in 1915 through 1919,
he came up with the general theory of relativity,
the theory that concerns gravity.
He said that a universe that has stars in it
means that it has light
and has matter and the planets, right?
There's matter in the universe.
We know we're made of matter.
We know the sun is made of matter.
We know the sun produces light.
If you only have those three ingredients,
matter like planets,
matter like stars,
and radiation from the stars
and almost nothing else,
then that universe is gravitationally unstable.
Isaac Newton knew that
because gravity is only attractive, right?
You have two masses,
they only attract each other.
Unlike electromagnetism,
you can have repulsive charges,
like two light charges repel each other.
There's no repulsive gravity in classical gravity
and Newton knew that.
So he said the universe has to be stabilized,
otherwise eventually it would collapse into,
we didn't know,
but eventually they'd say a black hole or a singularity.
Einstein, at that time in 1919,
they didn't have data that the universe was expanded.
So he said the universe is static,
except for the planets.
Planets in Greek means wandering, moving,
like plankton means wandering in the ocean,
or a plane means something that travels in the air.
Planet means wanders in the sky.
Those are the only things that moved.
That's what they thought.
The evidence only showed that.
Until the discovery of redshift,
of galaxies,
and of the motion by Hubble and Lemaitre and others in 1929.
As soon as Einstein saw those data from Hubble,
every galaxy in the universe,
except for the Andromeda galaxy and three other galaxies,
those four galaxies were moving towards the Milky Way,
everything else was expanding away at high speeds,
close to fractions of a percent of the speed of light,
very high speeds.
Not contracting, not static, but expanding.
Then Einstein said,
my theory of the cosmological constant of dark energy,
we call it now dark energy, is falsified.
But really, his model was falsified,
but dark energy was still possible.
I have two questions from our viewers regarding your observatory.
If Simon's observatory finds evidence for inflation,
we get a multiverse.
If it doesn't, we get a cyclic universe.
We get a world more.
So what's being said here is a statement of the fact
that the multiverse concept seems to come along
in most models of inflation.
Most models of inflation have something like a multiverse.
Not all, but almost all of them do.
And the creator of inflation in its modern form,
chaotic inflation, Andre Linde at Stanford University,
you should talk to him while you're up there.
He's like your neighbor, right, in Palo Alto.
He says it's wrong to think of a universe.
He's told me, he said,
why do you keep calling it the universe?
You should start with the multiverse.
His point is Copernican.
Copernicus would say that if you said
that there's only one planet, you're totally wrong.
We know that there's many planets.
In our solar system, there's seven other planets
besides the Earth, right?
Then there's literally 10,000 or more minor planets,
asteroids, meteoroids.
Then there's literally 100 billion stars in our galaxy,
and there's possibly 400 billion other galaxies
in the observable universe.
So if you said there's one galaxy, you're stupid.
If you said there's one planet, you're stupid.
If you said there's one...
So why say there's one universe?
That's his point.
You should start from a Copernican perspective
that we're not special.
The universe surrounds us,
and we're surrounded perhaps by other universes.
The problem is there's no way to prove that.
We don't think that there's a way
to necessarily prove that.
There's a couple of ways that you could get evidence for it.
For example, if our universe...
I usually use that beach ball.
You want to pass me that beach ball?
So here's...
Yeah, just throw it in there.
Oh, you dropped the universe.
So this is a fictitious God's eye view of our universe.
We're actually at the center of this beach ball,
and we're looking at it from the outside,
just like you could make a map of the constellations
just like this, right?
This orange band is our galaxy that surrounds us,
and that emits a lot of radiation,
and it can mimic some of the radiation signals that we saw,
and this is effectively what we saw with BICEP.
With BICEP, we were looking in this region,
which is very, very cold.
We thought it was far away from the galaxy,
but actually the galaxy extended its emission
into this region.
Now imagine there's another beach ball here.
There's a globe behind me,
and imagine that is another universe,
and it bumps into our universe.
Think about what's happening.
There'd be effectively two photospheres, right?
Imagine two beach balls,
and they intersect with each other.
The intersection point is a circular boundary,
so there are people that propose to detect the multiverse.
You look for these circles of patterns in the sky.
So far, we haven't seen them.
That's one example.
But this is unfalsifiable.
It's unfalsifiable.
It's potentially provable if we do someday see it,
but here's my point.
My point about the...
I make the analogy in my first book.
Losing the Nobel Prize.
I make the analogy about...
Imagine a Petri dish where bacteria are, right?
So there's bacteria,
and you put a couple of bacteria in there.
There are certain types of bacteria that make toxins
that kill off other bacteria, okay?
These are chemical warfare that the bacteria go on
so that their colony can thrive.
So they could be located physically close
to another colony, culture of bacteria,
but never know it.
They never interact with it
because there's a toxic layer that separates them.
The chemotoxic layer between the two of them.
But what they could say is that I exist on this Petri dish.
That means it's possible another culture of bacteria
could exist, even though I can't see it.
It's beyond my horizon.
I cannot directly observe it,
but I can admit that the possibility is there.
But you're right.
It's unfalsifiable.
So the existence of inflation would be sort of like that.
It would say we saw gravitational radiation.
It's a background.
It comes in all directions, all seasons,
all times, all days, all weeks, all months.
We see it all the time.
Every frequency, everywhere we look,
we see this background of waves of gravity.
That means it's very likely that we did emerge
from a region of space-time that began inflating.
The only way to have that inflation begin
is if the inflaton,
if this quantum field called the inflaton
exists in a wider cosmic space called the multiverse.
So our observable universe
is like one of those bacteria cultures.
There could be more.
Can we prove it?
Not unless we're very lucky.
Why you've decided to write a book, Lost Noble Prize?
So I wanted to document the religious side of scientists
that they never acknowledge,
which is that we all have an idol that we worship.
And the perils of worshiping idols
have been known since biblical days.
Is that if you part, you know,
your stance meant that you didn't win the Nobel?
Certainly people have claimed that.
I should say the following, though.
In my mind, I admit to huge character flaws that I had,
which was that I did worship the Nobel Prize.
It was sort of like becoming Richard Dawkins.
Like, imagine Richard Dawkins
is in a religious Christian household,
and then he becomes an atheist,
and he wants to say, like, here's what's wrong with that.
But unlike Richard, Richard's not trying to rehabilitate
the religion and Christianity and Islam.
He wants it to disappear, essentially, right?
But I'm trying to make the Nobel Prize
last for another hundred years,
and I'm doubtful that it will.
What's wrong with Nobel Prize?
Many things.
I wrote a book about it.
The book, by the way, is only three chapters
about the Nobel Prize,
but it's all based on interviews and conversations
and thoughts that I've had about Reformation.
Martin Luther, the Protestant Revolution,
in 1590, whatever,
this is his thesis, 32 theses, 39 theses,
whatever it was.
Again, forgive me.
My Catholicism and Christianity is long, long past.
He nailed these to the church door.
Why?
Because he wanted to reform it,
not to kill Christianity, God forbid, right?
He wanted to improve it so it would succeed and thrive,
and now it does,
and Catholicism still thrives too, right?
He said, you need to reform.
Here's how you reform.
I don't know all the details.
I said, you need to reform the Nobel Prize.
Otherwise, it may not be relevant.
Here's the biggest problem.
The Nobel Prize teaches people
that science is done by geniuses
working almost by themselves,
and it is done in a way that's not in agreement
with the father figure of the religion, Alfred Nobel.
Alfred Nobel said three things about the Nobel Prize.
He said it should go to one person.
He said a man who, in the last year,
made a discovery that had the greatest benefit
for all mankind.
So three conditions.
One person could be a woman.
A woman did win it the second year.
Greatest benefit to mankind in the past year.
So the first Nobel Prize was given to Wilhelm Roentgen
for developing the x-ray.
What does the x-ray do?
It takes diagnosis things inside of your body
so that you no longer have pain.
I just got a cavity fixed yesterday.
They took 70 x-rays.
I thank God they had an anal.
They found the filling, and they fixed it,
and they did all sorts of things.
I was in horrendous pain.
I couldn't do this interview yesterday.
Today, I'm fine.
I'm talking more than usual,
because I'm having fun talking to you.
I couldn't do it.
It had a huge benefit.
For 120 years, it's had a benefit.
It was 1896, he invented it.
Number two, single person.
He was a single person.
He was the only one who won it.
So it was a single person in the preceding year,
and it had the greatest benefit.
So he matched it.
Nowadays, we have people like Kip Thorne.
So Kip Thorne won it with my...
He's been in all three of them were on my podcast.
Ray Weiss, Kip Thorne, and Barry Barish.
Barry Barish won the Nobel Prize share
along for managing and building the instrument.
Kip Thorne for the theoretical aspects of it
and algorithmic aspects of it.
And Ray Weiss won half of it
for developing the LIGO concept.
So how many people built LIGO?
2,000 people.
Analyze the data, 5,000 people.
All of them were left out.
And I talked to Barry Barish.
He wrote the foreword to my second book
called Into the Impossible.
He wrote basically the introduction,
saying that he agreed with...
You believe that AI can change everything.
What will the Nobel Committee do in that case?
They've just won four Nobel Prizes
in the last year, right?
It's a new world.
I think that the AI, for me...
Before you guys, we started recording,
I was meeting with three
of the smartest students on earth.
Here.
Here in my weekly office hours,
meetings with them.
I have to say that they're 100% more
promising than I was at their age.
I would say right now,
I know 10 times more about AI than they do.
We're talking about building a quantum computer
that runs AI and launching it into space.
That's our project for the next couple of years.
I don't know if we'll succeed,
but we have to try.
Even basic things, like you said,
Claude, Gemini, Notebook LLM,
Chat GPT 5.2.
They don't perplex it.
They weren't using it.
They weren't doing the stuff
that I'm doing every day.
I have now changed my whole research projects.
Maybe in the future,
AI2 will be awarded by Nobel Prize.
It probably could be, yeah.
I would say that when you look at Demis Hassabis,
who won it,
when you look at him,
he used it to solve a scientific problem.
It wasn't the AI that won it.
It was for his research.
And his research is scaling up,
and it's very effective at doing that.
But until you know that Demis was awarded, no.
It wasn't deep mind, right?
So in the future, could that happen?
It could. It could happen.
Have you ever think of aliens?
Aliens, yes.
I think about aliens all the time.
Why all the time?
Well, I think there's three
very interesting questions to me.
There's three very interesting questions
that keep me up at night
that I'm infatuated with.
One of them is the origin of the universe.
That's what I get paid to do,
to study how the universe
could have emerged from nothing,
or it could have emerged from
perhaps there's a creator,
perhaps not, right?
So that's the origin of the universe.
Let's say we live in a single universe.
Forget about the multiverse.
How did a universe come from no universe?
Okay, that's the question.
Next question.
How did life come from pure chemistry?
How did life emerge in the universe
from hydrogen, helium,
stars collapsing, exploding,
iron, nickel, asteroid?
That explains how life got to the earth,
but it doesn't explain
how life originated itself, right?
So if life came on an asteroid
called panspermia or meteor,
came to earth,
I don't still know how it began somewhere else.
But what I'm saying is,
the second pillar of the question
is how did a universe create life from non-life?
First question.
How did a universe come into existence
from a non-universe?
No universe, some day was a Tuesday.
The next day is a Wednesday.
The universe comes into existence, okay?
That's a deep mystery.
That's what I get paid to do.
That's experimental cosmology.
Second question.
How did the matter produced there,
which is just hydrogen,
really for 380,000 years,
basically just hydrogen.
Eventually the stars are created.
They make helium.
Eventually those make heavier and heavier elements,
explode, make nitrogen, oxygen,
silicon, iron, et cetera.
And we're made of those materials.
We only exist because a star blew up
a billion years before our star was born, right?
We wouldn't be having this conversation.
I'm gonna give you some meteorites
that I give away to people that join my YouTube channel
and my mailing list.
I'm gonna give you some, of course,
because you made the journey here.
But those meteorites have the exact same iron
that's in your hemoglobin molecule in your blood, okay?
So life came from non-life.
Universe came from non-universe.
Next thing, consciousness came
from non-conscious matter, right?
Eventually we had amino acids, DNA,
but that's not a brain.
Remember I said-
Exactly.
So those are the three things that keep me up at night.
So the one that relates to aliens
has to do with the origin of life.
Did life only originate here?
I mean, it seems pretty lucky if that's true.
We don't have any evidence of aliens.
There's zero evidence of aliens.
People talk about unidentified aerial phenomenon,
UAPs, UFOs.
There's no one.
I've talked to the best astronauts,
the most famous, the most accomplished astronauts,
people that lived in space for a year or more.
No evidence.
None of them believe in actual aliens
are existing or coming from other planets.
But if we did discuss that,
remember I said before,
you can't do an average.
You can't make a standard deviation.
You can't do statistics unless you have more than one,
because you divide by N minus one.
So if you have two, if we found life on Mars,
that's a start.
If we found alien life,
even if it's just some mold,
living in chlorophyll on a planet on Proxima Centauri,
surrounding Proxima Centauri,
that would be a discovery for all time.
And then imagine you found dolphins,
some space dolphins on some other planet.
That's still not technological life, right?
Dolphins aren't making iPhones.
Then you find a technological life.
They would have to be so advanced.
They could tell me right now,
stop building the Simons Observatory.
We'll tell you how the universe began.
Now you can study the brain,
so you understand why it's so exciting,
why it would keep me up at night.
Because if you can find aliens,
again, there's zero evidence of aliens.
That doesn't mean there's zero evidence of aliens currently.
Yes, nobody said there's aliens with technology
or anything like that.
They're saying there's possibilities of it.
And I say, I agree, there is possibility.
But that possibility could be different from zero
by one part in 10 to the 24th, right?
There's 10 to the 24th planets in the observable universe.
There's maybe a trillion in our own galaxy.
We don't see any evidence, but we haven't searched.
We've searched a tiny little thimble
out of the entire ocean of knowledge, right?
But so I'm saying, right now we have no evidence.
That doesn't mean there's no possibility.
So possibility and evidence are completely different things.
What's the likelihood that we discover aliens?
Nobody knows.
There are things called the Drake equation,
SETI Institute.
They study these things.
Jill Tartar is a friend of mine.
I've been on the podcast many times.
They believe that we're just at the beginning.
It's like, we just discovered that germs exist.
You know, germ theory was only discovered in the 1800s.
And now we have DNA sequencing and we know about viruses.
We didn't know any of that.
We've only known it for a hundred years.
Maybe we're at that same point.
We're at 1869 in the equivalent study,
you know, before origin of species.
We're just at the beginning.
Again, no evidence, but huge opportunity.
Is the most likely scenario that the aliens we meet
will be AI astronaut?
So if you were to transport something across the cosmos,
the only thing that can travel at the speed of light is information.
Information in the form of radio waves or microwaves,
maybe neutrinos or some of the gravitational waves, radiation.
So it would be much cheaper to accelerate,
you know, a knowledge transfer and dissemination throughout the universe
if it was in the form of pure information.
Now, pure information can't do stuff.
We can't convert that information.
You know, if I broadcast to you, Lord of the Rings, right?
It's not going to start building, you know, a magical ring, right?
It'll tell you information about it.
But to have actual constructors, David Deutch and others have talked,
those that you've interviewed,
to have actual construction and manufacturing of things,
those take mechanical matter-based agents.
We might have the instructions for it.
In other words, if the aliens transmitted to me
what happened at the moment of the Big Bang,
that it was inflation,
here's how much gravitational waves there are,
look over there,
you'll see exactly what we should see at this frequency at this time.
That would save me.
I would not have to go through the effort of constructing the experiment.
Do you understand?
So this is what's so fascinating.
We can obtain their knowledge,
but maybe we won't actually meet them, quote-unquote, whatever they are.
Maybe we'll meet their information,
their avatars that'll traverse the galaxy at the speed of light.
Even at the speed of light,
it would take 30,000 years to cross into the center of the galaxy.
So the universe is enormous.
That's just one galaxy out of a trillion or more galaxies that we can see.
There could be even more than that.
So the question is, if we encounter something,
yes, to me, the most likely thing,
it would be not like a spaceship or a UFO or some flying saucer.
It would be some signal.
We would get some signal at the speed of light
that would reveal something that we know,
because we have to decode it.
You know, maybe broadcast the numbers of pi,
digits of pi or prime numbers or something like that.
And then we would have the answers potentially
to the physics of 10 centuries in the future.
If every lab laboratory or your Simons Observatory
finds aliens, would you change your religion,
your God understanding?
You know, that's such a counterfactual question.
It would have to be pretty substantial, right?
Because some people effectively...
How do you understand substantial?
Well, I would say this.
There are people that treat aliens
as if they would be a form of religion and of a God.
Supernatural.
Certainly, you know, the...
Like an AI.
Yeah, like us to ants or something.
Like my cat.
My cat doesn't really understand me.
He's very, you know, he's very solicitous of me
when he's hungry.
But most of the time, he doesn't really...
He knows I'm not gonna hurt him.
He knows I'll pet him.
And he seems to like me, but you know, who knows, right?
I don't know what's in his mind.
He doesn't have a theory of mind for me.
I can't tell you what the aliens would be like,
because they would be more different than me and my cat.
They would be like me, you know,
I would be like a crouton, you know,
on a piece of salad, you know,
compared to, you know,
compared to Stephen Hawking, Albert, you know, whatever.
We're not even in the same universe.
Literally universe.
And so for you to say, well, that changed you.
It really depends.
If there are science fiction authors,
like my friend Andy Weir,
who wrote The Martian and other books.
In one of his books, Project Hail Mary,
there's an alien,
but the alien is actually less advanced than we are.
It's kind of a count...
Normally, we think the aliens would be so advanced.
Certainly to get here, they would be more advanced,
like Arrival or, you know,
the three-body problem or things like that.
They would have to be more advanced if they came here.
But if they're just sharing information,
well, we've had, you know,
radio waves for almost 100 years.
So maybe they're only 100 years behind us,
or maybe they're 100 years ahead of us,
but not a billion years ahead of us in evolution.