February 26, 2026 • 48 mins
Jonny Dyer is the founder and CEO of a satellite company called Muon Space. The company’s first big project is a satellite constellation called FireSat.
Jonny’s problem is this: How do you capture data from space to help manage wildfires around the world in near-real time?
In this episode, Jonny explains:
- How smartphone tech inspired smaller, cheaper satellites.
- Why fighting wildfires is still shockingly primitive.
- Why China does not distinguish between commercial and military satellites.
- Why data centers in space might arrive by 2035.
Connect with us:
- Follow Jacob Goldstein on LinkedIn, X and Instagram
- Email us at problem@pushkin.fm
- Follow Pushkin on Instagram, LinkedIn or X
- Listen to Jacob’s other show, Business History
- To listen to the show early and ad free, sign up for Pushkin+
See omnystudio.com/listener for privacy information.
Listen
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:15):
Pushkin. I'm Jacob Goldstein and this is What's Your Problem
show about people using technology to solve problems that matter.
My guest today is Johnny Dyer. He's the co founder
(00:35):
and CEO of a company called Muon Space. Newon Space
was founded in twenty twenty one, and the company designs, builds,
and operates satellites. Their first big project is a constellation
of satellites called fire Sat. And Johnny's problem is this,
how do you build satellites that can capture data from
(00:56):
space to help manage wildfires around the world in near
real time. In our conversation, Johnny and I also talked
about his company's work with the US Department of Defense,
and more generally, the global geopolitics of satellites, which basically
means the US versus China in space. To start, we
(01:17):
talked about a satellite company called Skybox. Johnny joined that
company just after it was founded in two thousand and nine.
Up to around that time, satellites had mostly been giant, expensive,
bespoke things, but this was a moment when a few
people were starting to think that satellites could get much
smaller and much cheaper. Johnny told me that the idea
(01:37):
was partly inspired by the emergence of the iPhone.
Speaker 2 (01:40):
You shrink cameras down enough, you shrink compute down enough,
all of a sudden you have this thing that is
possible in a much smaller package than was ever possible before.
And so there was a strong analogy. We thought with
satellites that if you can use these commercial technologies, you
can do a lot of the things that these big
billion dollar systems are doing in a much smaller package.
If you can do it in a smaller package, everything
(02:02):
gets faster, cheaper, because most of the challenge of getting
stuff in space is launching, and that on a per
kilogram basis, like it just costs per kilogram to get
stuff in space. If you make it smaller, everything gets better.
And so we, you know, we did this and we
were very successful at it. We ended up building twenty
one satellites that were you know, took data of the
(02:23):
same quality that literally, you know, one hundred million dollar
systems before us were collecting. We got bought by Google
and in the process we learned a lot about what
it takes to do this, what works, what doesn't work.
But also, you know, there there was sort of a
lot of kind of frustrations we had about the fact
(02:46):
that we solved this problem, this technical problem, and figured
out how to deploy these things in space, but we
really were only applying it to one small use case,
which was this high resolution visible imagery. Think about what
you see in Google Maps. So that's what we were doing.
We're taking like submeters roughly one meter resolution visible images
like you see Google Maps.
Speaker 1 (03:04):
So when I look on Google Maps and I see
the trampoline in my backyard, which I do, Uh, is
that you taking that picture?
Speaker 3 (03:11):
So probably the trampoline in your backyard.
Speaker 2 (03:13):
Probably not that was probably taken it from an aircraft
because that's probably more like ten centimeter imagery.
Speaker 3 (03:19):
But if you go to.
Speaker 2 (03:20):
A big trampoline, yeah, I mean you would see it
in our imagery.
Speaker 3 (03:23):
You would definitely see a trampoline.
Speaker 2 (03:25):
If you go a little bit outside a major urban
area where it's too expensive to go collect aircraft imagery,
all of that imagery and Google Maps would be the
type of stuff we were collecting. So you can go
search around Google Maps and you can see a bunch
of that stuff. And so you know, we got we
went through that journey. Eventually it ended, but stuck in
the back of my head was we've shown, you know,
(03:46):
we've shown a different paradigm for doing this, and there's
so many possible applications of that new paradigm.
Speaker 1 (03:53):
So so kind of the basic narrative of how you
get to muon just to collapse it is Skybox gets
acquired by Google, and then at some point after that
you get a call from somebody who wants to build
satellites to track methane emissions, this project called Methane and Set,
which seems like it is sort of what puts you
on the path to starting mewon. The company're running, Now,
(04:15):
how did that happen? What happened there?
Speaker 2 (04:18):
I got a call from some people I knew from
Skybox at the Environmental Defense Fund saying we think there's
an opportunity, you know, approaching the problem the way that
companies like Skybox did, to go build a system that
can map methane globally in the world. And this is
you know, a huge first order a greenhouse driver. It's
in the near term, it's the most important greenhouse gas,
(04:38):
even more important than CO two. And we don't know,
we don't there's no good source of data for actually
how much is getting omitted by all these different sources
where not quantitatively at least, And so, you know, that
was a really exciting prospect for me to be part
of that because it seems like a really important problem
that needed to be solved. It was also really interesting
to me that here we are in this world where
(05:00):
there's a global nonprofitia the Environmental Defense Fund, so philanthropically
funded nonprofit that is basically believes they can go out
and build old what you know, traditionally would have been
only the realm of a government, like a major government
project like a NASA mission or an e submission, And
so that just seemed like an extremely compelling prospect, like,
if you can do that kind of thing and apply
(05:22):
it in all these different ways, it's going to blow
the door open on the types of things we imagined
doing from space.
Speaker 3 (05:27):
And so for all those reasons.
Speaker 2 (05:29):
I was really excited to you know, be part of
Methane SAD and join the advisory board for it. And
we went through this long process of basically, for lack
of a better term, trying to recreate some of the
things about what we did at Skybox, but with this
totally different outcome in mind. And it was really hard.
And one of the primary reasons it was really hard
(05:51):
was that at Skybox everything was within our control and
we built everything ourselves. We literally, you know, we were
literally building the circuit boards that went in the satellites.
We wrote all the software, we were building the cameras
that went into the telescope.
Speaker 1 (06:06):
You know.
Speaker 2 (06:07):
Environmental Defense Fund does not have a deep aerospace engineering
function that can go build all this stuff. And so
they essentially went out to the market and they said, Okay, well,
companies like Skybox and Planet exist. There must be now
a rich ecosystem of suppliers in the market that we
can go buy what we need to go do this mission.
The long and the short of it is that is
(06:29):
not true. And so what they what we kind of
ran up against was the state of the overall ecosystem
was still very much like that old mainframe in space,
you know, billion dollar ecosystem or these little tiny cottage
industry startups that we're trying to build some small piece
of a component or a little bit of software or something.
(06:51):
And the thing about space is really the hardest part
in many ways is integrating all these things together and
making them work as a system that can ultimately deliver
you know, an answer or a solution to a problem.
And so you know, if you fast forward and we
spent like five years helping edf kind get up the
learning curve doing that, and it was I would say, ultimately,
(07:12):
you know, successful in the sense that a satellite got built,
it got launched, it collected data. It was way slower,
it was way more expensive, it was way harder than
everybody thought, and it really kind of oriented me towards
there's a huge opportunity here. Like you know, we learned
how to solve a lot of these problems at Skybox.
There's clearly a need and so we have this opportunity
(07:34):
to frankly flood the sky with you know, entirely new
types of sensors collecting data that could be incredibly impactful
on how we manage the planet. But the supply chain,
the ecosystem to go execute that efficiently. You know, when
we started Mwon, I felt had not mature, it had
did not exist at the state that it needed to
(07:55):
to realize that. And so that was a lot of
the kind of origin thought process that led us to
start the company, is we can go build an entity
that is really able to deliver on these really hard,
difficult type of missions, but in the way that we
did it at Skybox. So we're ten times faster and
ten times cheaper. And if we do that, it's going
to blow the door open on the types of applications
(08:15):
people are doing in space.
Speaker 3 (08:16):
And that's what we're ultimately trying to do.
Speaker 1 (08:19):
And so you create me on in twenty twenty one,
is it right? And the first project that I'm aware
of that you were working on is fire SAT. Was
that sort of part of the creation of the company.
Speaker 2 (08:34):
Well, it predated the creation of the company a little
bit in the sense that what it came out of
was there's a group of people I worked with at
Google when I was at Google and research that we're
looking at how to is kind of AI applications, so
looking at how they could take a lot of the
AI work going on and applied to very particular impact
(08:56):
applications like climate, like wildfire. One of the biggest problems
with wildfire is there's these data gaps. It's very hard
to understand just how little they know when they're making
decisions during a wildfire, like where the fire is, how
fast it's moving, how hot, it's burning, what's in danger.
(09:16):
Usually it is largely from reports coming in from boots
on the ground.
Speaker 3 (09:21):
They have firefires on the line.
Speaker 2 (09:22):
Did the end of the day come in and say,
this is what we saw, this is roughly where the
fire line was, and people are hand drawing it on maps.
Speaker 3 (09:28):
That's literally often like it's.
Speaker 1 (09:30):
Nineteen fifty or at nineteen hundred.
Speaker 3 (09:32):
Nineteen hundred.
Speaker 2 (09:33):
It really is not that different than I think the
way wildfires were fought in nineteen hundred. And you know,
if you think about most elements of society, most elements
of our lives, they've progressed dramatically past that point. You know,
we could pull up a map on our phone that
has incredible near real time date on traffic.
Speaker 1 (09:52):
I mean, let me ask us a dumb question about that.
They fly planes over wildfires. Once these images of that,
can the planes not give them pretty good data about
what's happening? At least once they know it's there.
Speaker 3 (10:05):
They can.
Speaker 2 (10:05):
At times, the planes are very limited, so they they're
off not position in a place where they can get
them to an active fire fast enough to get the
data they need. So there's all those challenges, and they
often can't fly due to the conditions too, so if
it's too smoky, they can't fly. If there's really high winds,
which was obviously the case in the LA fires last year,
so they weren't able to fly the aircraft over those
(10:26):
fires because the winds.
Speaker 3 (10:26):
Were too high.
Speaker 2 (10:28):
So they can provide very good situational awareness data, but
they are very limited in a lot of ways, both
from a cost per perspective, from sort of a breadth
and geographic passy perspective, and then just from an operational perspective.
The thing that satellites give you that is unique as
this kind of persistent overhead view.
Speaker 3 (10:46):
Right A satellite's giving you this.
Speaker 2 (10:48):
You know, constant kind of persistent overhead map like view
of where the fire boundary is, how hot it is,
how fast it's moving, and that's just, you know, really
transformational for detecting the fires early. You don't have to say, hey,
I'm going to go fly a plane and look for
a fire. The satellite's going to cover the space and
detect a fire if it starts. And that's one of
the goals of firesat is to build to attack to
(11:09):
fire within about twenty minutes of its starting, and then
once it's burning we have this you know, ongoing time
series of understanding of how it's moving, which is going
to be absolutely transformational in terms of how they think
about positioning resources, how to manage risk. The other thing,
the other part of this that I think is important,
is that the goal is not to eliminate fire like
(11:32):
fire is really important. In fact, we don't have enough
fire in many in California, we don't have enough fire.
Historically there was a lot more natural wildfire, and we've
kind of eliminated that.
Speaker 1 (11:43):
Just to be clear about that, like the ideas, California,
like much of the world, all of the American West,
evolved to burn, to happen right, fires periodically, and we
had this century essentially of not letting anything burn that's right,
and you get too much fuel, too much underbrush, and
then when there is a fire, it's too hot and
it burns too big in a way that it would
not have if we had let fires burn in the
(12:05):
first place.
Speaker 3 (12:05):
Perfect, man, you've clearly studied this topic.
Speaker 1 (12:08):
Yeah, it's really interesting, right, it's not entirely intuitive, but
not into about it and it's like, oh interesting, Yeah.
I spend a lot of time in the Sierras where
It's like you see like Manzanita are like pumping out
oil because they want there to be a fire.
Speaker 2 (12:21):
Yeah, and Manzanita is like a blow torch when it
starts on fire. So yeah, so yeah, I mean that's
a huge problem. And I think the point is that
it's not simple enough to say when a fire starts,
put it out. There has to be a decision made
about there's you know, the way that this has best
been described to me. We work with an advisor named
Kate Dargan. He used to be the fire chief of California,
(12:42):
and she talks about good fire and bad fire. Right
when you get to what you were describing, where you
have these over fuel loated forests that burst into flame
and are burning basically so hot that they're killing everything
in a very unhealthy way.
Speaker 3 (12:54):
That's bad fire, and like you want it.
Speaker 2 (12:57):
You don't want that to burn, But we need a
lot more low intensity fire through the forest to keep
the forest healthy. And so right now, there's basically no
way for the fire agencies to distinguish that for the
most part. I mean, they're typically because of the risks involved,
they're just going to try and put the fires out
because there's so much at risk.
Speaker 1 (13:16):
Well, like politically, if you think about it, politically, you're
never going to get in trouble for putting the fires, right,
But if you're like, well, you know we made the call,
and you know our model said it should burn, like,
that's just not going to play if anything bad happens.
Speaker 3 (13:31):
That's right exactly. Yeah.
Speaker 2 (13:33):
Kate's big thing is if we can measure the intensity
on a near real time basis, we can make way
better decisions about how we deploy resources to combat the
bad fire while allowing the good fire to burn. And
that's that's a very heavily kind of risk based decision
process that needs to be much better informed by data.
And I think that's in many ways we love fire
(13:56):
Set because I think it is such a tangible example
of the power of these space based data collection systems
to not only tell us something about the world or
you know, oh that's interesting, but actually drive decision making
on the ground. We now have actionable data that can
be used to say we're going to position firefighters here
(14:16):
instead of there, and that's not only going to protect this,
you know, this town, but it's also going to allow
this other part of the forest to burn.
Speaker 3 (14:22):
So there's less risk next year that it's going to burn.
Speaker 1 (14:25):
Right. So you say we now have actionable data, but like,
we don't literally have it now, right, because the project
is still ongoing. And so just to be clear, the
goal of the project is to have So it's an
infrared image, right, which makes sense of every piece of
the earth out of resolution. You said about the size
of a swimming pool, right, which is a nice image.
Speaker 2 (14:47):
Yeah, the actual resolution is a lot bigger than that.
It's about fifty meters, so think about half of a
football field. But we can see we can detect a
fire within that fifty meters down to about the size
of a swimming pool. So if there was a swimming
pool size fire on half a football field, we would
detect that fire.
Speaker 1 (15:05):
Okay, and to do that of the whole earth every.
Speaker 3 (15:08):
Twent the minutes, that's right.
Speaker 1 (15:10):
It sounds like a lot.
Speaker 3 (15:12):
It is a lot of year.
Speaker 1 (15:13):
It is a lot, right. And so so you start
this company five years ago around the time this project
is kind of coming into view, right, and you the
company and the project sort of get going together.
Speaker 3 (15:25):
Yes, that's right, yep.
Speaker 1 (15:28):
And so I mean you talked about how you had
solved a lot of these things at Skybox. But plainly
there's a lot to figure out here. So when you
launch the company, what do you have to figure out
to do this?
Speaker 2 (15:38):
Yeah, so there's a couple of things. I think one
thing is be able to build a small and low
cost infrared camera, which is a critical enabler for this
to work. And that's something that we bid off very
early in the process, very much using again the same
kind of approach that we use at Skybox.
Speaker 3 (15:55):
At Skybox, we were.
Speaker 2 (15:56):
Much we were focused on visible imagery, which is a
different thing that's very much a solved problem now for space.
But if you ask anybody when we started the company,
what does it take to build like a very high
sense it'sivity infrared imager from space, they'd probably tell you
it's like a ten to thirty million dollar problem per unit.
Speaker 1 (16:15):
Wait, you mean it's going to cost you ten million
dollars for each camera?
Speaker 2 (16:19):
Yeah, I mean if you went out to Lockheed or
Northrope or whoever builds these things, because you know commercially
they're not built.
Speaker 3 (16:25):
Yeah, that's what they would have told you. That's what
you know.
Speaker 2 (16:27):
In many ways, like the government is paying today for
that type of payload.
Speaker 3 (16:32):
And we couldn't do.
Speaker 2 (16:33):
That because, you know, ultimately, to get to that twenty minutes,
which everybody agrees was a critical goal, you need like
fifty satellites and so if you're if just the camera
on each satellite costs twenty million dollars, it's pretty hard
to imagine that making sense.
Speaker 3 (16:46):
But we needed like a ten x.
Speaker 2 (16:48):
I mean, we needed to get this thing down to
like a million dollars or you know, no more than
two million dollars. And so again we had to go
to the out and say, like, what are technologies that
have been developed for other industries that we can apply
to this problem that haven't been used in space before,
because typically there's a lot of industries that have spent
billions of dollars on technology and getting the cost down
(17:10):
and the performance up. And so what we found was
there was there had been work on It was largely
for tactical military applications, so not things that go in space,
but things like the night vision that a soldier wears
in the field, where they're building hundreds of thousands of
these things at fairly low cost, and the underlying camera
technology is applicable to it's what we needed for this
(17:33):
it's a very different way to do it. Again, it
was another case of like we literally had people saying
this will never work, you can't do it that way.
Speaker 1 (17:40):
While as you say that, well, I mean when people
say that too.
Speaker 2 (17:43):
I love it when people tell me something's not possible,
because that's just like that's a great challenge, like Okay,
I'm gonna prove you wrong. So yeah, So that was
a big challenge that we had to solve and we did.
I think another challenge that we really had to solve,
which is related to fire set, but it's also related
to the bigger picture for the company is you know,
(18:05):
these things produce enormous amounts of data, and they often
are constrained in various ways in terms of the ability
to get that data back to Earth and then into
the hands of people that need it, and certainly on
reasonable time scales, so like typical satellite systems maybe only
might take an hour and a half or two hours
from the time a picture is taken to when it's
(18:27):
actually usable in somebody's hand.
Speaker 3 (18:30):
And again for the fire.
Speaker 2 (18:30):
Problem, obviously that's not going to work if we're going
to put all these satellites up to fly over every
fifteen minutes, Like, waiting an hour and a half for
somebody to get that data is not helpful. So we
had to put a lot of time and effort into
thinking about how you can build extremely high bandwidth, extremely
low latency pipelines basically to get data from satellites and
out to users. And the reason I say it's more
than just fire sat is because I do think that's
(18:52):
something that is an unlock for a whole bunch of
other things too.
Speaker 1 (18:56):
So how'd you figure that one out?
Speaker 2 (18:59):
So there's kind of two different pieces of it. There's
kind of a phase one and a phase two. The
phase one is clever, you know, basically clever archets picture
of where we put antennas around the world, and how
we design our radios to be able to continuously downlink
while the cameras are collecting, which most satellites don't do.
Most satellites collect some data, store it, down link it.
(19:22):
Our system is designed to be able to continuously do both,
and to do it to antennas that are basically placed
in geographies where wildfire really matters, so it's a near
real time connection all the time.
Speaker 3 (19:33):
That's kind of phase one.
Speaker 2 (19:35):
So the other thing kind of the phase two of
this that we're now bringing on board is we have
this partnership with SpaceX where we'll basically be connecting these
satellites into the Starlink network in space.
Speaker 1 (19:44):
Is that the laser part? Are you not saying lasers?
Speaker 3 (19:47):
It is lasers and space. Lasers and space. Yeah, well,
lasers in space.
Speaker 2 (19:52):
So now rather than having to wait to come over
a ground station or try and place one in exactly
the right spot to keep that latency down as we're
collecting data and space, we will have essentially what's like
an always on broadband connection to the Internet and we
can just continuously stream that data back for users. That's
completely transformative, right in terms of how you think about some.
Speaker 3 (20:12):
Of these problems.
Speaker 1 (20:13):
So where is fire SAT now?
Speaker 2 (20:16):
So we launched what we called the protoflight satellite, so
a first instance last year and it's up collecting data.
We're launching the first three of what we call the
operational constellation in June and then basically we have a
you know, a roadmap of launches that gets us out
to the first kind of big like milestone we're shooting
(20:37):
for is hourly So having a picture every hour of
the entire planet, which we should get to in sort
of late twenty seven early twenty twenty eight, and there
will ultimately be about twenty in that hour hourly constellation.
And then after that the goal is to expand that
ultimately to fifty two.
Speaker 3 (20:55):
That'll get us to fifteen minutes.
Speaker 1 (20:57):
Why might it not work?
Speaker 2 (21:00):
Yeah, So one of our biggest worries and concerns, because
there's a lot of history of this is up to
by the operational community.
Speaker 1 (21:12):
And I mean if there's no customers essentially.
Speaker 2 (21:14):
Yeah, our users, or if it's difficult for them to
incorporate into their operational flows, and.
Speaker 1 (21:20):
So somehow it's not actually helpful in fighting fires or
understanding fires.
Speaker 2 (21:24):
Right, And I think we're very confident that the data
itself is helpful, but it's only helpful, you know, if
you can get you know the term the water to
the end of the road, right, So it's not helpful
if all this data ends in some cloud bucket somewhere
that nobody can get to. It's only helpful if it
actually ends up out in these operational command centers, in
(21:45):
the tools that the fire operators are using on a
daily basis. And so that's actually not a trivial problem
because it's a huge patchwork of different systems and tools
even in the US, like there is no unified system.
So like our priority earliest integration partner is CalFire For
(22:06):
that reason, CalFire is the most sophisticated fire agency.
Speaker 3 (22:09):
They have the best tools to do this.
Speaker 2 (22:11):
But you know, ultimately we want to do a lot
more than that, Like, yeah, this is not just for California,
it's not just for the US, it's for the world.
And so that long tail of how do you actually
get the water to the end of the row with
the users that need it, I think is actually the
heart and not the hardest problem, but it's one of
the harder problems, and I think it's the biggest risk to.
Speaker 3 (22:30):
Ultimately having impact.
Speaker 2 (22:32):
Like, you know, if we get all these satellites in
space and they're wonderful and they're producing data, but we
don't solve that problem, it's all for nothing.
Speaker 1 (22:39):
Yeah, that could totally happen.
Speaker 2 (22:42):
Well, we're going to try and make it not not
a question.
Speaker 1 (22:45):
Yeah, not technically a question to say that could totally happen.
We'll be back in just a minute. Hey, it's Jacob,
and I want to tell you that I am hosting
a new show called Business History. It's about the incredible
(23:05):
innovations and massive failures and unbelievable characters in the history
of business, and I hope I think the show provides
insights about how business works today. At the end of
today's episode of What's Your Problem, We're going to play
you a clip from Business History. It's the story behind
the video game company Atari, and they're surprising early hire
(23:26):
of a young hippie named Steve Jobs. The show's called
Business History. You can listen to it wherever you're listening
right now, and we'll play that clip at the end
of today's episode. Tell me about your work with the
Department of Defense.
Speaker 3 (23:42):
Yeah, So there's a couple of worlds out there right now.
Speaker 2 (23:48):
There's the commercial world, and I would put fire sat
in that.
Speaker 3 (23:51):
World, and then there's sort of the national security world,
and they're related in a lot of ways.
Speaker 2 (23:57):
It's kind of hard to separate them completely, but there
is right now a I would say existential need, especially
in the US, for new technology and space. And the
way I would think about this is that we've spent
sixty or seventy years building very expensive, high performance, exquisite
(24:18):
systems for space. On the national security side, most people
don't even realize exist, and they're absolutely absolutely critical to
our country's ability to operate in the world, like in
a way that you know, we're very very sensitive too.
So historically we have dominated space as a country that's
(24:40):
not really been contested, and so this has been both
a superpower but also sort of a capability. It's kind
of taken for granted. The armed services don't even think
about the idea that some of these things, these tools
that they have that they rely upon, would not be
available or useful.
Speaker 1 (24:56):
Like that some enemy might shoot them down exactly what
you're alluding to, Yeah, or jam them or jam them, right,
I guess jam them is actually what happened.
Speaker 2 (25:05):
Well, I mean the Chinese shot a satellite down a
couple of years ago, right to demonstrate they can. So
like there's a lot of concern about that. So right
now we're in the midst of this period where those
old systems have were designed in an era where there
weren't threats. There's now threats. There's also a lot of
like China has become much more capable in space and
(25:27):
it's now more like a peer to us. Versus like
a strong second.
Speaker 3 (25:32):
And so there's this.
Speaker 2 (25:33):
Huge push to basically modernize and refresh all of the
kind of space infrastructure that the US uses on the
national security side, but very much leveraging the commercial approach
that we've been talking about that I think we kind
of pioneered at Skybox a decade ago, and at the
time it was perceived as too risky and too early,
(25:54):
but there's now a recognition that that is the future,
this is how these systems need to be built, and
that there is very quickly becoming this vibrant commercial industry
that can can build a lot of these systems.
Speaker 1 (26:06):
Analogous to what SpaceX did for rocket rockets used to
be whatever, three two whatever. It was one military contractor
and then a private company game. It was like, look,
we can do this cheaper, right, That's happening for satellites
now for the government, for the military.
Speaker 2 (26:20):
Yeah, And very similarly, SpaceX, you know, did it in
a way that was commercially oriented. They're launching commercial communication
satellites but also supporting national security missions. So I think
that's space inherently is very difficult to separate from that perspective.
Speaker 1 (26:38):
Separate private from the military essentially, Yeah, but just because
that's where the money is.
Speaker 2 (26:45):
That's where the money is. But also because anything you
put in space inherently has the ability to do both, right,
I mean in the limit, if you really want to
think about the limit. And it's interesting just to make
this make this point pretty clear, China does not really
distinguish between commercial satellites and military satellites. We historically have
(27:05):
distinguished that. But one of the big fears more and
more is that that you know, our government should be
depending on commercial satellites. Do those then become targets in
a conflict?
Speaker 4 (27:16):
Right?
Speaker 3 (27:16):
And so there's it's a little another analogy.
Speaker 1 (27:19):
I think that's a good thinking of the Lusitania.
Speaker 2 (27:22):
I I was gonna say naval like kind of pre
World War two naval conflict, right, birch and ship can
be used for goods, but it can also be used
as smuggle weapons and and you know, in a lot
of ways, our navy after the Second World War sort
of police the world's oceans to kind of guarantee a
lot of the kind of separation between commercial and military
(27:45):
affairs in the ocean. There's no equivalents of that in space.
And so that's I think that's a big concern people
have right now, is that even if you want to
say this is a commercial system, this is a military system,
it's very hard to draw that line in a way
that China will respect it or some other adversary will
respect it. And so all to be said, the technologies
(28:06):
are very similar. The technology are the same, right you know,
if you put something in space, it's it's very hard
to tell what it's for for somebody else looking on
like it's very easy for.
Speaker 1 (28:17):
It's just a satellite. It's satellite country taking a picture.
Speaker 2 (28:21):
Scout radios and sensors on it. Are those radios and
sensors communicating to commercial stuff? Are they community in the military?
Are they taking pictures of you know, people's trampolines in
their backyard. Are they taking pictures of military sites? It's
very hard to separate. And I think we're seeing a
big convergence right now in and a recognized convergence of
those those spheres. And so there's you know, there's both
huge opportunities and risks with that. You know, the opportunity
(28:43):
for us is that the things we're building are applicable
to both There's a huge need on the national security side.
And to give a very specific example of this, you know,
we want a contract with the Space Force last year
to basically take what we developed for fire SAT, this
infrared camera, this low cost infrared camera, and adapt to
a mission set that they've had for forty years that
(29:06):
has been very challenging for them to upgrade, which is
what they call They call it theater imagery and cloud characterization,
but it's basically weather forecasting. So the DoD has to
forecast weather. If they want to fly planes, they need
to know what the weather is, so they have their
own weather forecasting capability, and they have satellites to feed that.
And historically these have been these billion dollar, twenty year
(29:27):
bespoke things. And what we've been able to identify as
the same technology we've built for fire Sat is very
applicable to deploying satellites to solve that mission for them.
So I think that's a really good example of kind
of this crossover and how and how inherently dual use
a lot of these things.
Speaker 1 (29:45):
Are you were saying there's opportunities and risk. That's the opportunity.
What's the risk for you?
Speaker 2 (29:50):
I think there's like a perceptive risk, and I think
there's a physical risk. The perceptive risk is. You know,
we started the company very much to do positive in
the world, and I firmly believe that's what we're doing.
I think Fire SAD is a canonical example of how
we're trying to do this. I believe there's a lot
of things in national security sphere that we can do
(30:12):
that also have that characteristic that are positive things for
the world.
Speaker 3 (30:17):
Not everybody agrees with that.
Speaker 1 (30:19):
Yeah, well, and I mean, you know, somewhat famously, at
other big tech companies, right, people who worked at the
companies didn't like it when Google and Microsoft started doing
work for the Department of Defense.
Speaker 5 (30:31):
Right.
Speaker 2 (30:31):
I was at Google when the Project Maven thing happened,
like when there was the big backlash against their work
on that. So, like, have you encountered any of that
at MWON. Yeah, No, We've definitely had people with concerns.
I mean, I think that what we've tried to do
is just be very open about why we're doing it
and why we see it as a positive addition to
our mission set. And for the most part, that's been
(30:52):
very successful. I mean, people typically are engineers, are very
rational people usually, and so I think if we can transparently.
Speaker 1 (31:01):
Present rational I mean you rational people can disagree about
the most prevailance of the American Department of Defense.
Speaker 2 (31:08):
Oh, you're right, And so I mean again, like my
perspective on this is like I never want to jam
this down anybody's throat, Like I think people need to
make their own decisions and be very comfortable with what
we're doing, and we're just trying to be as transparent
about that as we can.
Speaker 1 (31:22):
Let's talk now about the broader state of the industry,
of the space industry, such as it is. I mean,
it's particularly interesting given that you have seen now this
rather long arc, right when there has been this big
transformation in the industry over what fifteen years or twenty years? Yeah,
twenty years. So you know, how has it changed? That's
(31:45):
the big dumb question. Like, yeah, I mean there's the
obvious stuff that you've done, but like, what can you
do now that you couldn't have done?
Speaker 4 (31:50):
Then?
Speaker 1 (31:51):
What's you know, what's interesting, what's surprising.
Speaker 2 (31:53):
So I'll give some anecdotes. So I interned at SpaceX
in like two thousand and four. I was there's like
forty people there, and you know, my first job was
helping build the first Merlin, which is like the first
stage engine for Falcon one, and testing it. And you know,
this was kind of when I was there, there was
(32:14):
this excitement about how space This was very early, right,
there's forty people, and there was this belief that they
were going to totally transform the industry that early. You know,
I learned very quickly that we were a long way
away from that. The first time we put this engine together,
probably tore itself apart, burned the test stand down, and
you know, I spent the rest of the summer like
(32:35):
rewiring this test dan in the summer heat in the
Texas eat and it took him another like five years
before they even got a first rocket into space after that,
And if I fast forward from there and then think
about starting Skybox in two thousand and nine, which is
roughly at the same time, again we had this perspective
that it's like, okay, well it took a while, but
(32:56):
SpaceX is solving this problem. They solve this launch problem
where access to space is going to be easier. They
now have Falcon one. We actually were going to buy
two Falcon ones. It's Skybox. Early on and right at
the last minute, Elon said, there's not a Falcon one
is not business competitive.
Speaker 3 (33:12):
We're gonna build a bigger rocket.
Speaker 2 (33:13):
We're gonna build Falcon nine, and all the contracts that
we're gonna do for Falcon one are now Nolan bought
Void and so this was existential for us at Skybox.
Speaker 3 (33:20):
We ended up having.
Speaker 2 (33:21):
To go to Russia, of all places, and buy these
converted Cold War era ICBMs to launch our satellites with.
That's we were literally launching our satellites out of a
missile base in southern Russia off ICBMs.
Speaker 3 (33:35):
And so again like.
Speaker 2 (33:36):
We were like, okay, well, this problem we thought was solved,
it was going to blow the door open, is still
not solved. It is now solved fundamentally. And I mean
there's all kinds of stats that show this from the
cost of launch that has come down, and launch launch
has been the big choke point for most of the
history of space. So that's a huge unlocked.
Speaker 1 (33:57):
Hard to get there space hard to get there now,
not as hard, not as hard.
Speaker 2 (34:01):
I don't think you can underestimate the impact of solving
the launch problem. I'll call it in terms of how
the world opens up, in terms of what you can.
Speaker 3 (34:11):
Do with space.
Speaker 2 (34:12):
The other thing I would say is I think we're
even when we were doing Skybox and thinking very differently
about how to build satellites, we were still building in
a silo. We were building this thing that the analogy
I like to use most of stuff in space is
kind of like species on the Galapagos Island. They're like
these pieces of technology that have been like separated from
(34:35):
the mainland for so long. They've kind of evolved in
these like perverse ways.
Speaker 1 (34:39):
And if you actually like this's whatever, like the motors
and the cameras and whatever, they're like, oh, I understand
how this used to be like a motor or camera
on Earth, But now it's totally different, totally.
Speaker 2 (34:48):
Like literally to the point where if you looked at
like the computer on even a satellite that got launched
ten years ago, it would look like something that TI
put in a calculator in the nineteen seventies, right, I mean,
it's crazy. So there's this kind of like frozen in
time nature to a lot of the technology in space,
and because of the launch, but also because of these
things that we've been talking about, there's a big turnover
(35:09):
where now the technology in space is looking a lot
more like the stuff that goes into data center, where
that goes.
Speaker 3 (35:14):
In your smartphone.
Speaker 2 (35:16):
And because of that, there's also a lot more interconnectedness
between those technologies and the technologies on Earth. And I
think we're just in the early stages of this and
it's going to cause it's already causing, but it's going
to cause this huge convergence in what we think about
as space technology and every other bit of technology in
our life, to the point where we're going to realize that.
(35:37):
You know, right now, I do I use space technology
on here GPS, right, satellites are in space and when
I look at my map, it's telling me where I
am based on satellites and space. But nobody thinks about
those satellites, you know, when they look at their map
and ask where they're going. That's going to be true
of like a whole bunch of different things, Okay, what
things and when one of the most exciting things for me,
(36:00):
there's been a lot of discussion about like orbital data
centers right people are talking about orbital data centers right now.
Speaker 1 (36:04):
Yeah, And just to be clear, like an actual like
AI data center that is some giant.
Speaker 2 (36:09):
Gigawatts of solar power and space. Yeah yeah, yeah, and
it sounds crazy.
Speaker 3 (36:14):
I think it's going to happen.
Speaker 1 (36:16):
Okay, when and why?
Speaker 3 (36:19):
So the why there's there's a bunch of reasons right
now to build a data center takes between five and
seven years okay, between permitting, construction, power, interconnect, all of
these other things.
Speaker 1 (36:31):
I love that it's permitting, permitting, permit will drive data
centers to space.
Speaker 2 (36:36):
It seems crazy, but I mean it's true, right and
so like again, and I'm not saying this is easy,
but if you could get a factory set up and
rockets launching frequently enough that you can go deploy that
one hundred megawatts of data center to space in two
years instead of five years, there's a huge time value
money component to that.
Speaker 1 (36:53):
That's it's so wild. What give me the over under?
What year is that going to happen?
Speaker 3 (36:57):
Twenty thirty five?
Speaker 1 (36:59):
That soon? That soon?
Speaker 3 (37:01):
I think that scale could happen that soon and.
Speaker 1 (37:03):
Is part of what has to happen there, Like it
keeps getting cheaper to go to space absolute kilogram.
Speaker 3 (37:09):
Yeah.
Speaker 2 (37:09):
So so like we've done a bit of work on
this with some hyperscalers.
Speaker 1 (37:13):
So they've come to you and asked hyperscalts is whatever.
This is four giant tech companies. They're all actually calling
you and saying, yeah, I mean it take to put
a data center in space.
Speaker 3 (37:23):
Google has announced it publicly.
Speaker 2 (37:25):
Okay, so you know the cost per kilogram to put
something in space is the key metric. And I think
for economic parody, if you think about, like how what
it would take to get to a similar cost per
unit capability of compute and a data center on Earth.
Right now, we're still off by about a factor of six.
(37:46):
So like it's six, but it's not one hundred, it's
one hundred. It's not even an order of magnetus.
Speaker 3 (37:51):
It's not even an order of magnitudes.
Speaker 5 (37:53):
And so is the.
Speaker 1 (37:54):
Reason it's cheaper once you're up there? What the energy permitting?
Like why is it? Why is it only six? So
it only effect us.
Speaker 2 (38:03):
So here's here's a really important factor. If you put
a satellite into space in the right orbit, you can
have solar rays that are in this sign one hundred
percent of the time. So it's baseload power. Now it's
solar is no baseload power. I mean, solar is extremely
cheap right now, but you can't use just solar to
power data center because it's not consistent, so you end
(38:23):
up needing massive some other form of massive energy storage,
and it.
Speaker 3 (38:27):
Has become very expensive. So I think that's a big deal.
Speaker 2 (38:30):
I think that's one of the reasons in my opinion,
that it at six x you know, it's cost competitive
because it's when you actually factor in all those other
costs to terrestrial data centers for things like power, it's
actually higher than you would think. Okay, So again, the
other thing I just think is an important point is
and I think this is being missed a little bit, like,
(38:52):
we don't go from here to gigawat data centers in space. Right,
there's this huge spectrum between where we are now and that,
and I have no doubt we're going to be going
way down that spectrum over the next ten years. Right,
so where it actually lands, Are we actually going to
have gigawat data centers in space?
Speaker 4 (39:09):
Like?
Speaker 3 (39:09):
Who knows? Right, who knows if the models will continue scaling.
Speaker 1 (39:13):
The scaling laws will hold.
Speaker 3 (39:14):
Yeah, there's a stuff. What's the midpoint?
Speaker 1 (39:17):
What's the like easier to imagine launches don't have to
get that much cheaper. We're almost there now, but still
cool thing yeah.
Speaker 2 (39:25):
I mean I think it's that there are these kind
of multi tiered layers of what we would think about
as like cloud infrastructure in space, right, so very high
speed network. So now you have this high speed network
that everything is connected to all the time in space,
which is not the case.
Speaker 1 (39:38):
Now what we thought of a data center ten years ago,
like where they're like whatever.
Speaker 2 (39:45):
You know what you're getting Netflix served from? And I
mean Starlink's already doing this a little bit. Starlink cash's
Netflix on the satellites, right. And what will happen is
as that builds out, there will be reasons to move
certain things that are in the cloud on data centers
and Earth now up to space to make the latency
lower or the processing more efficient or whatever. But we
won't even think about it. That'll just be in the background.
(40:08):
We'll use the thing and nobody will know whether it's
happening in space or happening in the ground.
Speaker 1 (40:15):
We'll be back in a minute with the lightning round.
Let's finish with the lightning round. What's the secret to
having one of the best breaking balls in the Stanford bullpen?
Speaker 3 (40:39):
Oh man, it's top spin.
Speaker 1 (40:46):
What's your favorite satellite that you didn't build.
Speaker 3 (40:51):
The Corona satellites.
Speaker 1 (40:53):
Tell me about the Corona satellites.
Speaker 2 (40:55):
So the Corona satellites were the first imaging satellites we
built in the early sixties. And they were film based systems,
so they had film in them that ran through them
that you know, this incredible film technology that ran through
them that you know, one hundreds defeat per second to
keep up with the speed that they were flying. And
that would get loaded into these canisters and dropped off
(41:17):
in reentry vehicles parachute down to Earth and then we
would go pick them up by scooping them out of
the air behind a plane.
Speaker 3 (41:23):
To go develop the film and distribute it.
Speaker 1 (41:26):
Why was I just reading about that? I was reading
about the special planes, right, they made special planes to
catch the canisters as they're.
Speaker 3 (41:33):
Coming down, these like tohoks.
Speaker 1 (41:35):
It's so wild and like physical, right. You're a mechanical engineer, right,
that's like very yeah, so mechanical, yeah, yeah.
Speaker 2 (41:41):
I mean, and you look inside these things and there's
just like all these rollers and motors and stuff to
like move the film around and develop it and you know,
put in the.
Speaker 3 (41:48):
Canister and drop the canister off.
Speaker 1 (41:50):
So analog, right, it's just all physical stuff. What's something
you learned working on autonomous vehicles that lift that's been
useful to you at new.
Speaker 3 (42:00):
Yeah, I mean I actually thought about this a lot.
Speaker 2 (42:03):
The way that robotics and av used data is totally
revolutionary from somebody that's come from aerospace because of some
of the things we talked about about how primitive a
lot of historically, like compute on aerospace systems is, you
end up having to be very conservative about data and
(42:26):
processing and all of these things. And avs are one
hundred percent the opposite. They're just these like data vacuums
and so, but what it does is it means that
you think about these problems totally differently, like solving problems
totally differently.
Speaker 1 (42:39):
So what does it mean to you in your work?
How do you use that?
Speaker 2 (42:42):
I mean by architecting our systems to look more like
avs than satellites.
Speaker 1 (42:48):
Just to be able to handle more data.
Speaker 2 (42:50):
Yeah, I mean we can use the same technology. So
we are putting massive amounts of compute and networking and
you know, bandwidth on our satellites, and so we can
do that.
Speaker 1 (43:02):
Will evidence of extraterrestrial life be discovered in your lifetime?
Speaker 3 (43:06):
Yes?
Speaker 1 (43:08):
Okay, Uh, there is there anything you know that I
don't know that makes you.
Speaker 3 (43:12):
Say that, No, I just think that.
Speaker 2 (43:16):
I think I think we'll discover very simple extraterrestrial life.
I think there's life arose so early in the history
of the planet, simple life that it just seems like
it's very likely to have risen in other places or
have been transplanted here from other places.
Speaker 3 (43:34):
I think complex life is a lot harder.
Speaker 5 (43:36):
Like.
Speaker 3 (43:36):
I do not think we will discover complex life in
my life.
Speaker 1 (43:39):
We'll find like bacteria on a moon of Jupiter or something.
Speaker 3 (43:43):
Yeah, I think like single cell procaryot like things.
Speaker 1 (43:46):
Yeah. Will you go to space before you die?
Speaker 3 (43:55):
I would love to? I think probably not.
Speaker 1 (43:59):
Why not?
Speaker 2 (44:02):
I'm not convinced there's great reason to send humans to
space A lot I think. I think there's a lot
of things we can do without sending humans to space,
all right.
Speaker 3 (44:11):
I think where we do, it'll be limited.
Speaker 1 (44:14):
Why would you love to?
Speaker 2 (44:16):
I'd love to see Earth from space? I think the
overview effects you I've heard that astronauts talk about the
total transformation and perspective that comes from seeing the Earth
from space.
Speaker 3 (44:28):
So I'd love to experience that.
Speaker 1 (44:36):
Johnny Dyer is the co founder and CEO of muon Space.
Today's show was produced by Gabriel Hunter Chang. It was
engineered by Sarah Bruguire and edited by Lydia Jane Kott.
I'm Jacob Goldstein. You can find me on x at
Jacob Goldstein. You can find me on LinkedIn. You can
also email us at problem at pushkin dot FM. We'll
(44:58):
be back next week with another episode of the show.
I'm Jacob Goldstein and right now we're going to play
you a clip of a new show that I co host.
The show's called Business History. My co host is Robert Smith,
(45:20):
and this clip is from an episode we did about
how Nolan Bush, now a stoner turned entrepreneur, created the
video game company Atari and hired a young, inexperienced Steve Jobs.
I really hope you like the clip, and if you
want to hear more, you can find Business History wherever
you're listening to this show. Right now, this like nineteen
(45:42):
year old hippie kid walks in and he says he
won't leave until they give him a job, and the
receptionist calls the head engineer and she goes, yeah, we
got a hippiekid in the lobby says, he won't leave
until we hire him. Should we call the cops or
let him in? And the engineer says.
Speaker 4 (45:59):
Bring him on in.
Speaker 5 (46:00):
It's nineteen seventies, a Silicon valley, and this guy wanders
in and he is whenever you tell a story like this,
you know who it is.
Speaker 4 (46:07):
It's Steve Jobs.
Speaker 1 (46:08):
It's Steve Job fun. It's so delightful and perfectly. Steve
Jobs is very good at his job, yes, and very
unpleasant to wrqu with. On surprising, he tells Bushnell that
like everybody's soldering wrong, right, they're actually putting together the hardware,
soldering the hardware.
Speaker 4 (46:25):
He's probably right, and bush Nell's like, yeah, he was right.
Speaker 1 (46:27):
He keeps calling his manager a dumb shit. Probably was
not kind, but not necessary. Bushnell winds up putting Jobs
on the night shift, partly so he won't bother so
many people, and partly because he knew that Jobs like
to hang out at night with his buddy Steve Wozniak,
who was a great engineer, would be great to have
(46:48):
hanging around atari And in fact, Jobs and Wozniak helped
to make Breakout great a target. Remember Breakout, of.
Speaker 4 (46:55):
Course you're trying to knock down the bricks in a wall.
Speaker 1 (46:57):
Still games like that, when you got a little paddle
at the bottom.
Speaker 4 (46:59):
End, there's a moment when it goes through the wall
and then goes so.
Speaker 1 (47:04):
Good. So Jobs worked at Atari for a little while
and then decided he wanted to go out to India
to find his guru. Perfect asked Atari to pay for
the trip. Nobody ever said he lacked moxie and wound
up making a deal with Atari where they pay him
to go part of the way there. They had exported
some games to Germany and there was some kind of
problem with the games in Germany, and they're like, we'll
(47:26):
send you to Germany to fix the games and then
you can get the rest of the way to India.
And you know, the Germans said Jobs was terrible to
work with, but he fixed the games. I was thinking
about like the link between Atarian Jobs and what did
he learn there, and it felt like maybe a little overdetermined,
but I do think, you know, clearly he had this
profound sense of aesthetics and of delight, right, like think
(47:49):
of the Macintosh, right, this breakthrough Apple machine in the eighties.
It was round and instead of squares, it was rounded,
and it cost them more money, but it was beautiful
and it was fun and you were engaged with it
like a game.
Speaker 5 (48:01):
It almost looked a little bit like an arcade game
with the curves. Yeah, but more than that, I think
that especially at Silicon Valley at the time when they
were dealing in action Silicon. Right, if you're making chips
for somebody, you're thinking about the future and the computers.
You're not thinking about the psychology of the customer, because
the customer was another electronics company, right, And so Atari
and the strength of it was really the first time
(48:22):
where they're just like, how will a regular human being
who has no training whatsoever interact with technology?
Speaker 3 (48:29):
Yeah?
Speaker 1 (48:30):
So Atari now mid seventies, they're selling all the machines
they can make. They need more space, and so they
rent an abandoned roller rink and they turn it into
this office slash video game factory.
Speaker 4 (48:45):
I've been ten years older, I would have loved to
have worked in a roller rink video game factory.
Speaker 1 (48:49):
Yes, everybody smoked weed. Yeah, there was a hot tub.
There was a pool party where everybody ended up naked
in the pool at Bushnell himself looked back on it
later and said if that isn't a horror show for
any HR person today, I don't know what is
Pushkin. I'm Jacob Goldstein and this is What's Your Problem
show about people using technology to solve problems that matter.
My guest today is Johnny Dyer. He's the co founder
(00:35):
and CEO of a company called Muon Space. Newon Space
was founded in twenty twenty one, and the company designs, builds,
and operates satellites. Their first big project is a constellation
of satellites called fire Sat. And Johnny's problem is this,
how do you build satellites that can capture data from
(00:56):
space to help manage wildfires around the world in near
real time. In our conversation, Johnny and I also talked
about his company's work with the US Department of Defense,
and more generally, the global geopolitics of satellites, which basically
means the US versus China in space. To start, we
(01:17):
talked about a satellite company called Skybox. Johnny joined that
company just after it was founded in two thousand and nine.
Up to around that time, satellites had mostly been giant, expensive,
bespoke things, but this was a moment when a few
people were starting to think that satellites could get much
smaller and much cheaper. Johnny told me that the idea
(01:37):
was partly inspired by the emergence of the iPhone.
Speaker 2 (01:40):
You shrink cameras down enough, you shrink compute down enough,
all of a sudden you have this thing that is
possible in a much smaller package than was ever possible before.
And so there was a strong analogy. We thought with
satellites that if you can use these commercial technologies, you
can do a lot of the things that these big
billion dollar systems are doing in a much smaller package.
If you can do it in a smaller package, everything
(02:02):
gets faster, cheaper, because most of the challenge of getting
stuff in space is launching, and that on a per
kilogram basis, like it just costs per kilogram to get
stuff in space. If you make it smaller, everything gets better.
And so we, you know, we did this and we
were very successful at it. We ended up building twenty
one satellites that were you know, took data of the
(02:23):
same quality that literally, you know, one hundred million dollar
systems before us were collecting. We got bought by Google
and in the process we learned a lot about what
it takes to do this, what works, what doesn't work.
But also, you know, there there was sort of a
lot of kind of frustrations we had about the fact
(02:46):
that we solved this problem, this technical problem, and figured
out how to deploy these things in space, but we
really were only applying it to one small use case,
which was this high resolution visible imagery. Think about what
you see in Google Maps. So that's what we were doing.
We're taking like submeters roughly one meter resolution visible images
like you see Google Maps.
Speaker 1 (03:04):
So when I look on Google Maps and I see
the trampoline in my backyard, which I do, Uh, is
that you taking that picture?
Speaker 3 (03:11):
So probably the trampoline in your backyard.
Speaker 2 (03:13):
Probably not that was probably taken it from an aircraft
because that's probably more like ten centimeter imagery.
Speaker 3 (03:19):
But if you go to.
Speaker 2 (03:20):
A big trampoline, yeah, I mean you would see it
in our imagery.
Speaker 3 (03:23):
You would definitely see a trampoline.
Speaker 2 (03:25):
If you go a little bit outside a major urban
area where it's too expensive to go collect aircraft imagery,
all of that imagery and Google Maps would be the
type of stuff we were collecting. So you can go
search around Google Maps and you can see a bunch
of that stuff. And so you know, we got we
went through that journey. Eventually it ended, but stuck in
the back of my head was we've shown, you know,
(03:46):
we've shown a different paradigm for doing this, and there's
so many possible applications of that new paradigm.
Speaker 1 (03:53):
So so kind of the basic narrative of how you
get to muon just to collapse it is Skybox gets
acquired by Google, and then at some point after that
you get a call from somebody who wants to build
satellites to track methane emissions, this project called Methane and Set,
which seems like it is sort of what puts you
on the path to starting mewon. The company're running, Now,
(04:15):
how did that happen? What happened there?
Speaker 2 (04:18):
I got a call from some people I knew from
Skybox at the Environmental Defense Fund saying we think there's
an opportunity, you know, approaching the problem the way that
companies like Skybox did, to go build a system that
can map methane globally in the world. And this is
you know, a huge first order a greenhouse driver. It's
in the near term, it's the most important greenhouse gas,
(04:38):
even more important than CO two. And we don't know,
we don't there's no good source of data for actually
how much is getting omitted by all these different sources
where not quantitatively at least, And so, you know, that
was a really exciting prospect for me to be part
of that because it seems like a really important problem
that needed to be solved. It was also really interesting
to me that here we are in this world where
(05:00):
there's a global nonprofitia the Environmental Defense Fund, so philanthropically
funded nonprofit that is basically believes they can go out
and build old what you know, traditionally would have been
only the realm of a government, like a major government
project like a NASA mission or an e submission, And
so that just seemed like an extremely compelling prospect, like,
if you can do that kind of thing and apply
(05:22):
it in all these different ways, it's going to blow
the door open on the types of things we imagined
doing from space.
Speaker 3 (05:27):
And so for all those reasons.
Speaker 2 (05:29):
I was really excited to you know, be part of
Methane SAD and join the advisory board for it. And
we went through this long process of basically, for lack
of a better term, trying to recreate some of the
things about what we did at Skybox, but with this
totally different outcome in mind. And it was really hard.
And one of the primary reasons it was really hard
(05:51):
was that at Skybox everything was within our control and
we built everything ourselves. We literally, you know, we were
literally building the circuit boards that went in the satellites.
We wrote all the software, we were building the cameras
that went into the telescope.
Speaker 1 (06:06):
You know.
Speaker 2 (06:07):
Environmental Defense Fund does not have a deep aerospace engineering
function that can go build all this stuff. And so
they essentially went out to the market and they said, Okay, well,
companies like Skybox and Planet exist. There must be now
a rich ecosystem of suppliers in the market that we
can go buy what we need to go do this mission.
The long and the short of it is that is
(06:29):
not true. And so what they what we kind of
ran up against was the state of the overall ecosystem
was still very much like that old mainframe in space,
you know, billion dollar ecosystem or these little tiny cottage
industry startups that we're trying to build some small piece
of a component or a little bit of software or something.
(06:51):
And the thing about space is really the hardest part
in many ways is integrating all these things together and
making them work as a system that can ultimately deliver
you know, an answer or a solution to a problem.
And so you know, if you fast forward and we
spent like five years helping edf kind get up the
learning curve doing that, and it was I would say, ultimately,
(07:12):
you know, successful in the sense that a satellite got built,
it got launched, it collected data. It was way slower,
it was way more expensive, it was way harder than
everybody thought, and it really kind of oriented me towards
there's a huge opportunity here. Like you know, we learned
how to solve a lot of these problems at Skybox.
There's clearly a need and so we have this opportunity
(07:34):
to frankly flood the sky with you know, entirely new
types of sensors collecting data that could be incredibly impactful
on how we manage the planet. But the supply chain,
the ecosystem to go execute that efficiently. You know, when
we started Mwon, I felt had not mature, it had
did not exist at the state that it needed to
(07:55):
to realize that. And so that was a lot of
the kind of origin thought process that led us to
start the company, is we can go build an entity
that is really able to deliver on these really hard,
difficult type of missions, but in the way that we
did it at Skybox. So we're ten times faster and
ten times cheaper. And if we do that, it's going
to blow the door open on the types of applications
(08:15):
people are doing in space.
Speaker 3 (08:16):
And that's what we're ultimately trying to do.
Speaker 1 (08:19):
And so you create me on in twenty twenty one,
is it right? And the first project that I'm aware
of that you were working on is fire SAT. Was
that sort of part of the creation of the company.
Speaker 2 (08:34):
Well, it predated the creation of the company a little
bit in the sense that what it came out of
was there's a group of people I worked with at
Google when I was at Google and research that we're
looking at how to is kind of AI applications, so
looking at how they could take a lot of the
AI work going on and applied to very particular impact
(08:56):
applications like climate, like wildfire. One of the biggest problems
with wildfire is there's these data gaps. It's very hard
to understand just how little they know when they're making
decisions during a wildfire, like where the fire is, how
fast it's moving, how hot, it's burning, what's in danger.
(09:16):
Usually it is largely from reports coming in from boots
on the ground.
Speaker 3 (09:21):
They have firefires on the line.
Speaker 2 (09:22):
Did the end of the day come in and say,
this is what we saw, this is roughly where the
fire line was, and people are hand drawing it on maps.
Speaker 3 (09:28):
That's literally often like it's.
Speaker 1 (09:30):
Nineteen fifty or at nineteen hundred.
Speaker 3 (09:32):
Nineteen hundred.
Speaker 2 (09:33):
It really is not that different than I think the
way wildfires were fought in nineteen hundred. And you know,
if you think about most elements of society, most elements
of our lives, they've progressed dramatically past that point. You know,
we could pull up a map on our phone that
has incredible near real time date on traffic.
Speaker 1 (09:52):
I mean, let me ask us a dumb question about that.
They fly planes over wildfires. Once these images of that,
can the planes not give them pretty good data about
what's happening? At least once they know it's there.
Speaker 3 (10:05):
They can.
Speaker 2 (10:05):
At times, the planes are very limited, so they they're
off not position in a place where they can get
them to an active fire fast enough to get the
data they need. So there's all those challenges, and they
often can't fly due to the conditions too, so if
it's too smoky, they can't fly. If there's really high winds,
which was obviously the case in the LA fires last year,
so they weren't able to fly the aircraft over those
(10:26):
fires because the winds.
Speaker 3 (10:26):
Were too high.
Speaker 2 (10:28):
So they can provide very good situational awareness data, but
they are very limited in a lot of ways, both
from a cost per perspective, from sort of a breadth
and geographic passy perspective, and then just from an operational perspective.
The thing that satellites give you that is unique as
this kind of persistent overhead view.
Speaker 3 (10:46):
Right A satellite's giving you this.
Speaker 2 (10:48):
You know, constant kind of persistent overhead map like view
of where the fire boundary is, how hot it is,
how fast it's moving, and that's just, you know, really
transformational for detecting the fires early. You don't have to say, hey,
I'm going to go fly a plane and look for
a fire. The satellite's going to cover the space and
detect a fire if it starts. And that's one of
the goals of firesat is to build to attack to
(11:09):
fire within about twenty minutes of its starting, and then
once it's burning we have this you know, ongoing time
series of understanding of how it's moving, which is going
to be absolutely transformational in terms of how they think
about positioning resources, how to manage risk. The other thing,
the other part of this that I think is important,
is that the goal is not to eliminate fire like
(11:32):
fire is really important. In fact, we don't have enough
fire in many in California, we don't have enough fire.
Historically there was a lot more natural wildfire, and we've
kind of eliminated that.
Speaker 1 (11:43):
Just to be clear about that, like the ideas, California,
like much of the world, all of the American West,
evolved to burn, to happen right, fires periodically, and we
had this century essentially of not letting anything burn that's right,
and you get too much fuel, too much underbrush, and
then when there is a fire, it's too hot and
it burns too big in a way that it would
not have if we had let fires burn in the
(12:05):
first place.
Speaker 3 (12:05):
Perfect, man, you've clearly studied this topic.
Speaker 1 (12:08):
Yeah, it's really interesting, right, it's not entirely intuitive, but
not into about it and it's like, oh interesting, Yeah.
I spend a lot of time in the Sierras where
It's like you see like Manzanita are like pumping out
oil because they want there to be a fire.
Speaker 2 (12:21):
Yeah, and Manzanita is like a blow torch when it
starts on fire. So yeah, so yeah, I mean that's
a huge problem. And I think the point is that
it's not simple enough to say when a fire starts,
put it out. There has to be a decision made
about there's you know, the way that this has best
been described to me. We work with an advisor named
Kate Dargan. He used to be the fire chief of California,
(12:42):
and she talks about good fire and bad fire. Right
when you get to what you were describing, where you
have these over fuel loated forests that burst into flame
and are burning basically so hot that they're killing everything
in a very unhealthy way.
Speaker 3 (12:54):
That's bad fire, and like you want it.
Speaker 2 (12:57):
You don't want that to burn, But we need a
lot more low intensity fire through the forest to keep
the forest healthy. And so right now, there's basically no
way for the fire agencies to distinguish that for the
most part. I mean, they're typically because of the risks involved,
they're just going to try and put the fires out
because there's so much at risk.
Speaker 1 (13:16):
Well, like politically, if you think about it, politically, you're
never going to get in trouble for putting the fires, right,
But if you're like, well, you know we made the call,
and you know our model said it should burn, like,
that's just not going to play if anything bad happens.
Speaker 3 (13:31):
That's right exactly. Yeah.
Speaker 2 (13:33):
Kate's big thing is if we can measure the intensity
on a near real time basis, we can make way
better decisions about how we deploy resources to combat the
bad fire while allowing the good fire to burn. And
that's that's a very heavily kind of risk based decision
process that needs to be much better informed by data.
And I think that's in many ways we love fire
(13:56):
Set because I think it is such a tangible example
of the power of these space based data collection systems
to not only tell us something about the world or
you know, oh that's interesting, but actually drive decision making
on the ground. We now have actionable data that can
be used to say we're going to position firefighters here
(14:16):
instead of there, and that's not only going to protect this,
you know, this town, but it's also going to allow
this other part of the forest to burn.
Speaker 3 (14:22):
So there's less risk next year that it's going to burn.
Speaker 1 (14:25):
Right. So you say we now have actionable data, but like,
we don't literally have it now, right, because the project
is still ongoing. And so just to be clear, the
goal of the project is to have So it's an
infrared image, right, which makes sense of every piece of
the earth out of resolution. You said about the size
of a swimming pool, right, which is a nice image.
Speaker 2 (14:47):
Yeah, the actual resolution is a lot bigger than that.
It's about fifty meters, so think about half of a
football field. But we can see we can detect a
fire within that fifty meters down to about the size
of a swimming pool. So if there was a swimming
pool size fire on half a football field, we would
detect that fire.
Speaker 1 (15:05):
Okay, and to do that of the whole earth every.
Speaker 3 (15:08):
Twent the minutes, that's right.
Speaker 1 (15:10):
It sounds like a lot.
Speaker 3 (15:12):
It is a lot of year.
Speaker 1 (15:13):
It is a lot, right. And so so you start
this company five years ago around the time this project
is kind of coming into view, right, and you the
company and the project sort of get going together.
Speaker 3 (15:25):
Yes, that's right, yep.
Speaker 1 (15:28):
And so I mean you talked about how you had
solved a lot of these things at Skybox. But plainly
there's a lot to figure out here. So when you
launch the company, what do you have to figure out
to do this?
Speaker 2 (15:38):
Yeah, so there's a couple of things. I think one
thing is be able to build a small and low
cost infrared camera, which is a critical enabler for this
to work. And that's something that we bid off very
early in the process, very much using again the same
kind of approach that we use at Skybox.
Speaker 3 (15:55):
At Skybox, we were.
Speaker 2 (15:56):
Much we were focused on visible imagery, which is a
different thing that's very much a solved problem now for space.
But if you ask anybody when we started the company,
what does it take to build like a very high
sense it'sivity infrared imager from space, they'd probably tell you
it's like a ten to thirty million dollar problem per unit.
Speaker 1 (16:15):
Wait, you mean it's going to cost you ten million
dollars for each camera?
Speaker 2 (16:19):
Yeah, I mean if you went out to Lockheed or
Northrope or whoever builds these things, because you know commercially
they're not built.
Speaker 3 (16:25):
Yeah, that's what they would have told you. That's what
you know.
Speaker 2 (16:27):
In many ways, like the government is paying today for
that type of payload.
Speaker 3 (16:32):
And we couldn't do.
Speaker 2 (16:33):
That because, you know, ultimately, to get to that twenty minutes,
which everybody agrees was a critical goal, you need like
fifty satellites and so if you're if just the camera
on each satellite costs twenty million dollars, it's pretty hard
to imagine that making sense.
Speaker 3 (16:46):
But we needed like a ten x.
Speaker 2 (16:48):
I mean, we needed to get this thing down to
like a million dollars or you know, no more than
two million dollars. And so again we had to go
to the out and say, like, what are technologies that
have been developed for other industries that we can apply
to this problem that haven't been used in space before,
because typically there's a lot of industries that have spent
billions of dollars on technology and getting the cost down
(17:10):
and the performance up. And so what we found was
there was there had been work on It was largely
for tactical military applications, so not things that go in space,
but things like the night vision that a soldier wears
in the field, where they're building hundreds of thousands of
these things at fairly low cost, and the underlying camera
technology is applicable to it's what we needed for this
(17:33):
it's a very different way to do it. Again, it
was another case of like we literally had people saying
this will never work, you can't do it that way.
Speaker 1 (17:40):
While as you say that, well, I mean when people
say that too.
Speaker 2 (17:43):
I love it when people tell me something's not possible,
because that's just like that's a great challenge, like Okay,
I'm gonna prove you wrong. So yeah, So that was
a big challenge that we had to solve and we did.
I think another challenge that we really had to solve,
which is related to fire set, but it's also related
to the bigger picture for the company is you know,
(18:05):
these things produce enormous amounts of data, and they often
are constrained in various ways in terms of the ability
to get that data back to Earth and then into
the hands of people that need it, and certainly on
reasonable time scales, so like typical satellite systems maybe only
might take an hour and a half or two hours
from the time a picture is taken to when it's
(18:27):
actually usable in somebody's hand.
Speaker 3 (18:30):
And again for the fire.
Speaker 2 (18:30):
Problem, obviously that's not going to work if we're going
to put all these satellites up to fly over every
fifteen minutes, Like, waiting an hour and a half for
somebody to get that data is not helpful. So we
had to put a lot of time and effort into
thinking about how you can build extremely high bandwidth, extremely
low latency pipelines basically to get data from satellites and
out to users. And the reason I say it's more
than just fire sat is because I do think that's
(18:52):
something that is an unlock for a whole bunch of
other things too.
Speaker 1 (18:56):
So how'd you figure that one out?
Speaker 2 (18:59):
So there's kind of two different pieces of it. There's
kind of a phase one and a phase two. The
phase one is clever, you know, basically clever archets picture
of where we put antennas around the world, and how
we design our radios to be able to continuously downlink
while the cameras are collecting, which most satellites don't do.
Most satellites collect some data, store it, down link it.
(19:22):
Our system is designed to be able to continuously do both,
and to do it to antennas that are basically placed
in geographies where wildfire really matters, so it's a near
real time connection all the time.
Speaker 3 (19:33):
That's kind of phase one.
Speaker 2 (19:35):
So the other thing kind of the phase two of
this that we're now bringing on board is we have
this partnership with SpaceX where we'll basically be connecting these
satellites into the Starlink network in space.
Speaker 1 (19:44):
Is that the laser part? Are you not saying lasers?
Speaker 3 (19:47):
It is lasers and space. Lasers and space. Yeah, well,
lasers in space.
Speaker 2 (19:52):
So now rather than having to wait to come over
a ground station or try and place one in exactly
the right spot to keep that latency down as we're
collecting data and space, we will have essentially what's like
an always on broadband connection to the Internet and we
can just continuously stream that data back for users. That's
completely transformative, right in terms of how you think about some.
Speaker 3 (20:12):
Of these problems.
Speaker 1 (20:13):
So where is fire SAT now?
Speaker 2 (20:16):
So we launched what we called the protoflight satellite, so
a first instance last year and it's up collecting data.
We're launching the first three of what we call the
operational constellation in June and then basically we have a
you know, a roadmap of launches that gets us out
to the first kind of big like milestone we're shooting
(20:37):
for is hourly So having a picture every hour of
the entire planet, which we should get to in sort
of late twenty seven early twenty twenty eight, and there
will ultimately be about twenty in that hour hourly constellation.
And then after that the goal is to expand that
ultimately to fifty two.
Speaker 3 (20:55):
That'll get us to fifteen minutes.
Speaker 1 (20:57):
Why might it not work?
Speaker 2 (21:00):
Yeah, So one of our biggest worries and concerns, because
there's a lot of history of this is up to
by the operational community.
Speaker 1 (21:12):
And I mean if there's no customers essentially.
Speaker 2 (21:14):
Yeah, our users, or if it's difficult for them to
incorporate into their operational flows, and.
Speaker 1 (21:20):
So somehow it's not actually helpful in fighting fires or
understanding fires.
Speaker 2 (21:24):
Right, And I think we're very confident that the data
itself is helpful, but it's only helpful, you know, if
you can get you know the term the water to
the end of the road, right, So it's not helpful
if all this data ends in some cloud bucket somewhere
that nobody can get to. It's only helpful if it
actually ends up out in these operational command centers, in
(21:45):
the tools that the fire operators are using on a
daily basis. And so that's actually not a trivial problem
because it's a huge patchwork of different systems and tools
even in the US, like there is no unified system.
So like our priority earliest integration partner is CalFire For
(22:06):
that reason, CalFire is the most sophisticated fire agency.
Speaker 3 (22:09):
They have the best tools to do this.
Speaker 2 (22:11):
But you know, ultimately we want to do a lot
more than that, Like, yeah, this is not just for California,
it's not just for the US, it's for the world.
And so that long tail of how do you actually
get the water to the end of the row with
the users that need it, I think is actually the
heart and not the hardest problem, but it's one of
the harder problems, and I think it's the biggest risk to.
Speaker 3 (22:30):
Ultimately having impact.
Speaker 2 (22:32):
Like, you know, if we get all these satellites in
space and they're wonderful and they're producing data, but we
don't solve that problem, it's all for nothing.
Speaker 1 (22:39):
Yeah, that could totally happen.
Speaker 2 (22:42):
Well, we're going to try and make it not not
a question.
Speaker 1 (22:45):
Yeah, not technically a question to say that could totally happen.
We'll be back in just a minute. Hey, it's Jacob,
and I want to tell you that I am hosting
a new show called Business History. It's about the incredible
(23:05):
innovations and massive failures and unbelievable characters in the history
of business, and I hope I think the show provides
insights about how business works today. At the end of
today's episode of What's Your Problem, We're going to play
you a clip from Business History. It's the story behind
the video game company Atari, and they're surprising early hire
(23:26):
of a young hippie named Steve Jobs. The show's called
Business History. You can listen to it wherever you're listening
right now, and we'll play that clip at the end
of today's episode. Tell me about your work with the
Department of Defense.
Speaker 3 (23:42):
Yeah, So there's a couple of worlds out there right now.
Speaker 2 (23:48):
There's the commercial world, and I would put fire sat
in that.
Speaker 3 (23:51):
World, and then there's sort of the national security world,
and they're related in a lot of ways.
Speaker 2 (23:57):
It's kind of hard to separate them completely, but there
is right now a I would say existential need, especially
in the US, for new technology and space. And the
way I would think about this is that we've spent
sixty or seventy years building very expensive, high performance, exquisite
(24:18):
systems for space. On the national security side, most people
don't even realize exist, and they're absolutely absolutely critical to
our country's ability to operate in the world, like in
a way that you know, we're very very sensitive too.
So historically we have dominated space as a country that's
(24:40):
not really been contested, and so this has been both
a superpower but also sort of a capability. It's kind
of taken for granted. The armed services don't even think
about the idea that some of these things, these tools
that they have that they rely upon, would not be
available or useful.
Speaker 1 (24:56):
Like that some enemy might shoot them down exactly what
you're alluding to, Yeah, or jam them or jam them, right,
I guess jam them is actually what happened.
Speaker 2 (25:05):
Well, I mean the Chinese shot a satellite down a
couple of years ago, right to demonstrate they can. So
like there's a lot of concern about that. So right
now we're in the midst of this period where those
old systems have were designed in an era where there
weren't threats. There's now threats. There's also a lot of
like China has become much more capable in space and
(25:27):
it's now more like a peer to us. Versus like
a strong second.
Speaker 3 (25:32):
And so there's this.
Speaker 2 (25:33):
Huge push to basically modernize and refresh all of the
kind of space infrastructure that the US uses on the
national security side, but very much leveraging the commercial approach
that we've been talking about that I think we kind
of pioneered at Skybox a decade ago, and at the
time it was perceived as too risky and too early,
(25:54):
but there's now a recognition that that is the future,
this is how these systems need to be built, and
that there is very quickly becoming this vibrant commercial industry
that can can build a lot of these systems.
Speaker 1 (26:06):
Analogous to what SpaceX did for rocket rockets used to
be whatever, three two whatever. It was one military contractor
and then a private company game. It was like, look,
we can do this cheaper, right, That's happening for satellites
now for the government, for the military.
Speaker 2 (26:20):
Yeah, And very similarly, SpaceX, you know, did it in
a way that was commercially oriented. They're launching commercial communication
satellites but also supporting national security missions. So I think
that's space inherently is very difficult to separate from that perspective.
Speaker 1 (26:38):
Separate private from the military essentially, Yeah, but just because
that's where the money is.
Speaker 2 (26:45):
That's where the money is. But also because anything you
put in space inherently has the ability to do both, right,
I mean in the limit, if you really want to
think about the limit. And it's interesting just to make
this make this point pretty clear, China does not really
distinguish between commercial satellites and military satellites. We historically have
(27:05):
distinguished that. But one of the big fears more and
more is that that you know, our government should be
depending on commercial satellites. Do those then become targets in
a conflict?
Speaker 4 (27:16):
Right?
Speaker 3 (27:16):
And so there's it's a little another analogy.
Speaker 1 (27:19):
I think that's a good thinking of the Lusitania.
Speaker 2 (27:22):
I I was gonna say naval like kind of pre
World War two naval conflict, right, birch and ship can
be used for goods, but it can also be used
as smuggle weapons and and you know, in a lot
of ways, our navy after the Second World War sort
of police the world's oceans to kind of guarantee a
lot of the kind of separation between commercial and military
(27:45):
affairs in the ocean. There's no equivalents of that in space.
And so that's I think that's a big concern people
have right now, is that even if you want to
say this is a commercial system, this is a military system,
it's very hard to draw that line in a way
that China will respect it or some other adversary will
respect it. And so all to be said, the technologies
(28:06):
are very similar. The technology are the same, right you know,
if you put something in space, it's it's very hard
to tell what it's for for somebody else looking on
like it's very easy for.
Speaker 1 (28:17):
It's just a satellite. It's satellite country taking a picture.
Speaker 2 (28:21):
Scout radios and sensors on it. Are those radios and
sensors communicating to commercial stuff? Are they community in the military?
Are they taking pictures of you know, people's trampolines in
their backyard. Are they taking pictures of military sites? It's
very hard to separate. And I think we're seeing a
big convergence right now in and a recognized convergence of
those those spheres. And so there's you know, there's both
huge opportunities and risks with that. You know, the opportunity
(28:43):
for us is that the things we're building are applicable
to both There's a huge need on the national security side.
And to give a very specific example of this, you know,
we want a contract with the Space Force last year
to basically take what we developed for fire SAT, this
infrared camera, this low cost infrared camera, and adapt to
a mission set that they've had for forty years that
(29:06):
has been very challenging for them to upgrade, which is
what they call They call it theater imagery and cloud characterization,
but it's basically weather forecasting. So the DoD has to
forecast weather. If they want to fly planes, they need
to know what the weather is, so they have their
own weather forecasting capability, and they have satellites to feed that.
And historically these have been these billion dollar, twenty year
(29:27):
bespoke things. And what we've been able to identify as
the same technology we've built for fire Sat is very
applicable to deploying satellites to solve that mission for them.
So I think that's a really good example of kind
of this crossover and how and how inherently dual use
a lot of these things.
Speaker 1 (29:45):
Are you were saying there's opportunities and risk. That's the opportunity.
What's the risk for you?
Speaker 2 (29:50):
I think there's like a perceptive risk, and I think
there's a physical risk. The perceptive risk is. You know,
we started the company very much to do positive in
the world, and I firmly believe that's what we're doing.
I think Fire SAD is a canonical example of how
we're trying to do this. I believe there's a lot
of things in national security sphere that we can do
(30:12):
that also have that characteristic that are positive things for
the world.
Speaker 3 (30:17):
Not everybody agrees with that.
Speaker 1 (30:19):
Yeah, well, and I mean, you know, somewhat famously, at
other big tech companies, right, people who worked at the
companies didn't like it when Google and Microsoft started doing
work for the Department of Defense.
Speaker 5 (30:31):
Right.
Speaker 2 (30:31):
I was at Google when the Project Maven thing happened,
like when there was the big backlash against their work
on that. So, like, have you encountered any of that
at MWON. Yeah, No, We've definitely had people with concerns.
I mean, I think that what we've tried to do
is just be very open about why we're doing it
and why we see it as a positive addition to
our mission set. And for the most part, that's been
(30:52):
very successful. I mean, people typically are engineers, are very
rational people usually, and so I think if we can transparently.
Speaker 1 (31:01):
Present rational I mean you rational people can disagree about
the most prevailance of the American Department of Defense.
Speaker 2 (31:08):
Oh, you're right, And so I mean again, like my
perspective on this is like I never want to jam
this down anybody's throat, Like I think people need to
make their own decisions and be very comfortable with what
we're doing, and we're just trying to be as transparent
about that as we can.
Speaker 1 (31:22):
Let's talk now about the broader state of the industry,
of the space industry, such as it is. I mean,
it's particularly interesting given that you have seen now this
rather long arc, right when there has been this big
transformation in the industry over what fifteen years or twenty years? Yeah,
twenty years. So you know, how has it changed? That's
(31:45):
the big dumb question. Like, yeah, I mean there's the
obvious stuff that you've done, but like, what can you
do now that you couldn't have done?
Speaker 4 (31:50):
Then?
Speaker 1 (31:51):
What's you know, what's interesting, what's surprising.
Speaker 2 (31:53):
So I'll give some anecdotes. So I interned at SpaceX
in like two thousand and four. I was there's like
forty people there, and you know, my first job was
helping build the first Merlin, which is like the first
stage engine for Falcon one, and testing it. And you know,
this was kind of when I was there, there was
(32:14):
this excitement about how space This was very early, right,
there's forty people, and there was this belief that they
were going to totally transform the industry that early. You know,
I learned very quickly that we were a long way
away from that. The first time we put this engine together,
probably tore itself apart, burned the test stand down, and
you know, I spent the rest of the summer like
(32:35):
rewiring this test dan in the summer heat in the
Texas eat and it took him another like five years
before they even got a first rocket into space after that,
And if I fast forward from there and then think
about starting Skybox in two thousand and nine, which is
roughly at the same time, again we had this perspective
that it's like, okay, well it took a while, but
(32:56):
SpaceX is solving this problem. They solve this launch problem
where access to space is going to be easier. They
now have Falcon one. We actually were going to buy
two Falcon ones. It's Skybox. Early on and right at
the last minute, Elon said, there's not a Falcon one
is not business competitive.
Speaker 3 (33:12):
We're gonna build a bigger rocket.
Speaker 2 (33:13):
We're gonna build Falcon nine, and all the contracts that
we're gonna do for Falcon one are now Nolan bought
Void and so this was existential for us at Skybox.
Speaker 3 (33:20):
We ended up having.
Speaker 2 (33:21):
To go to Russia, of all places, and buy these
converted Cold War era ICBMs to launch our satellites with.
That's we were literally launching our satellites out of a
missile base in southern Russia off ICBMs.
Speaker 3 (33:35):
And so again like.
Speaker 2 (33:36):
We were like, okay, well, this problem we thought was solved,
it was going to blow the door open, is still
not solved. It is now solved fundamentally. And I mean
there's all kinds of stats that show this from the
cost of launch that has come down, and launch launch
has been the big choke point for most of the
history of space. So that's a huge unlocked.
Speaker 1 (33:57):
Hard to get there space hard to get there now,
not as hard, not as hard.
Speaker 2 (34:01):
I don't think you can underestimate the impact of solving
the launch problem. I'll call it in terms of how
the world opens up, in terms of what you can.
Speaker 3 (34:11):
Do with space.
Speaker 2 (34:12):
The other thing I would say is I think we're
even when we were doing Skybox and thinking very differently
about how to build satellites, we were still building in
a silo. We were building this thing that the analogy
I like to use most of stuff in space is
kind of like species on the Galapagos Island. They're like
these pieces of technology that have been like separated from
(34:35):
the mainland for so long. They've kind of evolved in
these like perverse ways.
Speaker 1 (34:39):
And if you actually like this's whatever, like the motors
and the cameras and whatever, they're like, oh, I understand
how this used to be like a motor or camera
on Earth, But now it's totally different, totally.
Speaker 2 (34:48):
Like literally to the point where if you looked at
like the computer on even a satellite that got launched
ten years ago, it would look like something that TI
put in a calculator in the nineteen seventies, right, I mean,
it's crazy. So there's this kind of like frozen in
time nature to a lot of the technology in space,
and because of the launch, but also because of these
things that we've been talking about, there's a big turnover
(35:09):
where now the technology in space is looking a lot
more like the stuff that goes into data center, where
that goes.
Speaker 3 (35:14):
In your smartphone.
Speaker 2 (35:16):
And because of that, there's also a lot more interconnectedness
between those technologies and the technologies on Earth. And I
think we're just in the early stages of this and
it's going to cause it's already causing, but it's going
to cause this huge convergence in what we think about
as space technology and every other bit of technology in
our life, to the point where we're going to realize that.
(35:37):
You know, right now, I do I use space technology
on here GPS, right, satellites are in space and when
I look at my map, it's telling me where I
am based on satellites and space. But nobody thinks about
those satellites, you know, when they look at their map
and ask where they're going. That's going to be true
of like a whole bunch of different things, Okay, what
things and when one of the most exciting things for me,
(36:00):
there's been a lot of discussion about like orbital data
centers right people are talking about orbital data centers right now.
Speaker 1 (36:04):
Yeah, And just to be clear, like an actual like
AI data center that is some giant.
Speaker 2 (36:09):
Gigawatts of solar power and space. Yeah yeah, yeah, and
it sounds crazy.
Speaker 3 (36:14):
I think it's going to happen.
Speaker 1 (36:16):
Okay, when and why?
Speaker 3 (36:19):
So the why there's there's a bunch of reasons right
now to build a data center takes between five and
seven years okay, between permitting, construction, power, interconnect, all of
these other things.
Speaker 1 (36:31):
I love that it's permitting, permitting, permit will drive data
centers to space.
Speaker 2 (36:36):
It seems crazy, but I mean it's true, right and
so like again, and I'm not saying this is easy,
but if you could get a factory set up and
rockets launching frequently enough that you can go deploy that
one hundred megawatts of data center to space in two
years instead of five years, there's a huge time value
money component to that.
Speaker 1 (36:53):
That's it's so wild. What give me the over under?
What year is that going to happen?
Speaker 3 (36:57):
Twenty thirty five?
Speaker 1 (36:59):
That soon? That soon?
Speaker 3 (37:01):
I think that scale could happen that soon and.
Speaker 1 (37:03):
Is part of what has to happen there, Like it
keeps getting cheaper to go to space absolute kilogram.
Speaker 3 (37:09):
Yeah.
Speaker 2 (37:09):
So so like we've done a bit of work on
this with some hyperscalers.
Speaker 1 (37:13):
So they've come to you and asked hyperscalts is whatever.
This is four giant tech companies. They're all actually calling
you and saying, yeah, I mean it take to put
a data center in space.
Speaker 3 (37:23):
Google has announced it publicly.
Speaker 2 (37:25):
Okay, so you know the cost per kilogram to put
something in space is the key metric. And I think
for economic parody, if you think about, like how what
it would take to get to a similar cost per
unit capability of compute and a data center on Earth.
Right now, we're still off by about a factor of six.
(37:46):
So like it's six, but it's not one hundred, it's
one hundred. It's not even an order of magnetus.
Speaker 3 (37:51):
It's not even an order of magnitudes.
Speaker 5 (37:53):
And so is the.
Speaker 1 (37:54):
Reason it's cheaper once you're up there? What the energy permitting?
Like why is it? Why is it only six? So
it only effect us.
Speaker 2 (38:03):
So here's here's a really important factor. If you put
a satellite into space in the right orbit, you can
have solar rays that are in this sign one hundred
percent of the time. So it's baseload power. Now it's
solar is no baseload power. I mean, solar is extremely
cheap right now, but you can't use just solar to
power data center because it's not consistent, so you end
(38:23):
up needing massive some other form of massive energy storage,
and it.
Speaker 3 (38:27):
Has become very expensive. So I think that's a big deal.
Speaker 2 (38:30):
I think that's one of the reasons in my opinion,
that it at six x you know, it's cost competitive
because it's when you actually factor in all those other
costs to terrestrial data centers for things like power, it's
actually higher than you would think. Okay, So again, the
other thing I just think is an important point is
and I think this is being missed a little bit, like,
(38:52):
we don't go from here to gigawat data centers in space. Right,
there's this huge spectrum between where we are now and that,
and I have no doubt we're going to be going
way down that spectrum over the next ten years. Right,
so where it actually lands, Are we actually going to
have gigawat data centers in space?
Speaker 4 (39:09):
Like?
Speaker 3 (39:09):
Who knows? Right, who knows if the models will continue scaling.
Speaker 1 (39:13):
The scaling laws will hold.
Speaker 3 (39:14):
Yeah, there's a stuff. What's the midpoint?
Speaker 1 (39:17):
What's the like easier to imagine launches don't have to
get that much cheaper. We're almost there now, but still
cool thing yeah.
Speaker 2 (39:25):
I mean I think it's that there are these kind
of multi tiered layers of what we would think about
as like cloud infrastructure in space, right, so very high
speed network. So now you have this high speed network
that everything is connected to all the time in space,
which is not the case.
Speaker 1 (39:38):
Now what we thought of a data center ten years ago,
like where they're like whatever.
Speaker 2 (39:45):
You know what you're getting Netflix served from? And I
mean Starlink's already doing this a little bit. Starlink cash's
Netflix on the satellites, right. And what will happen is
as that builds out, there will be reasons to move
certain things that are in the cloud on data centers
and Earth now up to space to make the latency
lower or the processing more efficient or whatever. But we
won't even think about it. That'll just be in the background.
(40:08):
We'll use the thing and nobody will know whether it's
happening in space or happening in the ground.
Speaker 1 (40:15):
We'll be back in a minute with the lightning round.
Let's finish with the lightning round. What's the secret to
having one of the best breaking balls in the Stanford bullpen?
Speaker 3 (40:39):
Oh man, it's top spin.
Speaker 1 (40:46):
What's your favorite satellite that you didn't build.
Speaker 3 (40:51):
The Corona satellites.
Speaker 1 (40:53):
Tell me about the Corona satellites.
Speaker 2 (40:55):
So the Corona satellites were the first imaging satellites we
built in the early sixties. And they were film based systems,
so they had film in them that ran through them
that you know, this incredible film technology that ran through
them that you know, one hundreds defeat per second to
keep up with the speed that they were flying. And
that would get loaded into these canisters and dropped off
(41:17):
in reentry vehicles parachute down to Earth and then we
would go pick them up by scooping them out of
the air behind a plane.
Speaker 3 (41:23):
To go develop the film and distribute it.
Speaker 1 (41:26):
Why was I just reading about that? I was reading
about the special planes, right, they made special planes to
catch the canisters as they're.
Speaker 3 (41:33):
Coming down, these like tohoks.
Speaker 1 (41:35):
It's so wild and like physical, right. You're a mechanical engineer, right,
that's like very yeah, so mechanical, yeah, yeah.
Speaker 2 (41:41):
I mean, and you look inside these things and there's
just like all these rollers and motors and stuff to
like move the film around and develop it and you know,
put in the.
Speaker 3 (41:48):
Canister and drop the canister off.
Speaker 1 (41:50):
So analog, right, it's just all physical stuff. What's something
you learned working on autonomous vehicles that lift that's been
useful to you at new.
Speaker 3 (42:00):
Yeah, I mean I actually thought about this a lot.
Speaker 2 (42:03):
The way that robotics and av used data is totally
revolutionary from somebody that's come from aerospace because of some
of the things we talked about about how primitive a
lot of historically, like compute on aerospace systems is, you
end up having to be very conservative about data and
(42:26):
processing and all of these things. And avs are one
hundred percent the opposite. They're just these like data vacuums
and so, but what it does is it means that
you think about these problems totally differently, like solving problems
totally differently.
Speaker 1 (42:39):
So what does it mean to you in your work?
How do you use that?
Speaker 2 (42:42):
I mean by architecting our systems to look more like
avs than satellites.
Speaker 1 (42:48):
Just to be able to handle more data.
Speaker 2 (42:50):
Yeah, I mean we can use the same technology. So
we are putting massive amounts of compute and networking and
you know, bandwidth on our satellites, and so we can
do that.
Speaker 1 (43:02):
Will evidence of extraterrestrial life be discovered in your lifetime?
Speaker 3 (43:06):
Yes?
Speaker 1 (43:08):
Okay, Uh, there is there anything you know that I
don't know that makes you.
Speaker 3 (43:12):
Say that, No, I just think that.
Speaker 2 (43:16):
I think I think we'll discover very simple extraterrestrial life.
I think there's life arose so early in the history
of the planet, simple life that it just seems like
it's very likely to have risen in other places or
have been transplanted here from other places.
Speaker 3 (43:34):
I think complex life is a lot harder.
Speaker 5 (43:36):
Like.
Speaker 3 (43:36):
I do not think we will discover complex life in
my life.
Speaker 1 (43:39):
We'll find like bacteria on a moon of Jupiter or something.
Speaker 3 (43:43):
Yeah, I think like single cell procaryot like things.
Speaker 1 (43:46):
Yeah. Will you go to space before you die?
Speaker 3 (43:55):
I would love to? I think probably not.
Speaker 1 (43:59):
Why not?
Speaker 2 (44:02):
I'm not convinced there's great reason to send humans to
space A lot I think. I think there's a lot
of things we can do without sending humans to space,
all right.
Speaker 3 (44:11):
I think where we do, it'll be limited.
Speaker 1 (44:14):
Why would you love to?
Speaker 2 (44:16):
I'd love to see Earth from space? I think the
overview effects you I've heard that astronauts talk about the
total transformation and perspective that comes from seeing the Earth
from space.
Speaker 3 (44:28):
So I'd love to experience that.
Speaker 1 (44:36):
Johnny Dyer is the co founder and CEO of muon Space.
Today's show was produced by Gabriel Hunter Chang. It was
engineered by Sarah Bruguire and edited by Lydia Jane Kott.
I'm Jacob Goldstein. You can find me on x at
Jacob Goldstein. You can find me on LinkedIn. You can
also email us at problem at pushkin dot FM. We'll
(44:58):
be back next week with another episode of the show.
I'm Jacob Goldstein and right now we're going to play
you a clip of a new show that I co host.
The show's called Business History. My co host is Robert Smith,
(45:20):
and this clip is from an episode we did about
how Nolan Bush, now a stoner turned entrepreneur, created the
video game company Atari and hired a young, inexperienced Steve Jobs.
I really hope you like the clip, and if you
want to hear more, you can find Business History wherever
you're listening to this show. Right now, this like nineteen
(45:42):
year old hippie kid walks in and he says he
won't leave until they give him a job, and the
receptionist calls the head engineer and she goes, yeah, we
got a hippiekid in the lobby says, he won't leave
until we hire him. Should we call the cops or
let him in? And the engineer says.
Speaker 4 (45:59):
Bring him on in.
Speaker 5 (46:00):
It's nineteen seventies, a Silicon valley, and this guy wanders
in and he is whenever you tell a story like this,
you know who it is.
Speaker 4 (46:07):
It's Steve Jobs.
Speaker 1 (46:08):
It's Steve Job fun. It's so delightful and perfectly. Steve
Jobs is very good at his job, yes, and very
unpleasant to wrqu with. On surprising, he tells Bushnell that
like everybody's soldering wrong, right, they're actually putting together the hardware,
soldering the hardware.
Speaker 4 (46:25):
He's probably right, and bush Nell's like, yeah, he was right.
Speaker 1 (46:27):
He keeps calling his manager a dumb shit. Probably was
not kind, but not necessary. Bushnell winds up putting Jobs
on the night shift, partly so he won't bother so
many people, and partly because he knew that Jobs like
to hang out at night with his buddy Steve Wozniak,
who was a great engineer, would be great to have
(46:48):
hanging around atari And in fact, Jobs and Wozniak helped
to make Breakout great a target. Remember Breakout, of.
Speaker 4 (46:55):
Course you're trying to knock down the bricks in a wall.
Speaker 1 (46:57):
Still games like that, when you got a little paddle
at the bottom.
Speaker 4 (46:59):
End, there's a moment when it goes through the wall
and then goes so.
Speaker 1 (47:04):
Good. So Jobs worked at Atari for a little while
and then decided he wanted to go out to India
to find his guru. Perfect asked Atari to pay for
the trip. Nobody ever said he lacked moxie and wound
up making a deal with Atari where they pay him
to go part of the way there. They had exported
some games to Germany and there was some kind of
problem with the games in Germany, and they're like, we'll
(47:26):
send you to Germany to fix the games and then
you can get the rest of the way to India.
And you know, the Germans said Jobs was terrible to
work with, but he fixed the games. I was thinking
about like the link between Atarian Jobs and what did
he learn there, and it felt like maybe a little overdetermined,
but I do think, you know, clearly he had this
profound sense of aesthetics and of delight, right, like think
(47:49):
of the Macintosh, right, this breakthrough Apple machine in the eighties.
It was round and instead of squares, it was rounded,
and it cost them more money, but it was beautiful
and it was fun and you were engaged with it
like a game.
Speaker 5 (48:01):
It almost looked a little bit like an arcade game
with the curves. Yeah, but more than that, I think
that especially at Silicon Valley at the time when they
were dealing in action Silicon. Right, if you're making chips
for somebody, you're thinking about the future and the computers.
You're not thinking about the psychology of the customer, because
the customer was another electronics company, right, And so Atari
and the strength of it was really the first time
(48:22):
where they're just like, how will a regular human being
who has no training whatsoever interact with technology?
Speaker 3 (48:29):
Yeah?
Speaker 1 (48:30):
So Atari now mid seventies, they're selling all the machines
they can make. They need more space, and so they
rent an abandoned roller rink and they turn it into
this office slash video game factory.
Speaker 4 (48:45):
I've been ten years older, I would have loved to
have worked in a roller rink video game factory.
Speaker 1 (48:49):
Yes, everybody smoked weed. Yeah, there was a hot tub.
There was a pool party where everybody ended up naked
in the pool at Bushnell himself looked back on it
later and said if that isn't a horror show for
any HR person today, I don't know what is
Popular Podcasts
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
Betrayal Season 5
Saskia Inwood woke up one morning, knowing her life would never be the same. The night before, she learned the unimaginable – that the husband she knew in the light of day was a different person after dark. This season unpacks Saskia’s discovery of her husband’s secret life and her fight to bring him to justice. Along the way, we expose a crime that is just coming to light. This is also a story about the myth of the “perfect victim:” who gets believed, who gets doubted, and why. We follow Saskia as she works to reclaim her body, her voice, and her life. If you would like to reach out to the Betrayal Team, email us at betrayalpod@gmail.com. Follow us on Instagram @betrayalpod and @glasspodcasts. Please join our Substack for additional exclusive content, curated book recommendations, and community discussions. Sign up FREE by clicking this link Beyond Betrayal Substack. Join our community dedicated to truth, resilience, and healing. Your voice matters! Be a part of our Betrayal journey on Substack.
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com