August 18, 2024 • 32 mins
Magnus Müller, Co-founder and CEO of Greenw.ai, joins us to discuss a critical yet often overlooked aspect of our transportation infrastructure: traffic lights.
Currently, most traffic lights rely on outdated timer systems, basic IoT sensors, or even manual operation. This leads to inefficient traffic flow, excessive idling at intersections, and a cascade of negative effects: frustrated drivers, accidents, CO2 emissions, and chronic traffic congestion. Greenw.ai is on a mission to revolutionize traffic management.
By leveraging anonymized mobile phone data and cutting-edge artificial intelligence (AI), Greenw.ai aims to create a more streamlined traffic flow. Imagine fewer red lights, reduced emissions, calmer drivers, and safer roads – all achieved at a significantly lower cost.
Currently, most traffic lights rely on outdated timer systems, basic IoT sensors, or even manual operation. This leads to inefficient traffic flow, excessive idling at intersections, and a cascade of negative effects: frustrated drivers, accidents, CO2 emissions, and chronic traffic congestion. Greenw.ai is on a mission to revolutionize traffic management.
By leveraging anonymized mobile phone data and cutting-edge artificial intelligence (AI), Greenw.ai aims to create a more streamlined traffic flow. Imagine fewer red lights, reduced emissions, calmer drivers, and safer roads – all achieved at a significantly lower cost.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:16):
All right, Hi, welcome everyone, this is another episode. Let's started,
they'll say, and we're talking today with Madnus Mueller from Greenway.
Speaker 2 (00:26):
All right, Mad is welcome.
Speaker 1 (00:28):
Why don't we kick it off?
Speaker 2 (00:29):
Tell us a little bit about yourself.
Speaker 3 (00:35):
I thank you for having me here. Yeah, so two
years ago I was I was hitchhiking around the world
through Central America, through Middle East from Germany, tom lest
through India, Southeast Asia, and half of my time was
wasted at stupid traffic lights. So I thought back then,
(00:56):
like which century we're living in, Like this cannot be true.
So doing a hacker Ton in Germany with some colleagues
from my undergrad and cognitive science so neuroscientists, we tackled
this problem and we built the first PROTECT pilot project
for that we could actually manage in a simulation to
(01:17):
like release skyrocket, like all the metrics like Seal two emissions,
lowered downfield usage, waiting times, and we thought, wow, this
this is incredible. And yeah, since then we are on
this journey for almost two years now to tackle this
problem with a vision to both every single traffic light
(01:38):
on this planet. And while doing this, I moved from
my undergrad to my to my master program in cyric
at Etah, where I'm surrounded with many motivated people, especially
around the student project house. It's like an accelerator where
you're forbidden to study on your project on your university courses,
(02:01):
so you have to work on your court, on your project,
and I just love to be surrounded by people push.
That's that's how I came here today.
Speaker 4 (02:14):
So, Magnus, I mean, why did we double click on
the problem itself for a little bit, right, for people
to understand better? So you said you were hitchhiking. Obviously
that is a challenge by itself, right In addition to that,
you are stopping at two way too many traffic lights
and spending a lot of time. So uh, for people
(02:36):
who probably don't know too much about traffic lights and
how it works today, right, I mean, is what goes
on today in most of the world. You think, right,
why why people stop at these lights? And maybe it's
a primitive way of doing it, maybe it's not the
best optimized way to do it, But I mean, what
(02:57):
is the real problem itself? Can you maybe two minutes
on that?
Speaker 3 (03:01):
Yes? So I guess everyone who lives in a city
with traffic lights knows the situation, like fix them in
the middle of the night where you drive and you
face a red traffic light and they think, okay, house
is possible, like there's no car intersection? Why do I
have read? And of course different cultures deal different with
that situation. Some just go over some way. But if
(03:23):
you think that problem, which is now really clear to everyone,
fix some to the rush hour okay, where you have
thousand cars from every side and you have there a
stupid traffic lights which in many cities just run on
static cycles, so every side just gets the same amount
of brief time, no matter what's the current traffic, what's
(03:43):
the weather, if there was a big football game and
there are now thousand people coming from one side. So
in many cities they have like just still stupid traffic
lights and some for someone in Germany they now started
to tackle this problem by investing like millions of euros
and too expensive sensors, so they open the street, play
(04:04):
some pressure sensors or install cameras to then optimize single intersections.
But if you think about that like nowadays, especially why hitchhicking,
I realized every day around the world uses mobile phones
in their car and with that with Google Maps, tom
Tom Rixia many providers, you have globally available real time
(04:24):
data for the whole planet, and based on that data,
we build models to optimize those traffic lights, not just
for one intersections, but for whole cities. So those interactions,
they can talk to each other. You can say, hey,
I send to you now ten cars coming this way.
With this communication, they can have a much more efficient
(04:45):
optimization for the whole city.
Speaker 4 (04:48):
Got it? So any have you done any sort of
research magnets to see how big this market is? You know,
obviously you obviously seem passionate about the problem itself, right,
But I mean for you to spend time and say, oh,
I want to solve this. I want to solve this
by starting a company. I mean, what sort of numbers
(05:09):
market cab or market size rather any research numbers you have, yes.
Speaker 3 (05:15):
Yes, of course, so right now we'll do a lot
of fund raising. So I'm familiar with all this numbers,
Like the total available market is around ten billion dollars
in that space with a forty percent growth rate. Because
for example, in Germany there are now many smart city
Also in India, I saw there like one hundred new
smart cities so the government like really pushes those initiatives
(05:38):
as different reason has stealed two and sustainability issues, but
also air pollution. Also in Germany, Many, many cities are
just don't have enough people to manage their traffic lights
like manually, so they invest more and more in those
automated solutions.
Speaker 4 (05:55):
Got it. So it obviously affects us, say every day everyone,
I mean anyone who's on the road, I mean whether
you're a pedestrian or whether you're driving a car or
you're driving a commercial truck. So that's a very unique
problem that you're trying to solve. So obviously sorry, good.
Speaker 3 (06:16):
Yeah, yeah, you're completely right, Like it impacts all of us.
And if you think it impacts in so many different industries.
If you, for example, think okay, you're sitting banganor and
you're now ordering your food and your food delivery, it's
like maybe for you those couple of seconds which is late,
isn't a big difference because of stupid traffic lights. But
if you think, like for such a global business there's
(06:38):
like millions of trips every day, a couple of seconds
per trip is like immense and you can think that
from the logistic the truck through the bicycle industry like
it just impacts basically everything understood.
Speaker 4 (06:53):
So we've spoken about the problem. I think we have
a fair idea of what the problem is. So I
see greenway has an AI in it. So what is
your approach to solving this problem with greenway?
Speaker 3 (07:09):
Okay, so what was done until now was you look
at the city and you calculate, Okay, this traffic light
is five hundred meters away from this traffic light, so
we need to switch this traffic lightlight thirty seconds later green,
so that cars or bicycles, whatever you want to optimize,
have kind of a seamless flat green greenway, steamless flow.
(07:32):
But of course for us humans, that's hard to do
that with twenty different lodes, twenty different intersections and their
their complexity, the interactions between them, and that's where our
OAI comes into. Plank we have. We can create a
simulation of a city stack bangal or here, San Francisco, Berlin.
Create a simulation and then in a simulation we can
(07:55):
simulate everything. We can simulate cars, bicycle, trucks, buses, drives
around like in a real world. Now we can train
in AI, which in the beginning just completely stupid control
the traffic light. You can imagine it's like a baby
which comes to the world and there's just stupid actions
(08:17):
like putting the hand into the fire and gets feedback
and gets better and better and hair and like this.
You can also imagine our AI. We can throw thousands
of different like traffic situations on it. We can rush
our middle of the night, big football game, and all
those different data distributions can then get learned from our
(08:39):
system to get better and better. You can just choose
what it should optimize. Some cities they want to optimize
for speed, some for seal tool emissions, some they want
to improve buses or want to prioritize bicycles, and based
on what the city wants, you can then design your
optimization function to optimize statinism. Then, in the second step
(09:02):
in the real world, you don't take the data from
some Google maps or from sometimes feed it into your system,
which is learned in assimilation and then outputs the best
option to now in real time based on this data.
Control the traffic.
Speaker 4 (09:19):
Got it, So I think I get it, But at
the same time, I mean I'm a little I'm trying
to understand a little better on the real time data
of it.
Speaker 3 (09:30):
Right.
Speaker 4 (09:30):
Obviously with AI, you're training it with data that you
already have based on different permutations and combinations, like you said.
But at the same time, I mean, you know, traffic
is always unpredictable, right, I mean, you know, a group
of hundred might choose to have a party and they
might be going in you know, fifty different cars, which
(09:52):
is kind of an outlier, right, But I mean, how
do you so you have a combination of historical plus
real time data being fed and constantly kind of training
your model or updating area. Is that the right way
to think about it?
Speaker 3 (10:08):
Exactly? Yeah? So of course, like it's about making your
models better and better and better and fitting more and
more outlier cases into your model. And for example, if
you take like, yeah, traffic, like you say, is such
a huge diversity, Like you can just imagine, okay, here outside,
is now a car from Amazon Prime? Then then maybe
(10:29):
blocking the traffic light. Right, There's so many different cases.
But by creating a model which is trained on millions
of different.
Speaker 5 (10:37):
Data distributions and then you deploy them practice, and of
course you will figure out in the beginning cases which
are unexpected where which you need to handle for but
over time, like self driving cars to get better and
better and better from.
Speaker 3 (10:53):
Learning more and more and more regarded data distributions. Understand, yes,
this is interesting, the cool thing here?
Speaker 2 (11:05):
Make you have messed this up?
Speaker 3 (11:06):
I think?
Speaker 4 (11:11):
Then I'll go ahead.
Speaker 1 (11:13):
Yeah, I was gonna ask, at least the solution is
going to sit through with the state level agencies or
the traffic agencies. How has that interaction been, right? I
get like I would be curious to hear your reactions
with them.
Speaker 3 (11:34):
You mean to go to market, so how we sell
to the city.
Speaker 1 (11:36):
Yeah, like, because once you would have exposed the solution,
I'm sure they're like, oh, why our system is not
good enough? Like was that the reaction of people who
were very open to oh, look that there is a
better way of doing it.
Speaker 3 (11:52):
So for sure there are some like like you have
to understand and all our customer interviews, we realize those
cities especially they have nowadays a huge problem with skill shortage,
so they don't have enough people to handle those problems.
And for example, in normal small, medium sized cities, there
(12:13):
are hundreds of construction sites every year which changed the
road network and to handle those road blockages, those cities
they pay expensive factional engineering offices who then create new
traffic signals schedules by hand for those intersections. So by
(12:34):
doing those customer interviews, we realized, okay, with those skill
shortage where they really don't have enough people to look
for anything, we can have one of the biggest value
proposition to them. And so you really need to figure
out I mean, in the end of course you benefit
the whole society, but in a first step, like you say,
you sell the cities. So first you need to figure
(12:56):
out why would the city buy the solution, why would
they waste time on you? And one is skill shortage,
another one is still to a must reduction, and third
one is for example, direct cost savings in the the
public transport sector when they need less buses to keep
(13:16):
up the bus frequency. Imagine the bus in the city
wants that every thirty minutes is a bus in front
of your house which can pick you up. And if
now traffic is poor, they need more buses. But if
you improve your the system, then they need less buses.
And such a bus is super super expensive, like normally
(13:37):
they keep one for like ten years, costs around one million,
so to handle them for one year costs around one
hundred to more than two hundred thousand euros per year,
and that's where you can provide them like a value proposition.
And if you have like those three arguments, then you
(13:57):
can go to cities and of course, like we have
heart problems as a young startups that they to get
trust off those cities, and our.
Speaker 6 (14:05):
Strategy is to go with with pilot projects, which we
offer them almost for free to say, okay, let us
try this out so you can discover how much you
benefit from that and.
Speaker 3 (14:19):
Over time win more and more trust off those cities.
Speaker 4 (14:24):
Got it, Thanks, thanks my migness. So one of the
I don't know if it is you call it a
residual effect or a side effect. This actually can be
a big selling point I'm assuming is what goes hand
in hand is if you regulate traffic efficiently and you
regulate traffic safely, you can avoid a lot of accidents
(14:48):
as well. Right, I mean, you know why do most
of the accidents happen because you know, people are in
a rushia. People are frustrated because you know, the traffic
light is in red for so long and they take
a you know, a risk to just cut across the
street and go or you know speed or whatnot. So
I mean, is your solution with Green Bay trying to
(15:11):
pitch you know, public safety on the roads as also
one of these effects. I mean, is that something you've
thought about as well.
Speaker 3 (15:21):
So we are aware of studies which show that if
you improve traffic flow so less stopping code traffic, then
this increases road safety. And we can clearly improve traffic
flow somewhere in simulation in rush hour we can sometimes
reduce a number of waiting vehicles, but up to ninety percent.
(15:43):
And but asn't now we have never used that fact.
I think that's something which especially needs to figure it out,
like what clear percentage we can say, or like how
many accidents less are there. It's a thing which is
really hard to measure for us right now. But long term,
like this just goes hand in hand and with my vision.
(16:05):
And the cool thing here is that if you improve
one factor, then you improve so many other factors. You
improves your two waiting time, fuel usage, road safety, stability.
That's one thing which really amazed me about this problem.
Speaker 4 (16:23):
Awesome magnets. I mean, it's a it's a great idea.
Obviously you started this based off your experience traveling the world,
so you didn't see it in just one place. I
mean this is applicable kind of in every city around
the world. I mean, you know, you've you're obviously from Germany,
You've seen the traffic in in Bangalore. So do you
(16:46):
do you think, I mean greenway will work for any
permutation and combination sort of thing, or in other words,
it will it work for any any city you think,
any modern city so to speak, or you think you
have a sort of a template for an ideal city
where greenway will work now and maybe work elsewhere later.
(17:10):
What is your thought on that?
Speaker 3 (17:14):
So, of course we have our initial target customer group
where we start with this, but our plan is to
make this work everywhere, to solve of every single traffic
light on this on this planet. And of course some
cities are faster, somewhat slower, So we focus, especially in
the beginning, just on those who love us and really
want to develop it just together with us. And our
(17:36):
vision is like long term to make the whole process
around smart cities really easier for cities. Right now, imagine
you're a developer and you want to develop your app
for the App Store, so that everyone can benefit from it, right,
so you just go to Google play Store or App
Store and to upload your app. But in cities this
(17:58):
is right now, super super hard because every city uses
a different system and you have to do tenders. But
like my long term vision is that long young developers
and there'll be in India and Germany or in the
US if the ideas can come to a city and
an easy framework can bring in their ideas and easy
(18:21):
sandbox like format to run experiments on some infrastructure from
small they save parts of the city to test out
new stuff. And I think with that we can have
in so many different ways new innovation which benefits in
the end all of us will live in those cities. Great.
Speaker 1 (18:43):
I'm still really more of the fact that it's hitchhiking
that got.
Speaker 4 (18:46):
You to the solution.
Speaker 1 (18:47):
To be honest, I would be like, you weren't even
driving at that point, you still got bothered with the
traffic lights. That's just seriously amazing. But it's also it's
the scale of the problem too, right, This is you
said it yourself, right, You're starting with a small setups
that's the one city in then expand globally.
Speaker 2 (19:09):
That's a massive vision, right, how is it and how
is your day to day right when you go out
and like, Okay, I'm gonna achieve this one at a time,
how does that day look like.
Speaker 3 (19:22):
My day to day life. So it really depends on
the face of the business. Like sometimes I remember two
months ago, I was coding day and night, just thinking,
brainstorming on like a flip shot, you know, and drawing
architectures of our AI model. How's the whole pipeline going
(19:44):
right now? It's a whole day like fund raising. Right
now we have fund raising half a million euros and
so I reach out to many investors right now, I'm
here in Silicon Valley and just pitch I worked a
lot along pitch Deck, learned a lot to improve there.
About legal stuff, I learned a lot. So I think
(20:06):
that's it's one of the most amazing things to be
a founder, to just get insights into so many different domains,
learn about legal, learn about hiring, learn about tech, to
be a generalist. And I think especially for crewious people
like us, they're one of the most amazing things to do.
Speaker 4 (20:24):
So for anyone listening Magnus, you are fundraising and you
you're maybe again based on what I know I will
say it and you can correct it. You're looking for
angels for a start, right I mean you're in an
initial seed face or you can probably talk about what
sort of funding you would rather take in right now.
Speaker 3 (20:49):
Yes, So we got nominated for like a three million
grant in Germany where we want to hack at them
at the federal Ministry. And to secure this grant, we
are our saying five hundred k and we have already
the first angels committed. It's all around pre seed, like
we're especially looking for people who bring smart money on
(21:12):
the table, who wants to bring a network on a
table in the mobility space. Can be a tech space
advisor from from Google, can be also in the government
sector like everything. Just someone who is also frustrated about
this problem and wants to solve it. I'm happy to
talk with. Can be pre seed be cs or or
(21:33):
angels who really believe in that mission to solve that
problem globally. And it's crazy if you do the calculation,
like how many million tons of seal two we can
save with that solution, Like it's insane and that's really
what what motivates me.
Speaker 4 (21:49):
To somebody anyone with money in the bank and a checkbook. Right,
that's that's that's that's what you're looking for.
Speaker 3 (21:57):
Yeah, basically someone who brings a new new ideas.
Speaker 4 (22:02):
I'm always so I want to do a little bit
of a flashback maybe, right Obviously we've talked in about
enough about your hitchhiking, so to speak. But I mean,
did you always want to be an entrepreneur or it
just happened by chance?
Speaker 7 (22:24):
I think in the world where I grew up right now,
it's like a really hype it's all around ath Agent's
Rakes project house that was an esoprick.
Speaker 3 (22:35):
Like, Okay, maybe maybe I say this also because I'm
in this, baba, it could be also true because everyone
around me is like an entrepreneur. But yeah, I think
like also, when I was younger growing up and I
had problems, I always thought, okay, like what's the chance here,
Like I I remember I was sixteen, I'm selling like
(22:59):
Tisha designs online and I need to upload those designs
to like Ama Zone, and I just wasted my time
with uploading. So I just started, okay, writing my own
boughts to to upload that stuff. Just whenever if you
face a problem, I thought, okay, okay, what chance is here,
what solution could be here? Which also I don't better
(23:22):
fits from it, and I think this is what an
entrepreneur does in the beginning.
Speaker 1 (23:28):
Yeah, do you still think it's all hyped up being
an entrepreneur?
Speaker 3 (23:39):
I think right now a lot here in my age,
like many young people want to be really independent and
want to create their own company. And I think like
right now, especially online, like we have so many tools
which support us with that makes it easier and easier
to just go out and co found is went investors
(24:02):
great first MVPs, like the structure to do that is
become much.
Speaker 1 (24:08):
More Would you recommend caution or would you be more
like you know, hey, more power at this point, like
knowing what you know now?
Speaker 3 (24:18):
Yeah, if I would recommend to be an entrepreneur, yeah,
so I think, yeah, of course I wouldn't. I would
recommend it. Yeah, especially if you're like curious, like if
you feel inside of you that you want to be
it and there's just a little bit, you know, maybe
a little bit of fear or just can't be anything
which holds you back, then I would definitely recommend it
to you. But I think we often forget also in
(24:41):
big companies they are often like really entrepreneurs, you know,
like even if you're working in a big tech or
medium sized company, you can be an entrepreneur and really
innovate that company and run experiments and many examples where
they do everything with an entrepren basically does I think
(25:02):
the advantage of just being like in my situation is
just that you're kind of start from scratching. Really you
learn like everything from legal stuff.
Speaker 8 (25:11):
To product to investors through hiring, and that's just really
an amazing experiment experience which you don't get and working
in big type.
Speaker 3 (25:23):
But the mindset itself, you cannot everywhere.
Speaker 4 (25:28):
Anything you would have done differently in your journey. Magnets
looking back.
Speaker 3 (25:36):
A hundred things, one hundred things, like I think I
have every day. I have like a list with things
which I would do differently, little beaches, personal habits or
especially in the in the startup, like with some of
how we set up the legal stuff during the beginning,
we didn't have the best thing and then like no
(25:59):
way are really looking okay with what people you start
the company so you fully fully trust them, and thats
a I think a long long list with things that
would I've just learned on the on the goal. And
I think that's also completely fine because now I've learned
that stuff.
Speaker 4 (26:19):
Who were the.
Speaker 2 (26:22):
Yeah, I was gonna say, who are some of the
first people you told about, like, Hey, I'm going to
start this up, this is an idea I'm going to
take forward.
Speaker 3 (26:33):
The first people which I told this to. That's a
good question. I don't remember who was like their first one.
I think there's problem. This is in my brain alia
for such a long time, and I thought about, okay,
how to solve this, and then it just happened somehow,
(26:54):
and then I was just on it, just I guess
friends around me. But it wasn't like hey, I'm now
an entrepreneur or how I have now a startup. It
was never like this. It was more like, in the beginning,
like a side project which we started, and then I
got more and more and more. That's a good question.
Speaker 4 (27:17):
It's a given way. You are magnus with green Way.
What are some of your current challenges? What what can
people in the community who are listening to this approach
you with for help and growth and whatnot? So any
anything you want to ask from the community listening.
Speaker 3 (27:41):
Yeah, so, as a young startup, you what I think
someone also learned this. Every startup has assumptions which it
needs to valudate before it finds product market fit. And
for someone in our case, the assumptions, Okay, our city
is really willing to pay x amount of money for
(28:01):
our solution. Is our solution really gonna work scalable in cities?
And so we have a couple of assumptions, like seven
assumptions which we need to validate. And I would say
one of the biggest hurdles now Bottle makes, which we
need to overcome as fast as possible, is just to
demonstrate our solution and as many cities as possible. And
(28:26):
we are right now starting in Ostabrook, the first Chiland
in Germany. But if anywhere around the world, you know,
like cities which are innovative for me, even have contacts
which could be interested, or you just want our solution
in your city because you're frustrated about the traffic lights,
then I'm super happy to connect to and yeah, solve
(28:49):
this problem.
Speaker 4 (28:50):
And in terms of where you are with the solution itself,
I mean the solution is ready to go. You've written
the code, it's been tested and good to go to
we say uh, do a pilot in a city, Is
that what you'd say? I guess.
Speaker 3 (29:08):
So ouri model definitely works to optimize traffic lights in
the simulation, so broot type is demonstrated. But of course
the integration through the real world, that's what we are
right now doing in the in the first city in Ostabrook.
That's a challenge where we will definitely change and a
lot in the and our pipeline, our network. But yeah,
(29:35):
especially doing the next three months, we will be robuster
and robuster and you know, as a startup beginning to
start to hack something together. But as soon as you're done,
more growing business hundred percent sure in two years we
will need to rewrite our whole code and our whole
pipeline to make it more more robust and more scalable.
(29:57):
But that's that's right now, not the not the goal.
Speaker 4 (29:59):
How big is a team right now? Are you a
one man army Magnus or who else is helping you
with this right now?
Speaker 3 (30:10):
Yes? So to other co founders who work on this.
We have an advisor from really clever guy from from
Google who gives us much advice on how to set
up our pipeline, how to build like really good models
which can process in the right direction. And if you
(30:33):
get this grand it's right now. I'm supposed to start
in October. Then we have like a plan with to
eight other people which we will hire to tackle this problem.
Speaker 4 (30:42):
Awesome. So you once you get the funding in, I mean,
you are going to bring in some more smart, smart
people with that funding to make this better. So hiring
is on the horizon. That's that's cool. And I forgot
to ask, I mean, since you mentioned model intermined me.
So are these more is your own custom models magnets?
Are you using any of your any of the open
(31:05):
source models out there for to run your data and
all of that?
Speaker 3 (31:12):
Okay, this is all custom models. We tried many different
architectures from graphel networks to transformers, but yeah, the whole
pipeline is all custom a custom made for this problem.
Is also not that they are like some some large
language models which you know, we know from GVTS or
(31:35):
Lama model which you can just throw at that problem.
It's like more problem where you have to get to
custom development.
Speaker 4 (31:42):
Do you want to wrap it up? Sorry, if you
have any other questions to go for it.
Speaker 2 (31:46):
You can wrap it up with no, I think I
think this is good Nanos.
Speaker 1 (31:51):
Thank you so much for taking the time.
Speaker 2 (31:52):
I like you're really humble, and your like ingenuity here
is at display. So we wish you good luck with
all this. It looks like plan.
Speaker 1 (32:02):
And then again, thank you for taking the time.
Speaker 3 (32:09):
Thank you very much for all your questions and also
your interesting parts on on the safety aspect of it.
Really really looking forward to solving this problem together.
Speaker 4 (32:20):
All the best Magnets. We'll stay in touch and then
what we'll do is we'll publish all the details of
your website and your email and everything when the episode
goes out so people can reach out to you directly.
Thanks so much, Magnus.
Speaker 3 (32:35):
Thank you, Bye Mae.
All right, Hi, welcome everyone, this is another episode. Let's started,
they'll say, and we're talking today with Madnus Mueller from Greenway.
Speaker 2 (00:26):
All right, Mad is welcome.
Speaker 1 (00:28):
Why don't we kick it off?
Speaker 2 (00:29):
Tell us a little bit about yourself.
Speaker 3 (00:35):
I thank you for having me here. Yeah, so two
years ago I was I was hitchhiking around the world
through Central America, through Middle East from Germany, tom lest
through India, Southeast Asia, and half of my time was
wasted at stupid traffic lights. So I thought back then,
(00:56):
like which century we're living in, Like this cannot be true.
So doing a hacker Ton in Germany with some colleagues
from my undergrad and cognitive science so neuroscientists, we tackled
this problem and we built the first PROTECT pilot project
for that we could actually manage in a simulation to
(01:17):
like release skyrocket, like all the metrics like Seal two emissions,
lowered downfield usage, waiting times, and we thought, wow, this
this is incredible. And yeah, since then we are on
this journey for almost two years now to tackle this
problem with a vision to both every single traffic light
(01:38):
on this planet. And while doing this, I moved from
my undergrad to my to my master program in cyric
at Etah, where I'm surrounded with many motivated people, especially
around the student project house. It's like an accelerator where
you're forbidden to study on your project on your university courses,
(02:01):
so you have to work on your court, on your project,
and I just love to be surrounded by people push.
That's that's how I came here today.
Speaker 4 (02:14):
So, Magnus, I mean, why did we double click on
the problem itself for a little bit, right, for people
to understand better? So you said you were hitchhiking. Obviously
that is a challenge by itself, right In addition to that,
you are stopping at two way too many traffic lights
and spending a lot of time. So uh, for people
(02:36):
who probably don't know too much about traffic lights and
how it works today, right, I mean, is what goes
on today in most of the world. You think, right,
why why people stop at these lights? And maybe it's
a primitive way of doing it, maybe it's not the
best optimized way to do it, But I mean, what
(02:57):
is the real problem itself? Can you maybe two minutes
on that?
Speaker 3 (03:01):
Yes? So I guess everyone who lives in a city
with traffic lights knows the situation, like fix them in
the middle of the night where you drive and you
face a red traffic light and they think, okay, house
is possible, like there's no car intersection? Why do I
have read? And of course different cultures deal different with
that situation. Some just go over some way. But if
(03:23):
you think that problem, which is now really clear to everyone,
fix some to the rush hour okay, where you have
thousand cars from every side and you have there a
stupid traffic lights which in many cities just run on
static cycles, so every side just gets the same amount
of brief time, no matter what's the current traffic, what's
(03:43):
the weather, if there was a big football game and
there are now thousand people coming from one side. So
in many cities they have like just still stupid traffic
lights and some for someone in Germany they now started
to tackle this problem by investing like millions of euros
and too expensive sensors, so they open the street, play
(04:04):
some pressure sensors or install cameras to then optimize single intersections.
But if you think about that like nowadays, especially why hitchhicking,
I realized every day around the world uses mobile phones
in their car and with that with Google Maps, tom
Tom Rixia many providers, you have globally available real time
(04:24):
data for the whole planet, and based on that data,
we build models to optimize those traffic lights, not just
for one intersections, but for whole cities. So those interactions,
they can talk to each other. You can say, hey,
I send to you now ten cars coming this way.
With this communication, they can have a much more efficient
(04:45):
optimization for the whole city.
Speaker 4 (04:48):
Got it? So any have you done any sort of
research magnets to see how big this market is? You know,
obviously you obviously seem passionate about the problem itself, right,
But I mean for you to spend time and say, oh,
I want to solve this. I want to solve this
by starting a company. I mean, what sort of numbers
(05:09):
market cab or market size rather any research numbers you have, yes.
Speaker 3 (05:15):
Yes, of course, so right now we'll do a lot
of fund raising. So I'm familiar with all this numbers,
Like the total available market is around ten billion dollars
in that space with a forty percent growth rate. Because
for example, in Germany there are now many smart city
Also in India, I saw there like one hundred new
smart cities so the government like really pushes those initiatives
(05:38):
as different reason has stealed two and sustainability issues, but
also air pollution. Also in Germany, Many, many cities are
just don't have enough people to manage their traffic lights
like manually, so they invest more and more in those
automated solutions.
Speaker 4 (05:55):
Got it. So it obviously affects us, say every day everyone,
I mean anyone who's on the road, I mean whether
you're a pedestrian or whether you're driving a car or
you're driving a commercial truck. So that's a very unique
problem that you're trying to solve. So obviously sorry, good.
Speaker 3 (06:16):
Yeah, yeah, you're completely right, Like it impacts all of us.
And if you think it impacts in so many different industries.
If you, for example, think okay, you're sitting banganor and
you're now ordering your food and your food delivery, it's
like maybe for you those couple of seconds which is late,
isn't a big difference because of stupid traffic lights. But
if you think, like for such a global business there's
(06:38):
like millions of trips every day, a couple of seconds
per trip is like immense and you can think that
from the logistic the truck through the bicycle industry like
it just impacts basically everything understood.
Speaker 4 (06:53):
So we've spoken about the problem. I think we have
a fair idea of what the problem is. So I
see greenway has an AI in it. So what is
your approach to solving this problem with greenway?
Speaker 3 (07:09):
Okay, so what was done until now was you look
at the city and you calculate, Okay, this traffic light
is five hundred meters away from this traffic light, so
we need to switch this traffic lightlight thirty seconds later green,
so that cars or bicycles, whatever you want to optimize,
have kind of a seamless flat green greenway, steamless flow.
(07:32):
But of course for us humans, that's hard to do
that with twenty different lodes, twenty different intersections and their
their complexity, the interactions between them, and that's where our
OAI comes into. Plank we have. We can create a
simulation of a city stack bangal or here, San Francisco, Berlin.
Create a simulation and then in a simulation we can
(07:55):
simulate everything. We can simulate cars, bicycle, trucks, buses, drives
around like in a real world. Now we can train
in AI, which in the beginning just completely stupid control
the traffic light. You can imagine it's like a baby
which comes to the world and there's just stupid actions
(08:17):
like putting the hand into the fire and gets feedback
and gets better and better and hair and like this.
You can also imagine our AI. We can throw thousands
of different like traffic situations on it. We can rush
our middle of the night, big football game, and all
those different data distributions can then get learned from our
(08:39):
system to get better and better. You can just choose
what it should optimize. Some cities they want to optimize
for speed, some for seal tool emissions, some they want
to improve buses or want to prioritize bicycles, and based
on what the city wants, you can then design your
optimization function to optimize statinism. Then, in the second step
(09:02):
in the real world, you don't take the data from
some Google maps or from sometimes feed it into your system,
which is learned in assimilation and then outputs the best
option to now in real time based on this data.
Control the traffic.
Speaker 4 (09:19):
Got it, So I think I get it, But at
the same time, I mean I'm a little I'm trying
to understand a little better on the real time data
of it.
Speaker 3 (09:30):
Right.
Speaker 4 (09:30):
Obviously with AI, you're training it with data that you
already have based on different permutations and combinations, like you said.
But at the same time, I mean, you know, traffic
is always unpredictable, right, I mean, you know, a group
of hundred might choose to have a party and they
might be going in you know, fifty different cars, which
(09:52):
is kind of an outlier, right, But I mean, how
do you so you have a combination of historical plus
real time data being fed and constantly kind of training
your model or updating area. Is that the right way
to think about it?
Speaker 3 (10:08):
Exactly? Yeah? So of course, like it's about making your
models better and better and better and fitting more and
more outlier cases into your model. And for example, if
you take like, yeah, traffic, like you say, is such
a huge diversity, Like you can just imagine, okay, here outside,
is now a car from Amazon Prime? Then then maybe
(10:29):
blocking the traffic light. Right, There's so many different cases.
But by creating a model which is trained on millions
of different.
Speaker 5 (10:37):
Data distributions and then you deploy them practice, and of
course you will figure out in the beginning cases which
are unexpected where which you need to handle for but
over time, like self driving cars to get better and
better and better from.
Speaker 3 (10:53):
Learning more and more and more regarded data distributions. Understand, yes,
this is interesting, the cool thing here?
Speaker 2 (11:05):
Make you have messed this up?
Speaker 3 (11:06):
I think?
Speaker 4 (11:11):
Then I'll go ahead.
Speaker 1 (11:13):
Yeah, I was gonna ask, at least the solution is
going to sit through with the state level agencies or
the traffic agencies. How has that interaction been, right? I
get like I would be curious to hear your reactions
with them.
Speaker 3 (11:34):
You mean to go to market, so how we sell
to the city.
Speaker 1 (11:36):
Yeah, like, because once you would have exposed the solution,
I'm sure they're like, oh, why our system is not
good enough? Like was that the reaction of people who
were very open to oh, look that there is a
better way of doing it.
Speaker 3 (11:52):
So for sure there are some like like you have
to understand and all our customer interviews, we realize those
cities especially they have nowadays a huge problem with skill shortage,
so they don't have enough people to handle those problems.
And for example, in normal small, medium sized cities, there
(12:13):
are hundreds of construction sites every year which changed the
road network and to handle those road blockages, those cities
they pay expensive factional engineering offices who then create new
traffic signals schedules by hand for those intersections. So by
(12:34):
doing those customer interviews, we realized, okay, with those skill
shortage where they really don't have enough people to look
for anything, we can have one of the biggest value
proposition to them. And so you really need to figure
out I mean, in the end of course you benefit
the whole society, but in a first step, like you say,
you sell the cities. So first you need to figure
(12:56):
out why would the city buy the solution, why would
they waste time on you? And one is skill shortage,
another one is still to a must reduction, and third
one is for example, direct cost savings in the the
public transport sector when they need less buses to keep
(13:16):
up the bus frequency. Imagine the bus in the city
wants that every thirty minutes is a bus in front
of your house which can pick you up. And if
now traffic is poor, they need more buses. But if
you improve your the system, then they need less buses.
And such a bus is super super expensive, like normally
(13:37):
they keep one for like ten years, costs around one million,
so to handle them for one year costs around one
hundred to more than two hundred thousand euros per year,
and that's where you can provide them like a value proposition.
And if you have like those three arguments, then you
(13:57):
can go to cities and of course, like we have
heart problems as a young startups that they to get
trust off those cities, and our.
Speaker 6 (14:05):
Strategy is to go with with pilot projects, which we
offer them almost for free to say, okay, let us
try this out so you can discover how much you
benefit from that and.
Speaker 3 (14:19):
Over time win more and more trust off those cities.
Speaker 4 (14:24):
Got it, Thanks, thanks my migness. So one of the
I don't know if it is you call it a
residual effect or a side effect. This actually can be
a big selling point I'm assuming is what goes hand
in hand is if you regulate traffic efficiently and you
regulate traffic safely, you can avoid a lot of accidents
(14:48):
as well. Right, I mean, you know why do most
of the accidents happen because you know, people are in
a rushia. People are frustrated because you know, the traffic
light is in red for so long and they take
a you know, a risk to just cut across the
street and go or you know speed or whatnot. So
I mean, is your solution with Green Bay trying to
(15:11):
pitch you know, public safety on the roads as also
one of these effects. I mean, is that something you've
thought about as well.
Speaker 3 (15:21):
So we are aware of studies which show that if
you improve traffic flow so less stopping code traffic, then
this increases road safety. And we can clearly improve traffic
flow somewhere in simulation in rush hour we can sometimes
reduce a number of waiting vehicles, but up to ninety percent.
(15:43):
And but asn't now we have never used that fact.
I think that's something which especially needs to figure it out,
like what clear percentage we can say, or like how
many accidents less are there. It's a thing which is
really hard to measure for us right now. But long term,
like this just goes hand in hand and with my vision.
(16:05):
And the cool thing here is that if you improve
one factor, then you improve so many other factors. You
improves your two waiting time, fuel usage, road safety, stability.
That's one thing which really amazed me about this problem.
Speaker 4 (16:23):
Awesome magnets. I mean, it's a it's a great idea.
Obviously you started this based off your experience traveling the world,
so you didn't see it in just one place. I
mean this is applicable kind of in every city around
the world. I mean, you know, you've you're obviously from Germany,
You've seen the traffic in in Bangalore. So do you
(16:46):
do you think, I mean greenway will work for any
permutation and combination sort of thing, or in other words,
it will it work for any any city you think,
any modern city so to speak, or you think you
have a sort of a template for an ideal city
where greenway will work now and maybe work elsewhere later.
(17:10):
What is your thought on that?
Speaker 3 (17:14):
So, of course we have our initial target customer group
where we start with this, but our plan is to
make this work everywhere, to solve of every single traffic
light on this on this planet. And of course some
cities are faster, somewhat slower, So we focus, especially in
the beginning, just on those who love us and really
want to develop it just together with us. And our
(17:36):
vision is like long term to make the whole process
around smart cities really easier for cities. Right now, imagine
you're a developer and you want to develop your app
for the App Store, so that everyone can benefit from it, right,
so you just go to Google play Store or App
Store and to upload your app. But in cities this
(17:58):
is right now, super super hard because every city uses
a different system and you have to do tenders. But
like my long term vision is that long young developers
and there'll be in India and Germany or in the
US if the ideas can come to a city and
an easy framework can bring in their ideas and easy
(18:21):
sandbox like format to run experiments on some infrastructure from
small they save parts of the city to test out
new stuff. And I think with that we can have
in so many different ways new innovation which benefits in
the end all of us will live in those cities. Great.
Speaker 1 (18:43):
I'm still really more of the fact that it's hitchhiking
that got.
Speaker 4 (18:46):
You to the solution.
Speaker 1 (18:47):
To be honest, I would be like, you weren't even
driving at that point, you still got bothered with the
traffic lights. That's just seriously amazing. But it's also it's
the scale of the problem too, right, This is you
said it yourself, right, You're starting with a small setups
that's the one city in then expand globally.
Speaker 2 (19:09):
That's a massive vision, right, how is it and how
is your day to day right when you go out
and like, Okay, I'm gonna achieve this one at a time,
how does that day look like.
Speaker 3 (19:22):
My day to day life. So it really depends on
the face of the business. Like sometimes I remember two
months ago, I was coding day and night, just thinking,
brainstorming on like a flip shot, you know, and drawing
architectures of our AI model. How's the whole pipeline going
(19:44):
right now? It's a whole day like fund raising. Right
now we have fund raising half a million euros and
so I reach out to many investors right now, I'm
here in Silicon Valley and just pitch I worked a
lot along pitch Deck, learned a lot to improve there.
About legal stuff, I learned a lot. So I think
(20:06):
that's it's one of the most amazing things to be
a founder, to just get insights into so many different domains,
learn about legal, learn about hiring, learn about tech, to
be a generalist. And I think especially for crewious people
like us, they're one of the most amazing things to do.
Speaker 4 (20:24):
So for anyone listening Magnus, you are fundraising and you
you're maybe again based on what I know I will
say it and you can correct it. You're looking for
angels for a start, right I mean you're in an
initial seed face or you can probably talk about what
sort of funding you would rather take in right now.
Speaker 3 (20:49):
Yes, So we got nominated for like a three million
grant in Germany where we want to hack at them
at the federal Ministry. And to secure this grant, we
are our saying five hundred k and we have already
the first angels committed. It's all around pre seed, like
we're especially looking for people who bring smart money on
(21:12):
the table, who wants to bring a network on a
table in the mobility space. Can be a tech space
advisor from from Google, can be also in the government
sector like everything. Just someone who is also frustrated about
this problem and wants to solve it. I'm happy to
talk with. Can be pre seed be cs or or
(21:33):
angels who really believe in that mission to solve that
problem globally. And it's crazy if you do the calculation,
like how many million tons of seal two we can
save with that solution, Like it's insane and that's really
what what motivates me.
Speaker 4 (21:49):
To somebody anyone with money in the bank and a checkbook. Right,
that's that's that's that's what you're looking for.
Speaker 3 (21:57):
Yeah, basically someone who brings a new new ideas.
Speaker 4 (22:02):
I'm always so I want to do a little bit
of a flashback maybe, right Obviously we've talked in about
enough about your hitchhiking, so to speak. But I mean,
did you always want to be an entrepreneur or it
just happened by chance?
Speaker 7 (22:24):
I think in the world where I grew up right now,
it's like a really hype it's all around ath Agent's
Rakes project house that was an esoprick.
Speaker 3 (22:35):
Like, Okay, maybe maybe I say this also because I'm
in this, baba, it could be also true because everyone
around me is like an entrepreneur. But yeah, I think
like also, when I was younger growing up and I
had problems, I always thought, okay, like what's the chance here,
Like I I remember I was sixteen, I'm selling like
(22:59):
Tisha designs online and I need to upload those designs
to like Ama Zone, and I just wasted my time
with uploading. So I just started, okay, writing my own
boughts to to upload that stuff. Just whenever if you
face a problem, I thought, okay, okay, what chance is here,
what solution could be here? Which also I don't better
(23:22):
fits from it, and I think this is what an
entrepreneur does in the beginning.
Speaker 1 (23:28):
Yeah, do you still think it's all hyped up being
an entrepreneur?
Speaker 3 (23:39):
I think right now a lot here in my age,
like many young people want to be really independent and
want to create their own company. And I think like
right now, especially online, like we have so many tools
which support us with that makes it easier and easier
to just go out and co found is went investors
(24:02):
great first MVPs, like the structure to do that is
become much.
Speaker 1 (24:08):
More Would you recommend caution or would you be more
like you know, hey, more power at this point, like
knowing what you know now?
Speaker 3 (24:18):
Yeah, if I would recommend to be an entrepreneur, yeah,
so I think, yeah, of course I wouldn't. I would
recommend it. Yeah, especially if you're like curious, like if
you feel inside of you that you want to be
it and there's just a little bit, you know, maybe
a little bit of fear or just can't be anything
which holds you back, then I would definitely recommend it
to you. But I think we often forget also in
(24:41):
big companies they are often like really entrepreneurs, you know,
like even if you're working in a big tech or
medium sized company, you can be an entrepreneur and really
innovate that company and run experiments and many examples where
they do everything with an entrepren basically does I think
(25:02):
the advantage of just being like in my situation is
just that you're kind of start from scratching. Really you
learn like everything from legal stuff.
Speaker 8 (25:11):
To product to investors through hiring, and that's just really
an amazing experiment experience which you don't get and working
in big type.
Speaker 3 (25:23):
But the mindset itself, you cannot everywhere.
Speaker 4 (25:28):
Anything you would have done differently in your journey. Magnets
looking back.
Speaker 3 (25:36):
A hundred things, one hundred things, like I think I
have every day. I have like a list with things
which I would do differently, little beaches, personal habits or
especially in the in the startup, like with some of
how we set up the legal stuff during the beginning,
we didn't have the best thing and then like no
(25:59):
way are really looking okay with what people you start
the company so you fully fully trust them, and thats
a I think a long long list with things that
would I've just learned on the on the goal. And
I think that's also completely fine because now I've learned
that stuff.
Speaker 4 (26:19):
Who were the.
Speaker 2 (26:22):
Yeah, I was gonna say, who are some of the
first people you told about, like, Hey, I'm going to
start this up, this is an idea I'm going to
take forward.
Speaker 3 (26:33):
The first people which I told this to. That's a
good question. I don't remember who was like their first one.
I think there's problem. This is in my brain alia
for such a long time, and I thought about, okay,
how to solve this, and then it just happened somehow,
(26:54):
and then I was just on it, just I guess
friends around me. But it wasn't like hey, I'm now
an entrepreneur or how I have now a startup. It
was never like this. It was more like, in the beginning,
like a side project which we started, and then I
got more and more and more. That's a good question.
Speaker 4 (27:17):
It's a given way. You are magnus with green Way.
What are some of your current challenges? What what can
people in the community who are listening to this approach
you with for help and growth and whatnot? So any
anything you want to ask from the community listening.
Speaker 3 (27:41):
Yeah, so, as a young startup, you what I think
someone also learned this. Every startup has assumptions which it
needs to valudate before it finds product market fit. And
for someone in our case, the assumptions, Okay, our city
is really willing to pay x amount of money for
(28:01):
our solution. Is our solution really gonna work scalable in cities?
And so we have a couple of assumptions, like seven
assumptions which we need to validate. And I would say
one of the biggest hurdles now Bottle makes, which we
need to overcome as fast as possible, is just to
demonstrate our solution and as many cities as possible. And
(28:26):
we are right now starting in Ostabrook, the first Chiland
in Germany. But if anywhere around the world, you know,
like cities which are innovative for me, even have contacts
which could be interested, or you just want our solution
in your city because you're frustrated about the traffic lights,
then I'm super happy to connect to and yeah, solve
(28:49):
this problem.
Speaker 4 (28:50):
And in terms of where you are with the solution itself,
I mean the solution is ready to go. You've written
the code, it's been tested and good to go to
we say uh, do a pilot in a city, Is
that what you'd say? I guess.
Speaker 3 (29:08):
So ouri model definitely works to optimize traffic lights in
the simulation, so broot type is demonstrated. But of course
the integration through the real world, that's what we are
right now doing in the in the first city in Ostabrook.
That's a challenge where we will definitely change and a
lot in the and our pipeline, our network. But yeah,
(29:35):
especially doing the next three months, we will be robuster
and robuster and you know, as a startup beginning to
start to hack something together. But as soon as you're done,
more growing business hundred percent sure in two years we
will need to rewrite our whole code and our whole
pipeline to make it more more robust and more scalable.
(29:57):
But that's that's right now, not the not the goal.
Speaker 4 (29:59):
How big is a team right now? Are you a
one man army Magnus or who else is helping you
with this right now?
Speaker 3 (30:10):
Yes? So to other co founders who work on this.
We have an advisor from really clever guy from from
Google who gives us much advice on how to set
up our pipeline, how to build like really good models
which can process in the right direction. And if you
(30:33):
get this grand it's right now. I'm supposed to start
in October. Then we have like a plan with to
eight other people which we will hire to tackle this problem.
Speaker 4 (30:42):
Awesome. So you once you get the funding in, I mean,
you are going to bring in some more smart, smart
people with that funding to make this better. So hiring
is on the horizon. That's that's cool. And I forgot
to ask, I mean, since you mentioned model intermined me.
So are these more is your own custom models magnets?
Are you using any of your any of the open
(31:05):
source models out there for to run your data and
all of that?
Speaker 3 (31:12):
Okay, this is all custom models. We tried many different
architectures from graphel networks to transformers, but yeah, the whole
pipeline is all custom a custom made for this problem.
Is also not that they are like some some large
language models which you know, we know from GVTS or
(31:35):
Lama model which you can just throw at that problem.
It's like more problem where you have to get to
custom development.
Speaker 4 (31:42):
Do you want to wrap it up? Sorry, if you
have any other questions to go for it.
Speaker 2 (31:46):
You can wrap it up with no, I think I
think this is good Nanos.
Speaker 1 (31:51):
Thank you so much for taking the time.
Speaker 2 (31:52):
I like you're really humble, and your like ingenuity here
is at display. So we wish you good luck with
all this. It looks like plan.
Speaker 1 (32:02):
And then again, thank you for taking the time.
Speaker 3 (32:09):
Thank you very much for all your questions and also
your interesting parts on on the safety aspect of it.
Really really looking forward to solving this problem together.
Speaker 4 (32:20):
All the best Magnets. We'll stay in touch and then
what we'll do is we'll publish all the details of
your website and your email and everything when the episode
goes out so people can reach out to you directly.
Thanks so much, Magnus.
Speaker 3 (32:35):
Thank you, Bye Mae.
Popular Podcasts
CrimeLess: Hillbilly Heist
It’s 1996 in rural North Carolina, and an oddball crew makes history when they pull off America’s third largest cash heist. But it’s all downhill from there. Join host Johnny Knoxville as he unspools a wild and woolly tale about a group of regular ‘ol folks who risked it all for a chance at a better life. CrimeLess: Hillbilly Heist answers the question: what would you do with 17.3 million dollars? The answer includes diamond rings, mansions, velvet Elvis paintings, plus a run for the border, murder-for-hire-plots, and FBI busts.
Crime Junkie
Does hearing about a true crime case always leave you scouring the internet for the truth behind the story? Dive into your next mystery with Crime Junkie. Every Monday, join your host Ashley Flowers as she unravels all the details of infamous and underreported true crime cases with her best friend Brit Prawat. From cold cases to missing persons and heroes in our community who seek justice, Crime Junkie is your destination for theories and stories you won’t hear anywhere else. Whether you're a seasoned true crime enthusiast or new to the genre, you'll find yourself on the edge of your seat awaiting a new episode every Monday. If you can never get enough true crime... Congratulations, you’ve found your people. Follow to join a community of Crime Junkies! Crime Junkie is presented by audiochuck Media Company.
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.