35C3 - A Blockchain Picture Book

35c3 preroll music

Herald: I'm very happy to be allowed to
announce the following talk. It's called

"A Blockchain Picture Book" and it's held by
Alex. Blockchain is a topic we hear

constantly everywhere in the media. It's
sold as a solution for everybody and for

everything. Doesn't matter what. But how
it really works and what it really is...

...most people don't really have an idea. So,
today we got a beginner's introduction

with pictures. So, please welcome Alex on
stage with a big round of applause.

Applause

Alex: Hi. Yeah. My name is Alex and I'm
from the Technical University of

Brunswick. I'm working in the Institute of
application security. I suppose you're

here to see my blockchain picture book,
right? But right from the start, this talk

is not about the newest cryptocurrencies
out there and it's also not about in which

ICO should invest in. I'm absolutely
clueless about that topic. Sorry to

disappoint you, maybe. I'm here to give you
a very rough introduction to the idea and

the basics of this blockchain thing out
there. So, maybe you can judge by your own.

So. I hope there are no blockchain experts
in the audience, maybe. So this is a

foundation talk. That means I do not
expect any technical or mathematical

backgrounds at all. So, let's just give it
a try. Imagine the following situation.

Once upon a time, there were two people.
Let's call them "Mary", on the left, and "Alex",

on the right. They wanted to trade with
each other. Mary she wants to buy a house

from Alex. Thus , she brings the agreed
amount of money. And Alex – on the picture –

brings the keys for this house that
should be sold to Mary. But, since Mary and

Alex do not know each other, they do not
trust each other, as well. So, they set up a

contract for this trade. Just to be on the
safe side. Here on the right, you can see a

typical contract for trade, and it consists
of some important informations like a

description of the object that should be
sold: The house of Alex, maybe the address,

what's inside and stuff. the agreed amount
of money – let's say the house costs half a

million, whatever. The persons who are
involved in this trade – in this case Mary

and Alex. And of course the signatures. As
a final step, this contract must be

timestamp for being valid. So, nowadays
those kind of contracts usually exist as

digital documents. But for both the
digital and the analog document, if an evil

party gets hands on all copies of this
contract, he can modify it in a malicious

way. Let's say, removing the timestamp.
This can void the contract and thus make

the trade invalid. By the way, the evil
party can also be one of the persons

involved in this trade. So, let's say our
Alex – the evil one – he wants to take the

money from Marry but also keep the house.
Okay? This is our initial situation. So in

1990 Haber and Stornetta proposed in a
scientific paper a cryptographical

solution for binding a time stamp undevisably
to a digital document. What does

it mean? They both – the timestamp and the
digital document – are merged into a new

document and can't be separated from each
other without being damaged. So, we'll see

later how this exactly works. For now,
let's pretend this block, which you can see

here, contains our digital document and the
timestamp. And both are binded by some

mystical fancy cryptographic technique
which is represented by this blue frame.

Okay? So, unfortunately, we still face a
problem: That's our evil party – remember

Alex is the evil party? – He can just make
this contract and all copies disappear and

claim they never existed before. So, to
tackle this problem, Haber and Stornetta

and their paper further proposes a
chaining technique for those blocks. All

of which contains of course a different
contracts. So, the result is a chain of

blocks: The blockchain. Here we are. Hey!
And since this is a chain, it's not

possible to remove one of the blocks.
Removing a block from the inside of the

chain, will just break it completely, in a
way that you can't put it together anymore

without damaging it, as we will see later.
Okay? So, this doesn't prevent that anyone can

just make this whole chain disappear.
Maybe. So, but the longer the chain gets,

the more people are involved in this chain,
the higher the risk for evil parties

malicious task is discovered. OK? Clear so
far? Right. But the problem remains: This

chain has to be stored somewhere. You
remember it's digital? So let's say

another trusted party gets this task. So,
here we are. This is our trusted party

which is in charge for blockchain. But, the
thing is this trusted party can also fail

or become untrusted. And it doesn't have
to be malicious at all. It also may happen

that the storage for blockchain fails or
crashes or any data may be destroyed due to

any hardware failure, maybe, or a trusted
party just deleted accidentally. Maybe. So,

it's like with every centralized service
you have to trust. So, imagine your e-mail

provider just disappears tomorrow. How
many services are you using which depends

on the e-mail address? OK. So, for that
reason, the paper suggests to distribute

copies of our blockchain to a lot of other
parties. So, here we are. Here you see a

lot of other parties – whoever they are – who
owns a copy, an exact copy, of a blockchain.

Remember, this blue color represents
something cryptographical. OK. So, you can

see there is no centralized storage
anymore. Instead we have a large number of

participants which we can rely on. So,
let's call this a decentralized network.

Remember this term. It will get important
later. So, for now, even if some of those

parties fail, as long as the majority
stores are valid and of course the same

copy of a blockchain, we can keep going.
OK. You're still with me? Right. OK. So, the

question remains: How does mystical
cryptography thing, stuff, whatever works.

laughs

This is how such a block actually looks
like. And in fact, this is the first

blockchain block ever created. But, I
suppose you're not really interested in a

math lecture. OK. So, as I promised, let's
keep the mathematical details and stay

visual. But just for information: This
is a block. Okay. Let's pretend, we observe

another trade. Alice and Bob are doing a
trade and we can see, Alice wants to sell her

smartphone or whatever to Bob. So, do you
remember those polaroid cameras, where you

could take a picture and it was printed
instantly? You know them? I think they're

back in fashion now. So, while we observe the
situation, we take a polaroid picture of

it. It's printed, and this picture attests
this trade. So, we take this picture and

put it on a blackboard. Like this here. So,
let's say, this picture represents a block.

You remember these blue framed blocks like
we saw them before? OK. Somewhat later,

we observe another trade. Here, Charlie
wants to sell his car to Peter. And again,

we take a picture. But this time, we hold
the picture from before – the one, which is

already on the blackboard – into the
background of the situation. Like here. So,

now, we take the new Polaroid picture and
put it back to our blackboard. OK? So, what

do we see now? The first picture, which is
our first block, appears in the background

of the second picture, which is our second
block. So, let's say the second block is

chained to the first block. So, if someone
evil wants to change the content of the

first block, he must also change the
content of the second block. Which means

more effort for this evil party. Ok? So, of
course let's keep doing the same. And add

the third picture to the board. Here we
can see Eve and Alex, another Alex, trading

some goods. And in addition again we can
see the second picture. It's in the

background of this third picture, so it's
the same thing. The third is chained to the

second. OK. You're with me? Nice. And of
course for the sake of completeness we add

the fourth block where Armin and Ed are
trading. And this house is more expensive

this time. So again, we'll see the content
of the third block in the fourth block.

That's their chain. So now we can see the
big picture here. So if our evil party now

wants to change something on this first
trade on the first picture, that means he

must change the second and the third and
the fourth picture as well. Else this

modification is detectable. You agree? OK.
So one can say we created our very special

kind of blockchain here, consisting of
polaroid pictures, which represents blocks

of course. OK, so this effort can bring us
some fancy properties. And at this point

it's important to say that some of these
properties are optional. There are

different kinds of blockchain systems out
there and most of the properties depends

on what kind of blockchain is in use and of
course the use case. But let's just take a

look on some of the important properties,
in my opinion. So return to our blackboard

here, which we constructed earlier. So we
already observed this chaining mechanism.

This picture in a picture thing. You
remember. So this mechanism makes our

blocks immutable and thus the contracts
inside these blocks are immutable as well,

of course. So they can only be modified if
the complete chain is also modified. So

let's take our board and place it anywhere
in the public. This gives all the people

walking around the possibility to keep an
eye on everything happening on our chain

on the board, and maybe detect any
anomalies. So we can say, well, blockchain

is the coolest open source software,
maybe. Other parties who make up our

distributed network - you remember we
talked about that before - creates an exact

copy of our blackboard, and the updates,
they copy every time something new

is appended to the chain. So if the original
blockchain on this board becomes invalid

or disappears, or whatever, we can ask our
distributed network for the last valid

copy of this chain and put it back on the
board. But even if some of those parties

becomes malicious, as long as the majority
of all parties is honest and own valid

copy of our block chain we could prevent
fraud in it. And at least it's possible to

ensure anonymity for all involved parties.
Parties of our distributed network, as well

as the parties involved in the contracts
inside the blocks. OK, so in summary we can

say the contracts in our blockchain are
immutable. The control over the blockchain

is decentralized and it's really hard to
get the majority of the network to the bad

side, I suppose. Our block, our board and
thus the blockchain and side are exposed

to public scrutiny and it's not possible
to remove one of the blocks from inside

the chain without breaking it completely.
We can provide anonymity to our distributed

network and the parties
involved in the transactions or the

contracts inside the chain and at least we
don't have to be at the mercy of a single

institution, like a bank or government or
whatever. Okay, so looks like some fancy

properties. This are just some of a lot of
properties which we can achieve. So let's

take a closer look on a common blockchain
application. I suppose most of you here

heard of the so-called cryptocurrency
stuff, especially bitcoin. Maybe some of

you own some bitcoin. Of course there are
many more cryptocurrencies out there:

Woppercoin, OneCoin(?)... stuff. But
let's talk about financial transaction on

a blockchain in general. So this is a
classic German bank transfer form. Before

the times of online banking, this form was
used to transfer money from one bank

account to another. Maybe some of you
still remembered, you had to write

extremely clear - it was horrible - and you
filled out this form, gave it to your bank

and they take care of everything else. So
you can see some fields here: These are

the names of the sender and receiver of
the money. The sender is on the bottom.

Here you can see the account numbers. This
is the bank ID, and in this case you can

see it's the same bank. Of course the
amount of money and the signature of the

sender, as well as the date. So for
financial transactions via

cryptocurrencies like Bitcoin whatever, we
don't need that much information. The only

information we need here, are the sender's
and receiver's public keys. They are used

the same way as the account numbers, but
they're much longer, more cryptical.

Of course, still we need the amount of
money and the sender signature, which

also contains the timestamp. Remember this
stuff is digital. So back in the good old

days we just fill out this form, put it
into a special letter box in the bank and

forget about it. In cryptocurrencies, very
simply said, all commission transactions

like this here are collected by our
distributed network. So let's say our

distributed network collected seven
pending transactions. So, but these

transactions are not in our blockchain
yet. Let's get back to our blockchain.

We have five blocks currently with any
contracts or transactions inside. So now

we face the question which of those
transactions shall form our sixth block in

this chain. OK, you agree on this question.
Maybe some of you would say: Hey, we have

some timestamps in the transaction. Why
can't we use them. The thing is that our

distributor network is in different time
zones and, I don't know, network delays

and all this stuff. So we need another
solution for this question. I suppose some

of you already have an idea how to use
this. So if all involved parties follows a

rule on how to find an agreement on
something, whatever, we can call this a

consensus protocol. So maybe some of you
heard this term before. Okay, so remember

it. Let's replace transactions and
contracts with something much simpler.

Let's say some fancy symbols have been
proposed as candidates for our next block,

the red one in the chain. So our
distributed network, consisting of some

participants, now face the question. Which
of those symbols shall form the next

block. And there are different ways to
achieve this goal. So let's talk about the

best known consensus protocols. Maybe one
of the most popular used by the bitcoin

network, it's the proof of work protocol.
I see some people laughing. OK. So what

does it got to do with competition. Back
to our initial question, which of the

symbols shall form the next block. So very
simply said each time a new block shall be

created. And this means as long as there
are pending transactions we'll need new

blocks. Our network holds the public race
and everyone can take part in this race.

So the winner of the race can decide which
symbol gets into the new block. Looks

simply. And in this race we can see that
the winner party selects the heart symbol

for the next block. So it appears there.
So the outcome of this race is known to

our distributed network. So each of those
parties in this network appends the new

block into their own local blockchain copy
holding this heart symbol inside. OK?

Right. Unfortunately, as some of you may
know, this protocol has some drawbacks and

the most serious of these is probably the
negative impact on the environment. That

means holding such races consumes many
resources. And of course in reality not

motorcycles, but a very large number of
highly specialized, high performance

computers take part in these races. And of
course they need of a huge amount of power

which is not really great for environment.
So for last year it was estimated that the

bitcoin network needed, at annual rate, as
much power as the whole country of Denmark,

which is a lot. OK, so you don't like this
protocol. Let's take a look on another

one. So this protocol depends on the vote
of participants in the group. We call it

the consortium vote. This is one of the
names for this protocol. Back to our

initial question with the fancy symbols.
Which of them shall form the next block in

our chain? So instead of a huge network of
unknown participants driving motorcycles

and holding races, this time we invited
some parties in a fixed group. We called

this group consortium. So each time a new
block shall be created, the participants

of this consortium hold a vote. It's
similar to a democratic election. So the

symbol most parties voted for wins. It'
that simple. So in this round we can see

two parties voted for the circle two
parties voted for the heart and three

parties voted for a star symbol. So the
star is selected by the will of the people

and thus appears in the next block. You
agree? OK, so the outcome of this vote is

visible again to our distributed network
which holds copies of our blockchain. So

each party in a distributed network
appends a new block into their own local

copy holding the star symbol inside. So,
great, we see no environmental drawbacks

here. Maybe this is the best solution. OK,
some of you disagree. Yeah. Well the

topic here is the formation of the
consortium. This is one of the drawbacks.

We can't ensure that the invited parties
or participants of our consortium are

unbiased. Here we see a very huge doorman
and this doorman only allows participants

to our consortium party who promises to
vote for this circle symbol and nothing

else. So we can say he's kind of a filter.
So since there are only participants who

refers the circle symbol and a consortium
they also vote for it eventually. As you

can see here. So among all the drawbacks,
this protocol may be misused for

corruption or whatever. OK, you're still
with me? Nice. Okay let's check the last

consensus protocol, may we have more
success here. So this one depends on the

share or stake of the system that everyone
owns. This consensus protocol will be

called Proof of Stake. But what does it
have to do with randomness? Maybe you ask.

So one last time we repeat our question,
which of the symbols shall form the next

block in our chain? So let's pretend we
have any system, let's say all existing

symbols in this world form our system. We
can see the system. This is the circle on

the right side. And the whole system
belongs to five parties, which are on the

bottom left. This five color parties. So
this colored circle represents the stake

of each party on the system. So each round
these parties vote on which symbol shall

form the next block. But this time the
decision for who is the winner is made

randomly, taking into account the stake of
each party. So we can call this a weighted

randomness. OK, so how does this exactly
work? Let's imagine this roulette wheel. I

hope some of you know this kind of game.
OK. Usually you spin the wheel, you throw

a ball on it and some time the wheel stops
turning and the ball lands on any number

on this wheel and this number is the
winner then. But instead of numbers this

time we distribute the symbol candidates
on the wheel according of the stake of its

holder. OK, so let's play the game. That's
how our wheel looks like and we spin it

and throw a ball on it. It keeps rolling
and rolling and it stops at some point on

a symbol. Here it landed on the triangle.
So the triangle is the winner symbol for

our next block and appears there
eventually as you can see on the top

right. Okay? You're with me. All right. So
what happens when the stake of one of

these parties is significantly larger
than the stake of others? Here, our red

stakeholder is a very rich man and holds
the majority of the system, so as you can

imagine even if our system chooses
randomly, since this randomness is

weighted, his chance to be chosen is
significantly larger than the chance of

others. So, of course, there are even more
consensus protocols out there and you can

spend a few hours talking about them. But
to get to the point, every effort to reach

a fair consensus seems to come with
drawbacks and depending on our use case

and the blockchain system which we are
using those drawbacks can have negative

impacts. So why are we still putting all
this effort into this blockchain science,

you may ask. Let's get back to our initial
situation, where Mary and Alex traded and

Mary wanted to buy a house from Alex. You
remember? So here Mary has to trust Alex,

that the house he sells to her is really
his own property and that he won't revoke

the trade afterwards. On the other hand
Alex has to trust Mary that she gives him

the promised amount of money for this
house. So if a bank is involved in this

trade they both, Mary and Alex, have to
trust this bank for transferring the money

safely from Mary to Alex, and of course
the right amount of money and right

currency. And in big trades like house
selling, usually a notary is involved.

Mary and Alex have to trust this
notary to be impartial.

So we see, traditionally such
trades involves a lot of trust and

trusting in such things as trades is not
really popular as there has been a lot of

disappointments in the past. As you may
know. And all this blockchain science is

the most promising approach for mitigating
the need of this trust so far. But I think

we're still a long way from reaching this
goal. So let's see what the future may

bring us. Thanks for attention.
applause

Herald: Thank you very much for this nice
introduction. If you have questions please

go to the mics. There's three simple
rules to follow now. First, if you want to

ask a question, a question is normally one
sentence ended by a question mark. Nothing

else. Second if you speak into mic keep it
close to your mouth just like you would

like to bite into it. But please don't do
so. And third for everybody who is leaving

now, please do this quietly. Thank you very
much. So let's start with a question from

the Internet.
Q: Hi. There is a question from IoC. So

you talked about how blockchains provide
anonymity and could you explain what that

means in the context of a blockchain,
because after all it's like real people

having a transaction. So what is the
definition of anonymity in this context?

A: This is a very technical question, I
suppose. For the technical people

anonymity means we don't have names like
we've seen on the German bank transfer

form, where we can exactly see Mary and
Alex are trading. So you have the names

and you exactly know who those people are.
In the technical case, that means we use

public keys. And of course we can create
public keys without seeing the person

behind it. So in this case it's possible
to ensure anonymity, but as some of you may

know, comes some drawbacks. It's,
yeah, technical.

Herald: Microphone number three.
M3: You showed us different ways of

choosing which transaction is next. But I
don't know if there is a possibility that

some transaction will be lost? Like, you
know, it doesn't take -

A: You mean like one of the transactions
just disappears from this cloud thing?

M3: Yeah, because all others are chosen
before it.

Like - or how long does it take to do the
vote?

A: Oh that really depends on the protocol
as some of you may know in case of

blockchain which has this time minute
stuff. That means I think every 10 minutes

we'll do a new round because of this
cryptographical hash whatever, but it

absolutely depends on the protocol. As
this voting protocol works completely

different. So that's the reason why I
didn't want to go into details, because it

depends on the protocol. So we can talk
about it later if you want to but I think

this is out of scope here.
Herald: Microphone number two, please.

M2: Thanks for the talk. A lot of the
examples I see on blockchain are

basically database owned by one company.
So it doesn't really. So what is it for

you the most exciting usage of blockchain
that you see today?

A: I cannot answer this question because I
don't truly know what kind of solutions

are there. Of course I know the popular
stuff like hyper legend whatever. For me

personally the current solution are not
really cool, but because of technical

backgrounds. But I can imagine that the
distributed computing protocols behind

this blockchain thing can lead us to some
very cool stuff. But we need to put much

more science inside it and that means,
really science.

applause

Herald: We have another question at
microphone number two.

M2: Great talk. Yeah. For distributed
computation, there was already a lot of

work.
Herald: Sorry, can you go a bit closer to

the mic please?
M2: Sorry. If one of the most interesting

usage is distributed computation, there
was city at home there is a bunch of bla

bla bla at home so I'm not sure. Then
again, what's the competitive advantage

let's say of blockchain to those previous
distributed computation protocols?

A: I'm not sure if I understand your
question but do you ask, what exactly is

new about all
this thing?

M2: If the advantage is distributed
computation, that was done before. And I'm

not sure what's the advantage of doing it
by a blockchain.

A: Technically a blockchain is absolutely
nothing new. So distributed computing is a

quite old topic it's nothing else than
just a distributed database in fact with

some other properties which can be
combined. So that means blockchain is a

combination, for me personally, of a lot
of techniques. So I think we just should

use many more distributed computing
solutions for stuff like financial

transactions or smart contract stuff
whatever. So this blockchain term you must

be really careful with it.
Herald: That's a very nice warning. So

please give a warm round of applause for
Alex.

Applause

35c3 music

subtitles created by c3subtitles.de
in the year 2020. Join, and help us!