“In the Internet of Things (IoT), it is important to make
sure that we are connected to [the right devices].” “Do you think blockchains can help
in solving IoT security issues?” “Do you know of any IoT solutions
using blockchains to enforce security?” I am quite skeptical when it comes to
blockchains being used for IoT security. One of the interesting ideas we haven’t explored
is using proof-of-work to identify IoT devices… that must do a bit of work to [connect]. But in terms of using blockchains to help solve security
issues, I’m not quite sure what the benefits are. How is it any different than using a local database? What is the benefit of putting this
information on the blockchain? You may be able to use a public, decentralized
blockchain to log information from IoT devices… so that it can’t be changed in the future,
[if the blockchain is immutable]. But a lot of people use the word
“blockchain” to refer to a database… with digital signatures and
public key infrastructure (PKI). It is important to understand that the
purpose of a blockchain is more than PKI. We have had PKI for twenty-five years.
There is nothing new there. It is not particularly interesting to make a PKI database
public, unless you do something [more] with it. For example, a decentralized consensus system to have
immutability. Then again, what problem are you solving? Which problems in IoT security are you solving?
A lot of people are trying to mash these terms, IOT and blockchain, together, just like
they’re trying to mash other terms together. “X plus blockchain,” and pretend it is something new
that X couldn’t do on its own without a blockchain. I’m very skeptical. When it comes to IoT,
the security challenges are very big. There is a joke in security circles:
“the ‘S’ in IoT stands for security.” Of course, there is no ‘S’ in IoT. I am not
sure blockchains are the solution here. “Regarding solar energy trading on the blockchain,
what would be the benefits of using a unique… ERC-20 token versus ether
for buying and selling power?” That is a great question. I don’t know what the
benefits might be. Ether is produced through mining. If you have an ERC-20 token related to solar energy,
perhaps you can mint or issue that token… in response to people generating solar energy. But again, the only way you can measure
how much energy someone is producing… and issue them tokens, is [for someone else] to buy and
use that energy. In that case, they could just use ether. Not all things needs tokens. You should to be suspicious
when an application [incorporates] a token. Not many things need utility tokens.
Utility tokens have a number of limitations. Of the projects that are selecting to use tokens,
are doing so in order to fundraise. More so because they need [money] than because
they need a utility token to operate their application. This is exactly the right question to ask when
it comes to alternative uses of the blockchain. We should all be asking, “Do you
really need a blockchain for this?” Which part of an open, public, borderless, censorship-
resistant, and decentralized blockchain [do you need]? If you [don’t], if you don’t care that the system [has
those characteristics], you don’t need a blockchain. What do you need a token for?
What functions are you using that requires a token? Why can’t you just use either?
Why can’t you use something else? When it comes to tokens, is there something in
[your] operations that requires a smart contract… in order to monitor, adjust, or control the
usage of that token based on some rules? There are some cases where that [makes
sense]. Absolutely. But those are fairly rare. [Oftne], the token is just an excuse to fundraise.
You should be very skeptical of those things. [ANDREAS] Nice to meet you.
[AUDIENCE] Nice to see you. My question is… Recently, I [saw] a presentation on YouTube that the
killer app in blockchains will be a security market. [ANDREAS] Will be what? [AUDIENCE] The killer app in
blockchain technology will be a strong security market. [ANDREAS] A securities market. Was this
one of my videos or from somebody else? [AUDIENCE] Somebody else. [ANDREAS] Okay. [AUDIENCE] So, blockchain technology
will dominate [the securities] market… in ten years or so.
What do you think about the idea of that? [ANDREAS] I think that is very specific.
I try not to make specific predictions. The trick to making predictions is being
vague about the outcome or the timeline. That is how you can be “successful” [at predictions].
It is a bit like the Heisenberg uncertainty principle. If I make a prediction, I can tell you what will happen,
but not when. I can tell you when, but not exactly what. But not both at the same time. If I say both at the same
time, then I will probably be wrong 90% of the time. Or maybe not, in ten years. So, I think markets are one of the most
important applications for blockchains. Blockchains create free and
open markets for participants. Blockchains operate using markets
to calibrate various parameters. For example, there is a market for mining,
for difficulty calculation, and proof of work. There is a dynamic marketplace where people make
decisions every day as to whether they’re profitable. There are currency markets within the crypto space.
All these markets exist because of blockchains. Markets are a critical application of blockchains.
Blockchain technology will create better, more fair, transparent, and open markets,
wherever markets are needed. That also means in places where
markets are needed but not wanted. For example, some of you will be shocked and
upset [about this], but… drug markets. Why? Because drug markets are still markets.
Markets require two things to happen. Does anybody know what they are?
This is the quiz part of the presentation. Anyone? [AUDIENCE] Supply and demand. [ANDREAS]
If you have supply and demand, a market happens. It doesn’t matter if you try to stop it.
A market will happen. I could create a market right now. I could create a supply, and someone
in this room will demand what I supply. Markets just emerge from human behaviour, around
things we want, including things we shouldn’t want, like drugs, cigarettes, ice cream, and
whatever other things that are bad for us. Blockchains will create open, free, transparent,
public, borderless, censorship-resistant markets, even if people want to stop them. Unstoppable markets will cause some problems. Stock markets will eventually operate on
blockchains. Not all of them, not forever. How many people have used
a fax machine in the last year? Let me rephrase that question:
how many of you are Americans? Yeah, I still had to use a fax machine last year.
We are 35 years into the development of the internet. Fax machines still exist. I had to use a fax machine.
Okay, I am lying. I didn’t use a fax machine. I used a website where I uploaded
a PDF and they had a fax machine, which sent it to the government agency
still insisting on receiving via fax machine. I basically cheated and sent them
a PDF over somebody else’s fax. The point is, no technology ever goes away. I’m sure there is a park somewhere in Seoul
where you can go for a ride on a horse. Not because you need to,
but because they are still around. So the idea that all stock markets will
be on blockchains, that is ridiculous. There will be other types of stock markets that still exist.
We will still have the markets that you see today. We will still have tuna markets like the ones in Japan,
where people bid on the floor for an actual tuna. These will still exist.
Tradition is powerful. I think stock markets are one of the applications
that we will see sooner rather than later. The reason is very simple. If you can make
a more efficient, transparent, open, public system, then a lot of money can be made with that.
It is likely, but it is not very interesting to me. It is not the most interesting application of blockchains. “What do you think about blockchains
for transport and logistics projects, such as those being done by IBM and Maersk.” “Can we track items from the farm to the supermarket?
What are the challenges that need to be addressed?” “What is the point of using permissioned
ledgers or blockchains for supply chains?” We have heard [this idea] a lot. There are [many]
companies, mostly IBM, leading in terms of… using permissioned ledgers for logistics.
I am not convinced yet of the viability of these projects. I’m not quite sure what problems they’re solving. I think the primary idea is to use a blockchain for a
common standard and implementation of a database… between multiple partners with competing interests. In that case, it will be a permissioned, federated
ledger with various participants in the supply chain. They will control the security and validation of entries.
Is this better than a centralized facility? It [would be] less centralized, but with a federated
group of people, it doesn’t provide immutability… in the way that an open public blockchain would. There would still be the fundamental problem of
reconciling the real world versus [blockchain entries]. Cryptocurrencies are secure because the [rules around]
bitcoin, ether, and Monero are controlled by consensus. But if somebody writes an entry to the blockchain
which says, “One hundred eggs left the factory today,” there is no way for the consensus rules
to validate [that event independently]. The eggs don’t live on the blockchain.
You must trust the party that enters the information, or the series of other parties who are
auditing or validating that information. It is a very different security model
than that of an open public blockchain. Does that have a good application?
Does it solve real problems in the logistics industry? Does it create a novel and disruptive use case
for blockchains, better than a replicated database… with digital signatures? It is hard to answer those questions. We will see
whether IBM is able to make something [useful]. I am very skeptical of these types of projects.
So far, they haven’t delivered much in terms of utility. “What do you think about using blockchains
for securing vehicular communications?” [In response to] any of these [questions] like
“what do you think of using blockchains for ‘x’,” I would ask: “Which components of an
open, public blockchain are you using?” Remember, a blockchain is not just PKI.
This is not digital signature technology. We already have PKI and digital signatures.
[Open] blockchains give you decentralization. Open, public, neutral, borderless
and censorship-resistant operation… between parties that don’t trust each other. Which part of that solves a problem
in vehicular communication? Maybe by providing an open, public standard
that is independent of the car company. But the only benefit [would be] if it [used] a completely
open, public ledger and not a private ledger. If it is a private ledger, then it is difficult
to see what benefits would accrue from… using such a complicated technology
for vehicular communication. There are easier and simpler
cryptographic technologies: digital signatures, message integrity with hashes. These are all parts of a blockchain,
but they are not a blockchain. They were invented long before blockchains.
Not everything needs a blockchain. Just because something can use digital signatures,
hashes, or PKI, doesn’t mean it needs a blockchain. “Are there any convincing examples of
identity management on blockchains?” Identity is a very difficult [issue].
The problems [are bigger] than blockchains. Again, I’m not convinced that there are solutions
today for identity [management] on blockchains. The primary problem is, it is hard to verify someone’s
identity using consensus algorithms in a blockchain. Someone must verify the truth
before an identity is recorded. Depending on how you do that, you will
end up with either a centralized database… that maybe exists under an open standard,
but is controlled in terms of what data is entered. This is the same problem with oracles and Ethereum.
How do you trust this third party, the oracle of identity? [How do you know] the information is correct?
If you compromise the mechanism by which… the information is recorded, then the blockchain
will potentially record a lie that cannot be validated… by consensus, but is still propagated to
everyone who uses it. You haven’t gained anything. Again, which part of open, public, neutral, borderless,
censorship-resistant, and decentralized do you gain… by adding a blockchain to the
problem of identity management? It is not a database for PKI. There are some potential advantages with permissioned
ledgers if it is created under a common standard… that requires less trust in a centralized party,
based on the collaboration of federated members. A consortium of competitors within an industry.
There may be some advantages to that. An open-source standard mechanism for coordinating
information across parties that don’t trust each other. But again, trying to control, assess, or verify things
which are not [intrinsic] to the blockchain, such as people, products, logistics, or votes, will still
require that you trust whoever [enters] that information. This simply moves the [source of the] problem. You need some reason for doing that.
Otherwise, you will end up with a database… [that looks nice on] a pitchdeck for venture capitalists,
but doesn’t have any real use case. [AUDIENCE] My name is Jeff McDonald.
I’m the co-founder of the NEM Foundation. First of all, I sincerely want to thank you for
your service in making the world a better place. You have done a great job. [ANDREAS] I have
never heard someone call it a “service,” but okay. [Laughter] [AUDIENCE] It is.
[ANDREAS] I really appreciate it, thank you. [AUDIENCE] Do you believe that Bitcoin [can] someday
offer an open, borderless, censorship-resistant… form of identity? [ANDREAS] Identity is very tricky. It isn’t a singular thing.
Identity is a fractal. It is multi-dimensional. Depending on which aspect of human identity you
look at, you may be able to separate one aspect of it, but then if you look very closely, you will notice
that is actually composed of [more] smaller aspects. If you look at each one, they are also composed
of smaller [aspects], all the way down to infinity. Identity is a very tricky thing. It is not even a thing.
It is a social construct, very much like money, but with [even more] very important nuances. I think it is premature to be [developing] with identity
right now, and the projects which are attempting to… deal with identity [management] today are dangerous. They are dangerous because they are applying
a very mechanistic view of human behaviour. “If you do ‘X’ today, then you are
66.3% likely to do ‘Y’ tomorrow.” When you apply that [metric] to humans, it creates
a very difficult situation. The machines never forget. Every mistake you make will
follow you for the rest of your life. What did you do, that was really stupid, before you
were fifteen years old? [AUDIENCE] Everything. [ANDREAS] Everything. So did I, right? I do not want to even think about
what I did before I was fifteen years old, many of which could have landed me in jail. Certainly, I would want none
of it to be read at my funeral. “On December 3rd 1986, look what Andreas did.”
[Laughter] No! Please, forget about that. Forgetting is a very important part of the social
compact. Without forgetting, there is no forgiveness. Without forgiveness, you remove the possibility
for change. If you assume people don’t change, then you create an environment where
people can’t change. We have a word for that. It is called fascism. I don’t want that kind
of situation. Machine identity is dangerous. It is not unsolvable or impossible to find useful
applications, but we must be very careful. I am a strong believer that as technologists, we have
an obligation to study, understand, and consider ethics.. in the work we do. Without ethics, technology is dangerous. When it comes to identity or any other social construct,
which is about human behavior, ethics is important. These technologies will, if you allow them,
start changing the way society works… and have deep reaching implications. There are some experiments [happening] in China
with Sesame, which is a social credit score system. According to recent reports, it has now prevented
26 million people from using public transportation… because their political and social behavior
earned them a low score and they’ve been banned. That is dangerous, right? It is only [becoming] worse.
Those are the kinds of things we don’t want to do. “How can you see blockchains being
used for voting and election processes?” It is still very early for implementations of voting
or election processes using blockchain technology. There are a number of reasons why it is too early.
This technology is early-stage and poorly understood. Not many people have access to this technology. In developed economies, it is already a problem
that voting is done with electronic machines. Most people do not know how to operate them,
so they can’t be audited properly, which creates more opportunities for
voting fraud and lost votes within the election. We see this happening quite a lot,
even in developed economies. In order to use blockchains for voting and election
processes, you must think about a couple of things. First of all, would a blockchain
used for elections be decentralized? If it is not decentralized, what are the benefits
of using a blockchain versus a simple database? One of the advantages is perhaps the ability
to inspect and audit digital signatures, but that would also make it very difficult to
maintain the privacy and anonymity of your vote. If it is decentralized, which platform do you run it on? Which government would trust a decentralized
blockchain to run their election [processes] on? Hard to see [the answer] at this point in time. Finally, the general idea of moving
voting to digital systems has problems, especially in countries where not many
people have access this technology. I think we will see voting and election processes
[on blockchains] that are used for small organizations, not national elections. Perhaps a shareholder election within a company,
voting on governance decisions in… a decentralized autonomous organization (DAO),
with smart contract voting on a specific topic. We will see that happening more and more. Gradually, those types of voting techniques will be
used for larger and larger organizational entities… as people build confidence
that the system can be neutral. [They must trust] that it will deliver [the same]
benefits of the traditional voting systems. Maybe sometime in the very far future, we will see
that being implemented for national elections. “[There are] a lot of tokens, most of which are
just white papers and not yet fully executed.” “When do you think these ideas can
reach the implementation stage?” Many of them will never reach the implementation
stage. We must start thinking differently. In this new world where everyone can create a token,
we must think differently about value and use cases. It is like asking, “Everybody has a blog. When will
these blogs start [producing] serious journalism?” Well, not all blogs will do serious journalism. Just because everybody can create
a blog, many of them will still be bad. Just because everybody can create a token,
many of them will just be whitepapers, nothing more. Don’t expect any of them to develop
[beyond] the original idea. Ideas are cheap. People who understand the space and are thinking
about these problems can generate ten ideas per day. The real catch comes when they try to execute on these
ideas and turn them into real products or services. That is where things become more difficult. If somebody comes to you and says, “I have a fantastic
idea for something new in the blockchain space,” that idea and these white papers are worth nothing
without an implementation behind them. That is okay. You must accept [that] in this
space, where anybody can create cheap [ideas], then you don’t [always] need
to consider [their idea] seriously.