Blockchain 2.

0
Opportunities and Risks
Patrick Valduriez
The Hype

2
Bitcoin
•  Bitcoin: A Peer-to-Peer Electronic Cash System
•  Satoshi Nakamoto (pseudo), Oct. 31, 2008 (Halloween)
•  Cryptocurrency and payment system
•  Blockchain is the infrastructure
•  Since then
•  Many blockchains: Etherum in 2013, Ripple in 2014, etc.
•  Increasing use for high-risk investment
•  Initial Coin Offerings
•  But also in fraudulent or illegal activities !
•  Scam, purchase on the dark web, money laundering, tax
evasion, …
•  Warnings from market authorities and beginning of
regulation (China, South Korea, Japan, EU, …)
3
The Currency of Tomorrow?
•  Pros
•  Low transaction fee (set by the sender to speed up processing)
•  Fewer risks for merchants (no fraudulent chargebacks)
•  Security and control (protection from identity theft)
•  Trust through the blockchain
•  Cons
•  Unstable: no backing by a state or fed bank (unlike $ and €)
•  Unrelated to real economy, e.g. GPD: fosters speculation
•  High volatility, e.g. between 6K and 7K$ in 3 hours
•  Small user base: 20 million bitcoin wallets
•  Versus billions of users of e-payment systems like AliPay and Paypal
•  The Crypto Bubble (2017)*
•  Bitcoin price increased from $1k to 10K, then peaked almost at
$20K in December 2017 to collapse 4 months later to below $6k
(down 70% from the peak), and close to $6k since then
* Testimony for the Hearing of the US Senate Committee on Banking, Housing and Community
Affairs On “Exploring the Cryptocurrency and Blockchain Ecosystem”. Nouriel Roubini (NYU),
october 2018.
4
 Outline

•  Trust with blockchain

•  Consensus protocols
•  How the blockchain works
•  Blockchain 2.0
•  Use cases
•  Opportunities and risks
•  Issues
Trust in a Modern Economy

•  Context
•  How to exchange assets
safely between two
parties?
•  Centralized ledger
•  An account book that
records all transactions
•  Controlled by a trusted
central authority
•  E.g. a clearing house

6
Problems with Central Authority

•  Single point of failure

•  And easy target for attackers
•  Favors concentration of actors
•  Banks
•  Exploit our money to make big money
•  Web giants (GAFAM) and other intermediaries (Uber,
etc.)
•  Exploit our data to make big money

7
Trust with Blockchain

•  A distributed ledger
•  Shared by all participants
•  Replicated
•  Decentralized
•  Append-only
•  No update, no delete
•  Distributed transaction
validation
•  Consensus
•  Unfalsiable, verifiable

8
Blockchain Promises

•  Increased trust in value exchange

•  Trust the data, not the participants
•  No single point of failure
•  Increased security
•  Efficient, consistent transactions between
participants
•  Faster and cheaper than relying on a long chain of
intermediaries, with incompatible systems and rules

9
Public versus Private Blockchain
•  Public blockchain
•  Open P2P network
•  Participants can join and leave
without notification
•  Anonymous, untrusted participants
•  Large-scale distributed ledger

•  Private blockchain
•  Closed permissioned network
•  Identified, trusted participants
•  Regulated control
•  Small to medium-scale distributed
ledger
10
Background on Consensus
Protocols
Consensus
•  Critical applications
•  Replication, transaction validation, identity verification, etc.
•  Major problem of distributed systems
•  How to reach a consensus, i.e. agree on the same value, in the
presence of a number of faulty processes?
•  Problem statement: given n processes and one leader, how
to reach:
•  Agreement : all correct processes agree on the same value
•  Validity: if initiator does not fail, all correct processes agree on its
value
•  Types of failures
•  Crash: the easy case
•  Malicious (also called Byzantine)
•  The process gives different values to different observers
•  FLP (Fischer, Lynch, Paterson) impossibility result
•  With only one crash failure, termination is not guaranteed
•  Example: coordinator failure in 2PC
12
The Byzantine Agreement Problem

•  Suppose an army of the Byzantine Empire

•  Generals can only communicate by messengers and
must establish a common plan to attack the enemy or
retreat
•  A number of these generals may be traitors and vote
selectively
•  Example with 5 generals: 2 support the attack and 2
are in favor of retreat; the 5th can send an attack
vote to the first two and a retreat vote to the other
two and then …
•  Problem formulation
•  Find an algorithm (consensus) to ensure that loyal
generals can agree on a common battle plan

13
Paxos Algorithm
•  The basis for a family of protocols
•  [Lamport 1999, ACM Turing Award 2013]
•  Used to manage large-scale distributed data
•  Google Spanner & Megastore
•  IBM SAN Volume Controller
•  Microsoft Autopilot Cluster Mgr
•  Ceph (distributed file system)
•  Neo4J (NoSQL graph DBMS)
•  Inspired by the functioning of the Parliament of the
Paxos Island
•  The Parliament did work, despite the regular absence of
legislators and messages loss
14
Paxos Algorithm

•  Principle (simplified)
•  Initialization: a leader is elected by a majority quorum
•  Replication: leader replicates new updates to the majority
quorum
•  Leader failure: il the leader fails, a new leader is elected
•  To make progress, at least 1/2 of the participants should be
alive

•  Limitations
•  Permissioned settings: all participants should be known a
priori
•  Not appropriate for public blockchain
•  Tolerates only crash failures
•  Does not deal with malicious nodes
•  Progress is not guaranteed (FLP impossibility)

15
Practical Byzantine Fault Tolerance
(PBFT)
•  A three-phase protocol [Castro & Liskov 1999]
1.  Pre-prepare: a leader broadcasts a value to be
committed by other nodes
2.  Prepare: the nodes broadcast the values they are about
to commit
3.  Commit: confirms the committed value when more than
2/3 of the nodes agree in the previous phase
•  Assessment
•  Tolerates Byzantine failures
•  Permissioned settings

16
How the Blockchain Works
Blockchain Concepts
•  Blockchain
•  An immutable distributed database, i.e. a log of blocks,
which are linked and replicated on full nodes
•  A block
•  Digital container for transactions, contracts, property titles,
etc.
•  Transactions are secured using public key encryption
•  The code of each new block is built on that of the
preceding block
•  Guarantees that it cannot be changed or tampered
•  The blockchain is viewed by all participants
•  Enables validating the entries in the blocks
•  Privacy: users are pseudonomyzed

18
Blockchain Protocol (Nakamoto 2008)
0. Initialization (of a full node)
•  Synchronization with the network to obtain the
blockchain (185 GB on Q3, 2018)
1.  Two users agree on a transaction
•  Information exchange: wallet addresses, public keys, …
2.  Grouping with other transactions in a block and
validation of the block (and of the transactions)
•  Consensus using "mining"
3.  Addition of the validated block in the blockchain
and replication in the P2P network
4.  Transaction confirmation

19
 Transaction

Owner1 Owner2 Owner3

•  The coin owner signs the Transaction Transaction Transaction
transaction by
1.  Creating a hash value of PK2 Pk3 PK4
•  The previous transaction
•  And the public key (PK) h h h
of the next owner
2.  Signing it with its secret
key (SK) H-val signed H-val signed H-val signed
with SK1 with SK2 with SK3

20
Block Management
•  Transactions are placed into blocks, validated
(by checking inputs/outputs, etc.) and linked by
their addresses
•  Size of a bitcoin block = 1 Megabyte

Result of mining

Block Block

H-value Nonce H-value Nonce

T T
… T T T
… T

21
Validation by the Network

•  Each block is validated by network nodes, the miners,

by a consensus protocol (see next)
•  Problem: accidental fork
•  As different blocks are validated in parallel, one node can see
several candidate chains at any time
•  Solution: longest chain rule

Block 5 Block 6a Block 7a Block 8a

Block 6b

Transactions in a validated block are provisionally

validated; confirmation must be awaited
22
Intentional Fork
•  Main reasons
•  To add new features to the blockchain (protocol changes) => new
software
•  To reverse the effects of hacking or catastrophic bugs
•  Soft versus hard fork
•  Soft fork: backward compatible
•  The old software recognizes blocks created with new rules as
valid
•  Makes it easy for attackers
•  Hard fork
•  The old software recognizes blocks created with new rules as
invalid
•  Example: the battle between (new) Ethereum and Ethereum
Classic
•  In 2016, after an attack against the Decentralized Autonomous
Organization (DAO), a complex smart contract for venture capital, the
blockchain forked but without momentum
•  Battle is more philosophical and ethical than technical

23
Consensus Protocol: mining
•  Why not Paxos?
•  Remember: participants are unknown
•  To validate a block, miner nodes compete (as in a
lottery) to produce a nonce (number used once)
•  One of the first competing solutions is selected, e.g. the
one that includes the largest number of transactions
•  The winner miner is paid, e.g. 12.5 bitcoins today
(originally 50)
•  This increases the money supply
•  Mining is designed to be difficult
•  The more mining power the network has, the harder it is
to compute the nonce
•  This allows controlling the injection of new blocks
("inflation") in the system, on avg. 1 block every 10mn
•  Advantages powerful nodes
24
Mining Difficulty : Proof of Work (PoW)
•  PoW
•  A piece of data that is difficult to calculate but easy to
verify
•  First proposed to prevent DoS attacks
•  Hashcash PoW
•  Computed by each miner to produce the nonce
•  Goal: produce a value v such that h(v)<T where
•  h is a hash function (SHA-256)
•  T is a target value which is shared by all nodes and
reflects the size of the network
•  v is a 256-bit number starting with n zero bits
•  Low probability of success : 1/2n
25
The 51% Attack
•  Also called Goldfinger attack
•  Enables the attacker to invalidate valid transactions
and double spend funds
•  How
•  By holding more than 50% of the total computing
power for mining
•  Miners coalition
•  It then becomes possible to modify a received chain
(e.g. by removing a transaction) and produce a longer
chain that will be selected by the majority
•  Solution: monitoring by the community
•  In January 2014, Ghash.io reached 42%, then dropped
to 9% after the Bitcoin community alert

26
Transaction Confirmation

•  A provisionally validated transaction in a candidate

block ensures that it has been verified and is viable
•  Each new block accepted in the chain after the
validation of the transaction is considered as a
confirmation
•  A transaction is considered mature after 6 confirmations
(1 hour on average)
•  New bitcoins (mining products) are only valid after 120
confirmations, to avoid the 51% attack

27
Public Blockchain Limitations
•  Complexity and low scalability
•  Difficult evolution of operating rules
•  Increasing chain size
•  Low number of transactions per second (TPS)
•  5-7 TPS for Bitcoin versus 25K TPS for VISA
•  Unpredictable duration of transactions, from minutes to days
•  Cost
•  High energy consumption
•  Favors concentration of miners
•  Users are pseudonymized, not anonymized
•  Making a transaction with a user reveals all its other
transactions
•  Lack of control and regulation
•  Hard for states to watch and tax transactions
28
Blockchain 2.0
Evolution of Paradigm

30
Evolution of Paradigm

•  Beyond Bitcoin and other cryptocurrencies

•  Recording and exchange of assets without powerful
intermediaries
•  Example: smart contracts
•  Positioning in the internet
•  TCP/IP: the communication protocol
•  Blockchain: the value-exchange protocol?

31
Blockchain 2.0 Technology
•  Programmable blockchain, e.g. Etherum
•  Allows application developpers to build APIs on the
Blockchain protocol
•  APIs to allocate digital resources (bandwidth, storage,
etc.) to the connected devices, e.g. FileCoin
•  Micropayment APIs tailored to the type of transaction
(e.g. tipping a blog versus tipping a car share driver)
•  Private blockchain
•  Efficient transaction validation since participants are
trusted
•  No need to produce a PoW
•  Efficient management, e.g. in the cloud

32
Blockchain 2.0 Development
•  Support from all major industry players
•  Finance services: Mastercard, VISA, …
•  Audit firms: EY, KPMG, PwC, Deloitte
•  Consulting firms: Accenture, Capgemini,
•  Web giants: Amazon, Google
•  Software suppliers: IBM, Oracle, Microsoft, SAP
•  Technology platform companies: Cisco, Fujitsu,
IBM, Intel, NEC, Red Hat, VMware
•  New blockchain ISVs
•  Blockchain, ConsenSys, Digital Asset, R3, Onchain

33
Smart Contracts

•  "Code is law", Lawrence Lessig, Harvard Law School

•  Smart contract (Nick Szabo, 1993)
•  Self-executing contract, with code that embeds the terms
and conditions of a contract
•  Early application: digital rights management schemes
•  Deployment in the blockchain 2.0 (e.g. Etherum)
•  Participants can be unknown to each other
•  Contracts can be with many third parties, e.g. IoT devices,
at low cost
•  Challenges
•  Bug-free code, which requires code certification
•  Compliance with mandatory regulation, which requires
collaboration between programmers and lawyers

34
Hyperledger Project (Linux Foundation)
•  Started in 2015 (IBM, Intel, Cisco, …)
•  Open source blockchains and related tools
•  Major frameworks
•  Hyperledger Fabric (IBM, digital Asset): a permissioned
blockchain infrastructure
•  Smart contracts, configurable consensus (PBFT, …)
and membership services
•  Sawtooth (Intel): a new consensus "Proof of Elapsed
Time" that builds on trusted execution environments
•  Hyperledger Iroha (Soramitsu): based on Hyperledger
Fabric, with a focus on mobile applications

35
Blockchain Use Cases
Blockchain 2.0 Apps
•  Critical characteristics of the applications
•  Asset and value are exchanged (transactions)
•  Multiple participants, unknown to each other
•  Trust is critical
•  Top use cases
•  Financial services, micropayments
•  Digital rights using smart contracts
•  Digital identity
•  Supply chain management
•  Internet of Things (IoT)
•  POCs in many industries
•  Publishing, retail, music, healthcare, rental, real estate,
government, energy, agriculture, etc.
37
Diamond Supply Chain Management

•  Problems
•  How to trace diamonds, in an era of “blood diamonds”?
•  Complex and multi-tiered diamond and jewelry supply
chain
•  Objective of TrustChain
•  Provide trusted products with documented authenticity,
guaranteeing quality and environmental responsibility
•  Solution (IBM Hyperledger)
•  A permissioned blockchain that establishes a single
shared view of information without compromising detail,
privacy, or confidentiality

38
Opportunities and Risks
Opportunities

•  Disruptive technology
•  For recording transactions and verifying records
•  The ability to program applications and business logic in
the blockchain opens up many possibilities for
developers
•  E.g. smart contracts
•  Disruptive power
•  The goal of cypherpunk activists
•  It may establish a sense of democracy and equality for
individuals and small businesses in countries with non-
transparent, unsecure jurisdictions

40
Risks

•  Market disruption
•  Massive disintermediation of the current system,
replacing all procedures that deal with transactions with
a system where participants trade directly
•  Public blockchain
•  Consumer protection: significant volatility of Bitcoin and
other cryptocurrencies (no government backup)
•  Increasing use for fraudulent or illegal activities
•  Security concerns: if a private key is lost or stolen, an
individual has no recourse
•  Lack of control and regulation, and hard for states to
agree on what to do
41
Research Issues

•  Scalability of the public blockchain

•  Alternatives to PoW : proof of stake, proof of hold, proof of
use, proof of stake/time, …
•  New generation blockchains, e.g. Bitcoin-NG [Usenix 2016]
•  Smart contracts
•  Code certification and verification
•  Blockchain interoperability
•  Blockchain Interoperability Alliance (BIA), to promote cross-
blockchain transactions
•  Blockchain and big data
•  Blockchain-generated data analysis, e.g. fraud prevention
based on real-time transactions
•  Blockchain-based DBMS, e.g. BigchainDB

42
Ethical Issues

•  Blockchain can have strong (good or bad)

impact on the world
•  People, society, economy, environment, …
•  Remember: the public blockchain is great for crooks
and criminals
•  This raises ethical issues that we cannot simply
ignore
•  See the recent panel: A Debate on Data and
Algorithmic Ethics (VLDB 2018)

43
About Trust Again

44