CryptoKitties, Price irrationality and scaling

Cryptokitties mother and board (owned by an early crypto-kitty enthusiast and collector)


CryptoKitties is the first game created on the Blockchain.  The game works when individuals purchase a digital cat by spending ethereum. Individual cat owners can combine one or more cats to produce a third cat that shares physical features of  it’s parent cats. Additionally, there is a marketplace where one cat owner could trade in digital cats. During the initial stages of the game, each digital cat, that released onto the network sold for 1$ worth of ethereum.

Irrationality in Prices of Cats.
Over a short period of time (past 2 weeks) the marketplace exploded with millions of dollars worth kitties trading hands. More recently, cats that were sold during the early stages of the game were being sold at prices more than 100000 USD. The processes of breeding, selling, and buying were all configured as smart contracts wherein trade was autimated when suitable buyers and sellers were matched. An exponential increase in trade of kitties was responsible for almost 15% of the most recent transactions (1500 blocks) on the ethereum network as per the ETH Gas station.

This backlog of transactions resulted in excessive delays on the Ethereum network, even holding up some ICO offers.

Price Irrationality

Off late,  crypto coins (bitcoin, litecoin, ethereum, etc..) and crypto assets (e.g. crypto kitties, ICOs, etc.) have seen extreme price irrationality combined with volatility. This has caused significant press and attention to these assets. In this case of crypto kitties, linking the ownership of a digital asset (crypto kittie) through a smart contract and allowing this contract to reside permanently on a globally distributed Blockchain is the main feature.  The same functionality can be accomplished for other types of assets, e.g. titles of cars, titles of houses, or titles of land, birth certificates of individuals, etc… and is much more potent at changing real-world product markets. With crypto kitties, price irrationalities are potentially related to the scarcity of supply, since, the rate of production of new kitties is significantly lower than the demand from blockchain enthusiasts.


That being said, these technology demonstrator distributed apps and their problems have larger engineering and economic impacts.

  1. The large adoptions of the dapps surfaces infrastructure and engineering problems related to scalability that cannot be unearthed through any form of distributed software testing.  In the short term, the core engineering teams will have to increase their scalability efforts so as to support scale. Just a single successful app, can hold up transactions for hours from other applications on the network at the current rate. If the original goal of becoming a world computer that can theoretically execute any type of contract, has to be achieved, this scalability is imperative.
  2. This explosion of dapps, increases monetary gains for all segments of the crypto-ecosystem e.g. dapp writers, traders, and,  other owners of ether. Such a sudden increase in valuations for a tradeable digital asset has physical limitations, and is going to be corrected – hopefully by market conditions and not by regulation.

Important Computation problems in the ethereum blockchain

Important Computation Problems

Having the ability to process a large number of transactions has become the key focus area for many blockchain implementations.

In the following talk Vitalik talks about 4 categories problems that most blockchains have to handle.

  1. Privacy – a public ledger verifies data, but many nodes also access the data. This compromises the privacy of the user.  Many solutions such as a mixer, etc… are being worked on at this point in time.
  2. Consensus safety – the possibility of reducing the power consumed. (according to some estimates proof of work consumes more power than 1/2 the national power consumption of several small countries). Additionally achieving consensus has problems that arise from privacy violation, etc.
  3. Smart Contract bugs – loss of ethereum every year because of bugs in existing code bases of smart contracts. Therefore the proposal of Formal Verification (possibly automated) mechanism could reduce instances of these problems.

Scalability – the most interesting and complex problem to solve

Scalability – support for a large number of transactions. Many off-chain solutions similar to the lightning network, Raiden, plasma are possible- but their throughput is limited because of the main blockchain’s ability to handle attacks, etc… Therefore, the current proposed solution is to use “sharding” as a mechanism to scale the main chain solution. Sharding has its origins in large scale distributed databases and has been used to reduce query times from large databases, by partitioning data based on rules.

The following talk by Vitalik discusses the basics of the sharding approach for the Ethereum Blockchain. This approach suggested, is a hybrid approach wherein some part of the validation is done off-chain. The talk discusses the ability to target VISA scale validation for smart – contracts (not just pay and receive).

The Battle for Scaling Transactions

transaction support
A total number of transactions supported by ethereum has surpassed the combined transaction volume of all other cryptocurrencies in the past 24 hours.

Over the past few weeks, support for large transaction volumes has been at the heart of all design and other discussions in the cryptocurrency ecosystem. Scaling has two key aspects to it. Firstly, the ability to support large number of transactions (count). This means the underlying network should authorize the exchange of value, by Byzantine fault-tolerance and double spending prevention. Secondly, the time to validate each transaction on a network of nodes that are globally distributed should be within reasonable limits to allow for real-time trade. Though a sub-second (millisecond) transaction time is ideal, for most real-world systems such as a POS a confirmation on the blockchain is not possible and may not be required. Payment systems, point of sale systems, etc. will need to use completely different mechanisms to validate transactions  in real time, be it for either the bitcoin or for ethereum.

Currently, bitcoin’s block confirmation time is around 8 minutes as per this chart. If the Segwit2X performance enhancement was rolled out, this time would have reduced to under a minute.

In fact, in the real world it is much more delayed because at least 6 confirmations are needed for confirming a new transaction.

Comparing this with Ethereum’s block confirmation time, Ethereum has been able to, despite the surge in volumes, keep the block confirmation times to under 0.5 minutes (or less than 30 seconds). This is because of their most recent network upgrade that was not contentious as the Segwit2x and had wholesome support from the ethereum community.

Eventually, in my opinion, the cryptocurrency battle, will in the short term be based on scaling. The payoffs for the platform that scales the fastest in terms of two parameters i.e. transaction time, and, transaction count (volume) will be the highest.

Blockchain Security

As people depend on storing personal data on interconnected networks, the ability to protect networks becomes very crucial. The increasing number of hacks on corporate networks unveils the vulnerabilities of the computer system. With hard copies being replaced by system databases, companies undesirably invite trouble through exploits.

The recent Equifax (one of the nation’s 3 major credit reporting agencies) hack, and, the attack on Anthem (one of the nation’s largest health insurers) a couple of years ago has brought to the fore, the risks of storing sensitive data of individuals. The biggest harm of storing personal information is that once this information is gained, others can use  this identity to access credit. A hacker who has an individual’s name, Social Security number (a popular target for hackers), birth date, address, email, employment information (including income data) can now become that person.This information can be used for all legal purposes, such as accessing credit, providing information to legal agencies, etc.  On the gray/black market, combinations of Social Security numbers, birth dates and names sell for more than even credit card numbers.

With complex thefts, complex solutions to protect data are being  invented using the blockchainblockchain security

Blockchain-based security platforms such as Guardtime, REMME and Obsidian are restructuring cybersecurity.

Blockchain eliminates security flaws by taking away the weakest link, the human factor. The impact remains big with demand of more control of identities. Blochain based credit validation can potentially replace existing database based verification.  One such company working on Blockchain based credit validation is Bloom.

Bloom is an end-to-end protocol on the blockchain for identity attestation, risk assessment and credit scoring, thus making access to credit services more secure.  In addition, Bloom allows both traditional and digital currency lenders to serve people who cannot obtain a bank account or credit score. Bloom aims to address the “High Risk of Identity Theft” where borrowers are bound to expose all their personal info when applying for a loan. The attacker uses this same info to open new lines of credit.

The Bloom Protocol provides solutions allowing any lender authorized by a borrower to safely and securely issue credit to that borrower. The protocol work by validating prior stored information about credit access requests, on a secure network of nodes. Once this is authenticated the system proceeds to validate credit. This approach of validating a creditor’s prior history on the Blockchain after authentication will potentially do away with the need to independently use credit monitoring agencies to do such checks.

Forks – Bitcoin vs. Ethereum

Bitcoin’s forking problem

Today’s fork created Bitcoin Gold. A couple of month’s back there was a fork that led to the creation of Bitcoin Cash. Again in November, there is going to be another fork to Segwit2x. Many more forks can happen down the road -each claiming to have a new variation of the Bitcoin blockchain. Afterall, bitcoin’s source code is open source, and so is the blockchain accessible to all participants. If a group of people decides to host nodes with a particular version of the BTC code software with a particular change in the code (or logic) and start mining their own blocks a new chain is created.

The largest challenge for participants(traders, nodes, exchanges, miners) in the crypto-ecosystem is to identify that one of the forks is the original Bitcoin. Recently a couple of early adopters (Bitcoin millionaires) started claiming that Bitcoin Cash was the original Bitcoin and that they would fight for the rebranding.

Needless to say, this looks like the wild west in finance, where many versions of the same chain can exist independently, without anyone to control or coordinate these changes. (Time series Price correlations between Bitcoin Cash and Bitcoin has been < 0.6 since the birth of Bitcoin Cash).

Every exchange that supports Bitcoin will at the time of the fork have to decide at the  time of the fork to either support or to not support the fork. This is tedious work, to ensure that wallets of all users now access the corresponding keys on different blockchains to reflect the right balances. For exchanges will millions of users like CoinBase, Kraken, etc. this is a huge software update challenge.

Ethereum’s Fork (No! Problem)

Compared to Bitcoin’s forking problem, Ethereum’s hard fork to the Byzantium release did not create either a new chain or a new token variant of Ethereum. The Ethereum ecosystem, though, closely resembling Bitcoin’s ecosystem rallies around Vitalik and the Ethereum Foundation’s proposed changes. This is a wonderful model for large scalable open source systems, and, is slowly but surely leading to a self-regulated environment wherein a core team decides the roadmap for scaling (or other features) to be rolled out. The entire community of nodes, miners, and developers rally behind the fork and adopt the fork almost instantaneously.

Scaling the network, for all participants i.e. nodes, exchanges, transaction endpoints such as payment nodes, etc. become significantly simpler with Ethereum’s fork.

Lessons from bans – Uber and Bitcoin

Bans on Technology

The last two weeks have seen 2 major bans that affect global products. Uber, was banned in London, and Cryptocurrency exchanges and ICOs were banned in Mainland China. These two bans had one thing in common.

Wrong Usage

The wrong usage of a complex but highly efficient transactional and distributed computing system caused problems for all stakeholders concerned i.e. users, governments, service providers, intermediaries, economy etc.. In the case of Bitcoins and ICOs, there were lots of fraudulent schemes. In the case of Uber, the background checks of drivers weren’t good enough and had led to multiple crimes in a large city. These problems if left unchecked would lead to far more negative consequences for these institutions and for the governments that promise to protect citizens and their assets.


Bitcoin – mostly represented by the Bitcoin Foundation, didn’t make a public all-out effort to negotiate the ban. However,  Ethereum had some representation in China, to possibly talk with authorities. In this void created by a national ban in China, has risen a new crypto-platform – that of NEO. NEO is called the Chinese Ethereum because of its functional similarity to Ethereum’s Turing Machine.

Whereas Uber, and its new CEO led an amazing response, firstly by apologizing to the Londoners (both the citizens and the administration). On one hand, they mentioned that as a company they would introspect about their behavior, while on the other hand, they would appeal the ban in the courts of London.

Mature Response vs. Immature Response

This response from a mature and level-headed management team demonstrates a few things: Firstly, centralization of authority and power, within the structure of a modern firm is a very effective at running a decentralized and distributed technical system. Secondly, it might be difficult to overturn a ban legally, however, as ecosystems mature and the good users of a system far outnumber the bad users, law would potentially side the good. The possibility of lifting the ban increases proportionally with the corrective actions taken by the affected firm.  With the crypto-ecosystem – where there is no clear governance structure, it will be difficult to find mature representation and convince existing institutions about the need to overturn a ban. A few questions arise:

*How can ICO’s who have very little structure of control return the money raised from people on some ICO exchange?

*Who will self-regulate and mandate reparations caused by these fraudulent ICOs?

*Ethereum as a protocol is developing, but will Ethereum support a reputation model to support only valid ICOs, now that ICOs are by far the most popular Ethereum application in the market.

If a  structure of control is absent, maybe an ecosystem driven way to moderate applications of the decentralized technology is needed. If moderation of applications, users and participants do not happen – then the whole ecosystem suffers.

Technology Maturity and the Blockchain’s stability


Markets are rife with price speculations about Bitcoin and Ethereum. The last few days saw a unique phenomenon in this market – one where all miners, exchanges, and developers came together to accomplish Segwit2X roll out onto the Bitcoin Blockchain. Unlike Ethereum, which has a loveable and mature leader, Bitcoin’s development team has changed over the years. As the value has increased, more and more power has gone onto the miners of Bitcoin. In fact, this parody website Bitcoin Obituary that has recorded the death of Bitcoin 140 times – as reported by mainstream media i.e. Forbes, Yahoo! news, CNBC, NYtimes, WEF, etc..  Here’s the most recent obituary.

Complex Technology Maturity

Any new (complex) technology that depends on its network adoption, should have two concurrent mathematical time series functions expressing their adoptions.

The first time and most prominent one is the diffusion curve. This curve represents the value the technology provides to its users. There has been enough debate and literature around diffusion of innovation and this curve.

The second – and – most often ignored time series function is that of technology maturity. It takes many years for a technology to mature to operate behind the scenes, and, to become so ubiquitous that problems with the technology (or bugs in code) do not affect its users.

For example, consider your desktop computer or mobile phone. The TCP/IP network layer works mostly seamlessly without the need for any restarts. nearly flawless (or mature) TCP/IP is what makes it possible for anyone to depend on the email or browser or even stream videos or play games. The complex piece of software splits each packet of data into chunks and sends it over a network. On the receiving side, the same complex piece of code reassembles these packets and surfaces it to the application reading it.

Linux TCP/IP

Linux has been around for about 27 years, and the TCP/IP protocol that formed the backbone of the internet has been around for more than 40 years. One would expect that technology which powers 80% of all internet traffic (mobile and desktop based), and millions of apps should have zero bugs in it. The fact that TCP/IP itself is older does not help. Despite, its age, and the number of developers, apps, and installations TCP/IP on linux still has several bugs. The list of all TCP/IP bugs can be seen from Bugzilla. As recent as 2016, there have been bug fixes for TCP/IP code that could potentially cause a system to function.

Perspective for the Blockchain

To put this in perspective with the Blockchain’s technological maturity – society is at the very early stages of Bitcoin (about 7 years old) and Ethereum(just over 3 years old). The adoption curve is still at the very early stages (less than 0.1% of the world’s population own it), and, so is the technology maturity curve (the scale of transactions and the possibility of applications on the Blockchain is less than 0.1% of all monetary transactions therein). This volatility is expected for the next several years atleast.

The only difference this time is that this technology is fungible and both the adoption curve and technological maturity curve affect its value at any given point in time.

Bitcoin’s Social Scalability

Social Scalability
This concept of “social scalability” and Bitcoin’s (the Blockchain’s) role in enabling social scalability is fascinating. Nick Szabo – the original creator of the smart contract – defines it as follows: Social scalability is the ability of an institution –- a relationship or shared endeavor, in which multiple people repeatedly participate, and featuring customs, rules, or other features which constrain or motivate participants’ behaviors — to overcome shortcomings in human minds and in the motivating or constraining aspects of said institution that limit who or how many can successfully participate.

Why Social Scalability matters?
What’s amazing about this idea is that social scalability (as evidenced and plausible with a blockchain based system) can eliminate dunbar’s limitation of 150.Repeated conversations featuring customs, rules and/or other features that eliminate constraints (both internal and external) are now possible. In the essay here Nick makes the case for social scalability and how blockchains play the empowering role in ensuring that human and social/behavioral limitations can be transcended. This transcending happens because of various limitations that are completely eliminated in such an institution powered by the blockchain. There is more about this here – where Naval Ravikanth, Tim Ferriss and Nick Szabo engage in an all enlightening podcast here.

Nick Szabo a polymath genius, designed Bitgold – using Hashcash – the earliest proof of work system, now common-place in the blockchain world for transaction validation. His thoughts on various things are phenomenal to read here .

Corporates to startups – Part 1.

Many of my friends, colleagues and students want to either

a) start a high growth company OR

b)  work for a startup

Many of them have atleast a decade and more experience in working for large companies.

Some of them come to me with term sheets. Some others come with revenue models or with what looks like a draft copy of equity contracts that read like a pathway to retirement (and a life on a yatch) in 2 years.

Being an early stage employee has its benefits and pitfalls. In the next few series of articles I’ll lay out guidelines on what to look out for – things that I’ve learnt from my peers and friends.

1.What factors to look out for?
2.How do one choose the right startup?
3.What does one do with the term sheet or equity contract?

Bitcoin scaling agreements – Consensus 2017

Consensus 2017 concluded this past week in New York. Off the many excellent talks like the one here from Dr. Arvind Krishna at IBM or the one from  Stefan Thomas, Ripple’s CTO here Consensus 2017 marked a watershed moment in the history of computing per se – in terms of participation from the industry, from the research community and the overall crypto ecosystem. Many agreements were reached – and development paths for many altcoins such as LiteCoin, Ripple, etc.. have been put forth in front of the public.

Off the many trade agreements that were reached, there was one that was notable. The Bitcoin scaling agreement amongst the miners, developers and hosting partners of Bitcoin. The Segregated Witness proposal rollout agreement was reached – as stated here in this Medium post .  This is going to be the first consensus based hard fork for Bitcoin. The debate around BTU (Bitcoin Unlimited) versus SegWit2MB has with this (hopefully!) been settled. As of now 83.5% of the world’s hashing power – has agreed to this hard fork.

What we’re seeing now is really very large delays in transaction confirmation. For example, most exchanges or places that accept Bitcoin such as miners are forced to alter policies on payment using BTC because of transaction delays. The transaction delays increase costs for all parties involved in a bitcoin based transaction – thus NOT living upto Satoshi’s original vision for Bitcoin. The following is an excerpt from his seminal paper…

“Commerce on the Internet has come to rely almost exclusively on financial institutions serving as trusted third parties to process electronic payments. While the system works well enough for most transactions, it still suffers from the inherent weaknesses of the trust based model. Completely non-reversible transactions are not really possible, since financial institutions cannot avoid mediating disputes. The cost of mediation increases transaction costs, limiting the minimum practical transaction size and cutting off the possibility for small casual transactions, and there is a broader cost in the loss of ability to make non-reversible payments for nonreversible services. With the possibility of reversal, the need for trust spreads. Merchants must be wary of their customers, hassling them for more information than they would otherwise need. A certain percentage of fraud is accepted as unavoidable. These costs and payment uncertainties can be avoided in person by using physical currency, but no mechanism exists to make payments over a communications channel without a trusted party.”

With SegWit2Mb, once the hard fork is rolled out we will see Bitcoin scale to unlimited number of transactions at a much larger pace than ever before. This was one of the major problems with Bitcoin – documented in one of my research papers which I will share – a presentation off shortly…

This was one of the major barriers for adoption of Bitcoin that we documented in one of our research papers, that I presented recently at POMS 2017.