24 min read - Posted 20 May 19

random cryptocurrency / cryptoeconomics thoughts #4

The below is ungroomed, thinking out loud, a collection of thought bubbles written as a messy tangle and has not been processed for concision/to make it easier to read. Patience and indulgence is required. There is not a sequential flow, paragraphs are ordered randomly. No effort was made to format for easier reading/viewing.

2018 for crypto is often discussed in terms that echo, extend or reprise the memes and themes that inflected 2017. Having encountered those effects, causes and contours through the course of 2017, it seems counter-logical to expect that 2018 will be a faithful replay of those same effects in 2018 — why would crypto 2018 be limited to only those causes and effects that mutely repeat 2017 again; has crypto run out of new ideas? In Extremistan domains, it is normally an error to expect the future to be a faithful echo of the past. Crypto is an Extremistan world, dominated by unforeseeable causes that exhibit multi-sigma, fat tail Swans, whether White or Black. Chartists will refute this by showing you Bitcoin’s multi-year price chart with a log scale, as it shows a linear upward sloping line that offers evidence for continued telegraphable predictability. Chartists kinda want tomorrow to be an analogue of today.
Crypto will go either go exponential (J curve, hyperbola) again, else it will go to zero — Extremistan logic says it seems least likely that it will just passively follow a predictable log upward sloping line. Crypto will ingest all human social economic interaction, else we will deprecate it to near zero. The internet did the same: it started from zero, went J curve, then went J curve again — today, everything exists on the internet, all business is e-business, all humans exist in cyberspace. Barbell, winner takes all, White Swan, convexity.
This suggests that planet earth will either unleash new and previously unseen drama onto its crypto efforts in 2018 and 2019, else crypto will begin its fall to zero and will struggle all year. It falls to dreamers, futurists and investment analysts to model and extrapolate these possible future pathings. Or just hopeful crypto investors. For the anxious viewer, this quiet bifurcation will cause confusion: things will be quiet for a while, then large jumps/dislocations (earthquakes) will occur: the signal will be bursty, peaky rather than continuous, and most viewers will confuse periods of ‘sideways quiet’ with ‘crypto’s gone dead’.
Crypto needs to contemplate likely-inevitable “walls of worry” as 2018 unfolds, or in subsequent timeframes.

  • Governments, regulators and old world vested interests intensify efforts to “break the back” of crypto fever in citizenry; repel the invader else we will lose our fortress

  • Thefts, hacking, threat-follows-value: crypto is worth money, so thieves will flock to the easy lunch meal; and/or, insiders get corrupted by easy wealth and can’t resist “stealing the pooch”

  • Scaling

  • Privacy; once crypto solves scale, then investors will turn their minds to tax, audit, privacy and surveillance concerns

  • Governance; human tribes still have much to learn about how to arbitrate amongst competing interests

  • Trashtalking: as crypto increasingly becomes dominant, butt-hurt by old-world losing industries will escalate fiercely, and they will ask their serfs/alumni to also chime in to swell the chorus

  • Crypto does it learning-curve fumbles in front of your eyes, not behind closed doors: pundits will misinterpret that as failure of the whole premise, so noobs continue to panic at each fumble Trying to guess the pathing of the future also is important, as path dependency is a factor that inflects actual betting histories. It matters when each of the above occurs, in what sequence. This sensitivity effect compounds as margin and derivatives for crypto’s get applied.
    The above worry beads, whilst relevant and ascribable, do not really help us solidly predict the actual pathing in 2018. So far, we have seen the early January 2018 insanity peak, then the huge sell off wash off in following weeks/months. The rest of 2018 awaits us, the task is to lay out the footsteps and thus foretell the pathing, so as to attempt a predictive world history in the Feynman sense.

  • Investors allow thoughts of SEC, CFTC regulation cower them and thus cool demand for ICO’s. Insta-millionaire wanna-be founder hopefuls go into hibernation for a while, until one of their kindred figures out some other clever new way to bypass the traditional ‘venture cap’ circus and harvest the dumb money that eagerly wants to FOMO into new overhyped moon-shot shitcoin projects.

  • In-motion big bet projects — Tezos, EOS, Cardano, NEO — grind away. No big bangs, but no stumbles either. Soft launch as the way to avoid regulator wrath/scrutiny yet still monetise WiP.

  • Scaling candidates, state channel offloads, IBCs — Lightning, Plasma, Casper, Raiden, Truebit, MimbleWinble, DAGs, Cosmos, Polkadot — continue to gather bets. Each struggles to prove itself as “THE” answer to the BFT, CAP and/or FLP barriers, though.

  • First-world central bank dream teams (G20, IMF, etc) combine their resources and launch a concerted, full scale Alamo-style “kill all crypto” assault. All GitHub access to repo’s is banned/firewalled at national borders, VPNs are banned. Open Source encryption tools are banned. Some example citizens who poked holes in the walls get arrested, footage aired on tv to cower the remaining citizens.

  • Simultaneously, some example citizens in these countries get their tax affairs publicly audited, so as to define them as crypto tax cheats, to “send the message” to the remaining citizens.

  • Cryptocoin-centric terrorist act gets false flag unfurled on us via non-fake news media. Trump, Xi, Putin lecture us that all crypto is terrorist evil #1.

  • Cryptomarket shudders under the combined weight, the light goes dull. Only deepnet activity keeps crypto chains barely alive on life-support via massively reduced numbers of crypto tech insiders capable/willing to run TAILS, TOR/I2P, meshnets, deep underground. Bummer, dude. Is all hope lost? Similarly, Do-Gooders have been calling for years to nannify the internet, but so far it has proved antifragile against such censorship to its genetic evolutionary actions. Crypto will exemplify antifragility and convexity. (At least until the AI’s start writing and improving the code) cryptocurrency codebases are software. We can observe the past multi-decade history of the art, science and business of software, and consider to what degree those contours will inform and constrain the path of cryptocurrency software projects henceforth. The whole world used to trashtalk Microsoft for the boneheaded business/coding blunder choices Microsoft made in the 1990’s but eventually the codebase improved and now the whole world has moved on from Microsoft bashing and gotten on with business. Cryptocurrency codebases right now are a frontier town where banditos, cowpokes and townsfolk are all staring at each other wondering who has the best chance at improving things and leading things. Cryptocurrency codebases are being rewritten daily in front of our eyes, we observe the sausage as it is being made in the factory production line, and sometimes the townsfolk see unpleasant aspects of the sausage making art/business. To what degree this will mature and become more professionalised (better optics, less unsightly) remains to be seen, perhaps the art of writing cryptocurrency codebases will never allow itself to become a tame observance. Expect coding errors and governance failures. Crypto should never be tame, less it lose its fangs. Crypto is a controlled laboratory experiment that is deliberately designed and set up to pit individual incentivation (appetite, greed, self-interest) against external efficiency (maximised total welfare) + altruism i.e. social constructs. Cryptoeconomics are the dialog between the sovereign individual/entity — who cares only for its own needs — versus the need for the individual to also have an efficient ‘externalising system’ (a system that arbitrates the supply the required utility via an interface to the externalised world i.e. society, tribe, out-group). Inside our heads, we need/want things, then we use money to get the external world to give it to us. Cryptoeconomics is the dialog between those two aspects of the ego-self. Cryptoeconomics as locked down into coin codebases will twist and turn over time as that dialog unfolds and scales up, as the self-ego asks for, pleads, commands and hacks away against the attack surface of the external utility world and its ‘hard rule set’. Humans want their cake and they want it without cost or limit or obligation to themselves, and we will attempt continuously to reprogram crypto codebases into strange attempts to bypass previous constraints that, for instance, fiat currency fractional reserve banking systems included as part of their inherent nature. We may have been willing to accept that banks won’t give us easy loans for our desired doodads in the past, but now we might expect crypto to give us free coinz for nothing. CryptoInnerPerson won’t take ‘no’ for an answer and will fork off to a different coin that offers more hope for easy candy. We will hack our coinz and the corpus of cryptoeconomics until we exhaust its ability to either provide the magic miracle of endless wealth or else it collapses from what is essentially a (socially constructed) DDoS attack. CryptoInnerPerson is more so a moody teenager than a post-middle aged mature adult, and so we will attempt to force crypto into giving us easy wish fulfilment of our teenage need for being unaccountable. We all know how that script ends. “….. there is an entire sub-discipline of economics dedicated to studying how to design protocols that incentivize rational actors to behave in socially desirable ways. This is called mechanism design.”

    The aspect of mechanism design that needs to be highlighted for the cryptoeconomics context is that of anticensorship. A good cryptocoin is not so much just a 21st century go-go version of a dollar bill — more than that, as a primary aspect, a cryptocurrency is an external expression of the mental concept of anticensorship. Whilst mechanism design should inform and serve as an input ingredient to the edifice of cryptoeconomics, the end state that we need from our adventure into cryptocurrency is, above all, anticensorship. Utility function (quality of ‘mechanism design’) alone is not sufficient. Cryptocurrency is not just the race to design a better mousetrap: cryptocurrency is a monument we build to ourselves to show how much we value our own self-primacy and the censorship-free expression of that. (For our crypto purpose here), censorship is The State telling The Individual what’s to be what, or else. Cryptocurrency’s (antifragile) ability to avoid, delete, bypass and render harmless the (censoring) actions of The State is a sovereign utility value of crypto to us. Mechanism Design is a means to that end. Crypto is of course about solving the confluence challenges posed by incentives, consensus and Byzantine action. The question is what higher context gives those efforts their value and importance: the answer is the social and personal utility that blockchain and crypto gives us around money and wealth in a censorship-free manner. Who amongst us wants their wealth censorable (controlled, directed, taxed) by external actors? Anticensorship is the context that fuels the efforts now being expended writing crypto codebases as being better mechanisms than fiat currency and banking.
    Mechanism design, like much of the canon of economics, presumes actors are rational, will act along a rational basis. Crypto kills (or recasts) rationality in that sense, as anticensorship is the higher goal and something’s gotta give. Expect crypto participants and cryptocodebases to act against their otherwise ‘rational economic interests’ in some cases — these may only be edge or corner cases, true. Money, fiat currency, economic relations had — prior to Satoshi — always been about the mental conversation that one tribe of humans attempts to force onto another. Cryptocurrency recasts that conversation by introducing distributed, decentralised consensus into the engine room. Society will now have to learn how to conduct economic mental relations between tribes in a new, opt in, way — the old world, coercive approach to intertribal economic relations will no longer suffice. Money used to be the way I could virtually ‘force’ you to do something I wanted you to do, without actually physically forcing you: if I/we made the incentive large enough/impactful, then I/we could know in advance, with high confidence level, that you would cave in and comply. At a societal level, money (largely via debt) was the way the tribe made communal decisions about what to spend tribal wealth (resources, skill, time) on, then ‘encourage’ (read: compel) workers to do those things — politicians spent your tax dollars on wars you didn’t want and sent you the bill for the overage/vig (whoops, my bad, sorry!). Money was the way we tokenised our tribal thoughts about those decisions — we mentally asked our self what we wanted/we asked the tribe what it wanted, then we used dollar bills, taxes, lending laws and threats of jailtime to make those things happen. Cryptocurrency and cryptoeconomics, via Nakamoto consensus (distributed trust, noncoercive), now recast/remake/redefine that mental conversation. The new mental conversation will not so much be: “how do I make the other guy give me what I want, else I’ll get my captured Senator to pass laws and then sue him”, to instead be: “the other guy has complete decisional control over his private key, so how do I ask nicely and hope he’s willing to share with me a little?”. Humans understand how to use a stick to hit the other guy if he fails to comply; we’re not so good at asking nicely then hoping the other guy agrees. Crypto will highlight this to us. Cryptocurrency and cryptoeconomics is being birthed in the coming age of endless availability: AI’s, robots, drones and 3D printing will automate/ supercharge many previously-manual processes. The ‘rational economics’ thinking of the 20th century’s scarcity and ‘supply/demand curve’ models will not apply, or as strongly/in the same old ways. In a world where resources are automatically made and supplied for us by clever software, what then do we need money for, what changes? The signalling, allocation and throttling functions that money used to serve wither away, to be replaced by the new cryptoeconomics contours that unfold. The devs and codebase and AI’s will flex and pivot as they work to keep up with those shifting cryptoeconomics sands. Cryptocurrency built a trillion dollar reality: no one planned it, no one controlled it, no one directed it, there was no master plan, no authority. There was no Marshall Plan, no Bretton Woods. Because of those aspects of crypto, the human race can henceforth never again use those excuses as reasons why things can’t be done. Rather, crypto shows that perhaps its realities are the only ways by which anything useful will get built from here on. Crypto assumes that some actors within it and some of its users are Byzantine, and it designs its organs with that factor in mind — it expects interference from the very actors who ought not be cutting their own throats. Crypto is not fragile because the actors within it are occasionally Byzantine — crypto expects them to be so, is antifragile/robust to that and this is a key reason crypto does so well. It may be the case that anything new and large and complex attempted by humanity in coming years/decades can, will or is more likely to achieve strong robust outcomes only if it follows the tenets that crypto’s developmental path has shown us, along with its inherent Byzantine assumption. The specific ‘new idea’ of cryptoeconomics and cryptocurrencies is that it’s not enough just to ‘nudge’ [behavioural economics] entities into not being malicious (e.g. PoW), further, that it’s not enough even to punish [game theory] them for malicious attempts (e.g. slashing); the ‘new idea’ is to make the self-cost + the damage to the item being attacked so asymmetrically high compared to the possible payoff to be gained that the malicious act becomes logically and practically insensible to pull off, a complete calculus fail –the payoff is asymmetrically punitive to both sides, not just one. If every time you tried to rob a bank, you not only got caught, fine and jailed for life, but that the bank and its cash disappeared in a puff of smoke as you attacked it, then what would your incentive be to make the attempt (assumes non — pathological reasons)? A well designed cryptocoin/ cryptoeconomics is the paradoxically curious animal that is simultaneously a socially friendly (open sourced) money that anyone can own and control (decentralised, anyone can be a full peer), whilst being one that collapses to zero were you to attempt to overpower it + it simultaneously chops off your arms in the attempt.
    This ‘new idea’ of cryptoeconomics and cryptocurrencies leads to a proposed new heuristic to describe the logic of such crypto-attacks:
    • Your attempted attack will fail, up to some threshold (34% attack, 51% attack)
    • You are free to attempt an attack (crypto is open sourced, decentralised, p2p), but as punishment you will lose your stake at risk and/or the cost-to-you of mounting your attack
    • If you succeed in an attack, the degree to which your attack succeeds will paradoxically cause the prize to diminish in value by the same quantum or more [where Attacker = single or coordinated group of attackers]
    This ‘new idea’ of cryptoeconomics and cryptocurrencies may be the first instance of such a mechanism in an economic/social domain, and may, once knowledge of this aspect has rippled out further, come to be the heuristic that informs the defence posture of all such future systems/designs. Such “new idea” asymmetrical retaliation might start to infuse all complex system designs that need protective capability/robustness/antifragility against attackers. The intuition may be a heuristic:

  • this is not ‘defence by barbed wire’ — strong perimeter;

  • this is not ‘defence by asymmetrical reparations cost’ — successful attack costs the attacker exponentially; attacker self censors due to probable cost if attack attempt fails;

  • this is defence by ‘try attack me and not only will I kill you but I will blow myself up entirely too’. Perhaps this ‘new idea’ of cryptoeconomics is a new, 21st century flavoured, re-interpretation of the credo of the aspiring terrorist.
    This ‘new idea’ taps into the notion that you can’t just hope that economic rationality (incentives and counterincentives, nudges) are sufficient to cause actors to behave. Further, you can’t just hope that social punishment (slashing, forfeiture, penalties) will deter the remaining recalcitrants — the hardliners for whom economic rationality wasn’t a barrier. The ‘new idea’ is the capitulation that both of the above are not sufficient: essentially, the ‘new idea’ of cryptoeconomics repudiates the supposed efficacy of ‘rational self-interest’ and improves it via nuclear-strength mechanism design: namely, attacker self censors, never mounts an attack, due to the outsize cost to both attacker and their victim/prize. This is ‘Minority Report’ and ‘Clockwork Orange’ levels of engineering: get the attacker to never bother to make, or want to make, an attempt.
    It is not a lost observation that the irony of this ‘new idea’ mechanism in cryptoeconomics is that we nudge attackers into using self-censorship to get the attacker to not even bother to attempt an attack, yet the whole canon of cryptocurrencies is that all we p2p peer humans should be free to have our own censorship-free monetary system; we achieve a censorship-free currency/system by ‘encouraging’ attackers to self censor.
    The ‘new idea’ recognises that some attackers may have their own illogical reason/s still to proceed with an attack, and that we are mainly powerless to design a cryptocoin codebase that somehow they fear enough so as to not proceed with their attack. Instead, the ‘new idea’ taps into the martial arts-like notion that one should attempt to get an attacker to use his/her/their own strength to defeat them self or to talk themselves out of proceeding at all. Distributed ledgers as expressed by cryptocurrencies may well be the first time the human race has ever invented a truly incorruptible/immutable recording system. Most systems of record that humans have so far come up with can be brute forced or otherwise cracked (e.g. social hacking) by a sufficiently motivated entity or entities. A very well designed distributed ledger cryptocurrency should prove resistant even to a quantum computer that has ‘quantum supremacy’. The human mind normally is capable of functioning through cognitive dissonance — simultaneously holding two ideas in mind as both being true yet that logically conflict with each other. Human social structures have to deal with this ‘dual truth/lie’ as an ordinary part of their function; e.g. police and the justice system sorts through lies to arrive at a socially acceptable verdict; couples can exhibit spousal abuse yet they still love each other; presidents can lie yet their base approves of them for it. Distributed ledgers as expressed by cryptocurrencies may be the first system that relives us of the need to sift truth from fabrication. The social and psychological impact of this may be profound. Imagine the events, facts, measures and fingerprint of your whole day gets committed, like a running commentary, to an immutable ledger, by your AI-powered avatar, written to an IPFS-style (IPFS/Swarm/Sia/Storj/Filecoin et al ) disk/file system: the record of what you did, where you were, who you saw, what you said, all immutably recorded. Human social functions normally allow degrees of flex, to accommodate when we collectively develop a need to re-write/muddle/pervert what had previously been a consensus derived record of history. Distributed ledgers as expressed by cryptocurrencies when used as immutable recording systems mean that the past statement of fact can no longer be flexed ex post. It remains to be seen how humans and societies cope when this post-diction capacity is removed. A good lie is easier to believe than the truth. Asking humans to live in a gated community (e.g. IPFS as substrate for all knowledge storage) where one can’t twist/alter/reinterpret previously recorded facts is to ask humans to live in a kind of mental prison where there is no hope for the relief that lying can offer. Cryptocurrency and its constituent concepts are not money, they are not a financial system, they are not an economics: crypto is a pedagogue. Crypto teaches its students, through its subtle invasive mechanisms, what certain things really are: what it means to be decentralised, what it means to have censorship resistance, what it means to be self-sovereign, what immutability is. It is doubtful that there has previously been such a large scale attempt at infecting 7 billion humans with these particular memes. By analogy, we may consider the advent of mainstream internet consciousness was also a pedagogical phenomenon that taught new things on a mass scale: what it means to meme (self genes and selfish memes), to be private (Facebook, Google own you and your data), to hack (Snowden, Panama Papers, Equifax), what a singularity is (AI, ML, self-driving cars, DAO’s). It is said that we advance by going from failure to failure, that we fail our way into the future. Crypto exemplifies and proudly restates this, crypto is proud to offer itself as a vehicle to teach us new lessons. Legacy anchors (regulations, governments, central banks, old world economics canon) try to hobble or impede crypto’s progress, crypto patiently rewires itself so as to teach us what antifragility, convexity, mechanism design are. Another pedagogical aspect of cryptocurrency is that it is, by stealth or overtly, teaching millions of people to reconsider what money is, what an economy is, what is the purpose and function of a market, of exchange, of trade, of currency sovereignty, and many other economics aspects. Crypto is a lesson to its users about what they ought to believe money is. Cryptocurrency is nudging people to have a conversation inside their own heads about what they them self think money is, what they want money to be/do, about who is allowed to tell them what money is. Choose wisely. Vitalik Buterin/Gavin Wood popularised (many would say they invented) the concept of the EVM as being a “world computer”. The notion is that the EVM is a single engine that does blockchain-based p2p computation of smart contracts and state, but one that carefully happens to be distributed over thousands of individual Ethereum node systems running the node software. Taking this ‘singleton engine from multiple cells’ concept, we might also philosophise by saying that cryptocurrency is cannibalising own minds by redefining our previously-agreed notions of what money and economics are, are reinventing those into a single, “EVM/world computer-style” new notion. We are all, by stealth, being led into realising that we can/must invent our own blockchain version of economics, anew. Millions of minds have now redefined inside them self their concept of money, currency and p2p fiscal self-sovereignty. Cryptocurrency is now like a single smart contract running inside the EVM, where we all converge on the same new understanding of what money is, we compute the concept of money on a p2p basis, we all know how the blockchain works, we have shifted out of fiat to crypto, and now we all compute money as transactions on this new ‘money EVM’. Crypto’s use of consensus algorithms/engines is, in a blowback kind-of-way now, causing us humans to come to a new singleton EVM consensus about what money is to be to us henceforth. One of the most magic charms of cryptocurrency is that it inverts the human concept of trust. Up until cryptocurrency, humans in the main trust things: people, structures confirmed by ‘proof of work by long social history’ (companies, governments, the legal system) and lived experience (my ontology, my social structures, my beliefs). Crypto instead teaches us to trust only non-things: in crypto, trust is an emergent property. Trust emerges from the notion that you can trust deeply confirmed blocks in the chain, elliptic curve cypher mathematics as applied inside a cryptocurrency / the one-way calculation aspect of encryption cypher maths. Anyone can do the hash calculations and arrive at the same answer, anyone can recompute the blockchain’s history. Anyone can do the asymmetric key computational maths and arrive at the same answer. The maths works no matter who or what does the work, and it is invariant, 100% deterministic, inviolable. I trust confirmed blocks, blockchain finality, Ethers and Ethereum smart contracts and their mathematical integrity more than I trust any CEO, lawyer, politician or family member. The fact that the maths calculations are one-way computable is the aspect that feeds the emergent aspect of the trust: the trust is not explicitly encoded into the codebase, the trust emerges from the correct operation of the (one-way crypto) maths functions and the derived phenomenon: that whilst it is trivial to one-way compute the hashes, it is nigh impossible to reverse-compute that calculated answer. You can ‘trust’ this one-way maths’ irreversibility to be robust against reverse-calculation: the trust only exists whilst someone can not reverse the calculation, the trust emerges from that property. Solid trust achieved by not existing, but by being a computed, implied outcome only. The other leg of the chair that forms up the trustworthiness aspect of crypto is that hackers have had ten years now to assail the attack surface of that encryption maths and have yet to break the cyphers as applied, e.g. SHA256. One might say that the most secure (trustworthy) system is the one that has been probed for its weaknesses the most by the highest group of motivated actors. This forms up part of the “new mental trust equation”: humans will increasingly realise that only a system (e.g. crypto’s one-way cypher maths) that has been most heavily attack-probed by incentivised crackers is a system that one can apply the label ‘’trustworthy’ to. Humans will naturally resist the idea that they ought not trust real world things from their collective social experience, but instead should only trust the finality that a ECDSA-protected blockchain currency has confirmed through never having been hacked This notion of emergent trust will not sit easily with a human brain (that is wired by evolution, experience and social convention) to place trust in things rather than non-things. Trust will transition from being human >trusts> human to human >trusts> only finality via one-way maths.

    Disclaimers: just my ideas about possible scenarios for near-term future. This is not investment advice. I’m ego-driven, clueless and biased, so do your own thinking. I’m not qualified, I have no special privileged position to drive my insight, I’m a nobody, is what you should assume about me and what I say here.

Created with Sketch.Content is"CC-BY-SA 4.0" licensed
Article On-chain
Article Author





Related Articles
random cryptocurrency / cryptoeconomics thoughts #3

The below is ungroomed, thinking out loud, a collection of thought bubbles written as a messy tangle. It has not been processed for concision/to make it easier to read. Patience and indulgence is required. There is not a sequential flow, paragraphs are ordered randomly. No effort was made to format for easier reading/viewing. tl;dr I know you just want the money shot: incentive structures matter, crypto’s hard-encode them Crypto codebase is so smart, it will become a person at law who thinks for


20 May 19

random cryptocurrency / cryptoeconomics thoughts #6

The below is ungroomed, thinking out loud, a collection of thought bubbles written as a messy tangle and has not been processed for concision/to make it easier to read. Patience and indulgence is required. There is not a sequential flow, paragraphs are ordered randomly. No effort was made to format for easier reading/viewing. Perhaps the most significant advancement that Satoshi gave us is/was Nakamoto Consensus. Prior, we had only Classical Consensus (Lamport-Liskov). We now also are on the cus


20 May 19