That's because they stopped being useful in pushing back against global communism. Chile, for example, ditched Pinochet but kept the liberal US style economy.
And Argentina is kept in line via WHO loans and regular scourging through economic means.
Didn’t read the entire post but I think the characterization of PCP’s is a bit off - some SNARK constructions and STARKs both use PCP’s - using a linear PCP vs a PCP doesn’t impact transparency or proof size, that’s more a function of the commitment scheme
I meant to skip over PCPs because it would need a separate posts and make the whole longer, I also avoided explicitly mentioning PCPs or SNARKs constructions from PCPs[1] for the same reason.
Yes, concluding about the transparency vs proof-size trade-off is a mistake of my part.
Thanks for taking the time ! will push an update on that bit soon.
Author here, I did gloss over PCPs certainly, and agree that SNARKs also use PCPs and I might have made the terrible conclusion about proof sizes.
Will push an update soon ! thanks for taking the time.
It seems like a VRF might be a more natural choice than a commitment scheme for verifiable randomness, since it doesn't require any honesty assumption for participants, and Keybase already manages keys (though maybe it would be a problem if participants could change keys midway through the ceremony).
I thought that the post was fairly unclear, but for me, it seems like the main argument against 538 is that it makes unfalsifiable claims about the probability of individual elections - picking a winner is a falsifiable claim, but assigning a probability always allows Silver to claim something like “even events with a 10% probability occur frequently” even if his model assigns a high likelihood of victory to the loser of an election.
My favorite Nate Bronze takedown remains Carl Diggler[1].
Not a pilot, and not trying to be an armchair pilot, but why would he have ascended all the way to 36000 feet? Isn't there a real risk of depressurization with hull damage?
I'd rather not trust my life to "probably". If there was some damage to the hull or a cargo door that could lead to failure, it's under more stress and is more of a hazard as the altitude increases.
Engine power is a fraction of weight. Even at full thrust (which is rarely used in commercial service, even for takeoff), the engines are only capable of 0.25g or so.
The point that he's making is not that individual observation should trump objective measures, but the epistemic claim that if what we perceive contradicts those measures, it's worth interrogating the validity of those measures.
I don't think that the particular example that you're giving is "fraught with peril". Note the lack of an upper bound on the labor force participation rate that you cite - 16+ includes people of retirement age, whereas the measure given in the article shows that the official unemployment rate for men 25-55 is just 1/3 of the "true" unemployment rate that includes disaffected workers. This is significant.
>whereas the measure given in the article shows that the official unemployment rate for men 25-55 is just 1/3 of the "true" unemployment rate that includes disaffected workers. This is significant.
Using that as the "true" unemployment rate is, frankly, bullshit. The number they're citing is the OECD "employment rate"[0][1], defined as the employed share divided by the total population in that demographic. Thus it includes the disabled, stay-at-home-parents, the leisure class, those enrolled in education, and anyone else who is not working but has no desire to do so. The decline is almost entirely driven by the reduction in labor force participation rate from 97+% in that demographic in the 1960s to 88.8% today.[2] Is it unreasonable to expect males may on average voluntarily spend 3 years outside the labor force during their prime years?
If you want to count discouraged workers, use U4 — which currently stands at 4.1% vs. the U3 at 3.9%.[3] (U3 is 3.0% for the demographic in question.[4]) U6, which adds all other marginally attached workers and those employed only part-time who'd like to be full-time, stands at 7.4%.
If your concern is falling LFPR, say so. If your concern is discouraged workers, use U4/U6. But don't change the denominator on the unemployment rate from the labor force to the whole population and act like it's some huge hidden increase the authorities have been concealing from us.
I don't agree that using the OECD employment rate is bullshit. The argument is not that at a single point, the fact that this other measure of unemployment is much higher than the standard measure shows that the economy is not healthy, the argument is that the divergence between the OECD employment rate and the standard measure indicates that the economy is unwell. If you deny this, then you're forced to argue that somehow the number of "disabled, stay-at-home-parents, the leisure class, those enrolled in education, and anyone else who is not working but has no desire to do so" has increased within the same demographic (25-55 males, so generally post-school), while wages have remained flat.
>this other measure of unemployment is much higher than the standard measure shows that the economy is not healthy
No it doesn't. Aside from it being a nonsensical measure of 'health', the overall 'employment rate' is well above pre-1990 rates (as women have continued to enter the workforce).[0]
>If you deny this, then you're forced to argue that somehow the number of "disabled, stay-at-home-parents, the leisure class, those enrolled in education, and anyone else who is not working but has no desire to do so" has increased within the same demographic (25-55 males, so generally post-school),
Of course it has. Being a stay-at-home-dad was unthinkable in 1950. Post-graduate education is increasingly common, as is mid-life career switching. I can think of many reasons why we shouldn't expect a 98% labor force participation rate from that demographic in this century.
You misread my comment - I'm saying that the argument is not that the fact that this other measure of unemployment is much higher than the standard measure shows that the economy is not healthy, but that the sudden divergence over the past decade hasn't been plausibly explained by anything except an increase in disaffected working-aged males.
>Of course it has. Being a stay-at-home-dad was unthinkable in 1950. Post-graduate education is increasingly common, as is mid-life career switching. I can think of many reasons why we shouldn't expect a 98% labor force participation rate from that demographic in this century.
All of the points that you raised should not show such a sudden effect in the past < 10 years.
It was pretty interesting (and disquieting) to see a linear trend.
I don't know how the statistics are gathered, but I wonder if they separate those who wish to/need to work from those who don't. For those who have given up searching, that is.
> I wonder if they separate those who wish to/need to work from those who don't. For those who have given up searching, that is.
Indeed they do. The rate you most commonly see tries to measure unemployment in the active labor force. If you're not actively looking for work, you're not counted as unemployed. There are other measures which do count "discouraged workers" and others not usually included. They're much higher.
Yea, I was talking about further filtering "discouraged workers", or those who've given up searching for work. I think it'd be interesting to apply the "wish to/need to" filter to those who are employed, too!
This article[1] gets into a bit, but
"Some men choose not to work and can afford not to. That’s great, Furman says. But for many, probably most, dropping out of the work force not only means a lack of income but also a loss of the dignity that comes with not working."
>Discouraged workers are a subset of persons marginally attached to the labor force. The marginally attached are those persons not in the labor force who want and are available for work, and who have looked for a job sometime in the prior 12 months, but were not counted as unemployed because they had not searched for work in the 4 weeks preceding the survey. Among the marginally attached, discouraged workers were not currently looking for work specifically because they believed no jobs were available for them or there were none for which they would qualify.
True but has a common failure mode, aka "how X won the election if I don't know a single person who voted for him?". It's rather hard to directly perceive a diverse 300+M trillion-dollar economy without resorting to some aggregating measures.
>>The point that he's making is not that individual observation should trump objective measures, but the epistemic claim that if what we perceive contradicts those measures, it's worth interrogating the validity of those measures.
I don't know if I agree with this. I mean okay, if your house is on fire and your thermometer is showing a comfortable 75 F, you should probably question your thermometer. But in the overwhelming majority of situations, especially involving extremely complex phenomena such as a national economy, one should absolutely not trust their perception or use it to question the validity of empirical evidence that has been collected.
On the first Friday of every month when the Employment Situation report is released, we get new analysis on jobs created, hours worked and hourly wages. This is great information, the only issue is that the hours and wage information is an average, not a median [2].
If economic reality changes to a point where citizens/politicians/economists become concerned with income inequality, this measure begins to lose it's significance. Average wages could increase while median wages fall. It is still "empirical evidence" but it isn't necessarily measuring what we think or want, labor force participation rates before/after women joined the workforce is an example. We should always be questioning the validity of our economic measures over time.
Why not? Determining how we measure the health of our economy, and implicitly, what we want to optimize for, is a political process, not simply an objective, scientific one.
The emphasis on humor in Mozart and Beethoven is interesting, particularly since it seems almost absent in the later Sonatas. I remember hearing Martha Argerich note that it's present in early Beethoven, and then disappears in Chopin and Liszt, and then reappears a bit in Ravel, but truly with the Russians. I would be interested in whether Brendel distinguishes between the ironic humor (Mediant vs Dominant) and any other types.
When discussing how crypto-assets should be valued:
> In that case, you can take the simple model and say a cryptocurrency’s valuation is the net present value of the transaction fees that it’s getting. This, by itself, surprisingly does give fairly decent valuations.
> For example, Ethereum’s transaction fees tend to be about $500,000 a day recently, which is about $180 million a year. If you tried to value the ether market cap as some kind of corporation, then the “P/E ratio” is only somewhere in the low 200s, which is high for a company, but not off-the-charts absurdly high.
Isn't this completely circular? He's saying that the value of the transaction fees per year measured in dollars equates to some reasonable fraction of the market cap measured in dollars... but wouldn't this be true if the market cap were much higher or much lower? It's just a comparison of the total supply to the amount being circulated, rather than a substantive claim about how each token (or the market cap) should be valued in dollars.
No, because transaction fees are an ongoing negotiation between miners and the people issuing transactions. That market can adjust to change in the ether price, and keep the real-world transaction cost the same. Or the fees can wildly fluctuate due to variations in traffic, even when the ether price is stable.
That position isn't supported by volatile transaction fee data - excepting spikes for spam or high traffic, the transaction fee denominated in dollars generally tracks the price of ETH. Probably because the overwhelming majority of ETH activity is speculative (60% of ETH held on exchanges), and traders view transaction fees as simply a percentage of gains/losses and are more amenable to paying higher fees.
For example, if the market cap were 1/100 of its current value, then the transaction fees per year, denominated in dollars, would be $1800000, yielding the same "P/E" multiple, and justifying the valuation at 1/100 the current market cap.
If people are paying transaction fees in exchange for some real utility then you'd expect them to pay the same dollar amount for them even if the price went up or down (ignoring second-order effects like increased trading activity).
That's a big "if" though. I'm not sure what portion of transactions are related to trading/speculation vs. something more useful.
Sure, I'm assuming that most of the transactions are trading/speculation, given that 60% of ETH is held on exchanges, and if the question is how we should value ETH, a measure that depends on the current market value of ETH seems flawed. Plus, the base transaction amount is denominated in ETH: .00042, and is fixed (though I think that the gas conversion changes?).
Maybe I'm just being cynical (or jealous!), but when I read "Vitalik Buterin, who has managed to synthesize insights across those fields into successful, real-world applications like Ethereum", I think it's worth asking what qualifies as successful, because to me, the inflated market cap of a speculative asset isn't an objective measure.
I think that Ethereum's market cap has perhaps created an unearned reputation for solving an essential problem (running any verified, distributed computation) and being the best solution to large-scale verifiable computation, when I would argue that verifying a computation in a distributed system by having each node (or some large subset of nodes, if sharded) perform every step of every computation is not an optimal solution.
Of course it's not optimal. But consensus mechanisms resilient to Byzantine attack are an active area of academic research, and traditionally they've all required all nodes to hold the complete state. Bitcoin's contribution was a new method that scaled to a much large number of nodes and didn't require a known set of nodes; the tradeoffs were low throughput, long latency, and offering only probabilistic finality.
But it wasn't Vitalik who created that; what he did was extend it into an application platform. That platform is working, with live applications including decentralized exchanges, Maker (stable currency and collateralized loans) and Augur (prediction market with decentralized bet resolution).
Sharding is one of the main things he and the research team are working on now (along with proof of stake). The current design uses 4000 shards, so unless you consider 1/4000 to be a large subset, it would answer your objection if they manage to get it working.
Verifiable computation is an active and developed area of research in cryptography, and the validation method of having every node run every step of every computation is the most naive approach. There are plenty of ways to prove to you that I faithfully executed a computation that don't rely on your also having performed the computation.
I don't know how active the Maker userbase is, but Augur has 700 monthly active users.
It seems like it would be possible to manipulate the validator set for a particular shard - also, isn't part of the point of ethereum to be able to access information from other contracts? How do you guarantee that that contract is on the same shard?
People are working on things like Truebit and zksnarks for Ethereum, but those still depend on base layer consensus for posting the proofs. They just reduce on-chain computation.
Augur just launched a couple weeks ago.
Preventing validator set manipulation is an important part of sharding research. Part of the solution is getting good random numbers that can't be manipulated, and there are several approaches for that.
Cross-shard communication gets complicated. There are some ideas for it, but that's several years down the road. It'll still be a big step up to have lots of shards, each with the capacity of the full blockchain today, sharing the security of all of them but not getting swamped with traffic if something gets really popular on another shard.
Money. A few engineers are impressed by the technology but most people are excited by the money and prospect of a lot more money in the future. That's what drives crypto fanatics instead of trying to generate value and that's why I think it will ultimately fail.
"Trustless" is a very tricky concept. Instead of trusting a centralized authority bound by laws and regulations, you trust an assumption that enough independent parties will continue spending significant amount of computation power on mining without merging or colluding to execute a 51% attack. It holds for now, but once the spirit of the 2017's hypergrowth fades away, things may start to look differently.
Voting has nothing to do with it, let alone votes per person. PoS doesn't require a voting system as EoS has implemented. The problem is having transaction inclusion decided by individuals that are voted in. It's analogous to having a democratic-republic government running a computer program.
Digital scarcity is finally invented, but people feel like it will have no use-cases?
Automated finance bots (smart contracts) that work without keeping a server up has no use-cases?
All fiat currencies are deflated continuously, but you believe 100% of people will prefer this?
The ability to embed money into a digital good has no value?
How can it ever go to zero if there's someone out there who thinks it might be worth accumulating just for the remote chance that one of these ideas works out?
The idea of it going to "zero" seems irrational to me.
Totally untrue. Cryptocurrencies are artificially scarce, not actually scarce; this is most obvious when two chains hard-fork and the sum of their combined value exceeds the total value of the parent chain. You can't "fork" a scarce resource.
Even without a fork, the scarcity is still an illusion because the consensus protocol can be modified to create an infinite number of tokens if that is the prerogative of the developers or community.
> Automated finance bots (smart contracts) that work without keeping a server up has no use-cases
Of course an automated finance bot needs a server, it's just a distributed and computationally expensive one instead of a centralized one that costs pennies to operate (and I say this as someone who has written profitable trading bots)
> All fiat currencies are deflated continuously, but you believe 100% of people will prefer this
The reason people prefer fiat money is because fiat money can be used buy goods and services and cryptocurrencies generally cannot without jumping through a bunch of hoops that offer no clear benefits in return. The OpSec that goes into "being your own bank" is not worth the cost to the overwhelming majority of people.
> How can it ever go to zero
I do agree that it is unlikely to ever go to zero simply because the limited utility cryptocurrencies do provide (irreversible pseudonymous online payments) will always have a niche use case that will keep the price above zero.
A fork gets you nothing on the original chain, just like inventing "new dollars" gets you nothing except what the market decides "new dollars" are worth. Digital scarcity is still there. If I have an item on that blockchain, you can't take it from me by forking. The only way to "move" that asset would be to get my private key. Forks are an anomaly the market is still figuring out.
> Of course an automated finance bot needs a server, it's just a distributed and computationally expensive one instead of a centralized one that costs pennies to operate
This misses the point. If I want to make a trading bot, I have to find a host, initiate a business relationship with them, and keep it running and highly available. If I make an ethereum smart contract, I set and forget. If I want people to interact with my bot and know it's source code, I can do that on ethereum. Is there another way you can think of to do this and have the bot be transparent--in other words a guarantee that the source code you looked at is actually the source code you're interacting with?
> The OpSec that goes into "being your own bank" is not worth the cost to the overwhelming majority of people.
This is only true until people notice the deflation. The ability to convert electricity to money that doesn't automatically shrink (even if it is not ready-cash) is very valuable in places where it is shrinking quickly.
Forks aren't as simple as you might think - who gets to determine what the "original chain" is in the event of a fork? Suppose that ETH forked and one version changed some consensus rules to increase the supply of currency, and Vitalik, the Ethereum Foundation, all of the miners/stakers, and users moved to that chain - wouldn't that be ETH? But it would violate the principle of fixed digital scarcity.
Confiscation of assets is obviously possible via forks as well, but the fact that someone can't take your money without your private key doesn't relate to scarcity.
As long as crypto is pegged to a fiat currency like USD then the scarcity effect is going to be destroyed by speculative valuations.
USD isn't going to stop being fiat because it's too useful for it to be fiat when doing market regulation.
USD might stop being standard currency, but I expect other countries are going to maintain fiat standards because being able to vary the nominal amount of money supply is a better policy tool than merely controlling gold, because it opens up another theater of war beyond maintaining hard power.
>All fiat currencies are deflated continuously, but you believe 100% of people will prefer this?
I have no idea if 100% of people will prefer it, certainly, some who hold immense amounts of a currency will not, but it was one of the smartest ideas mankind ever had, because it introduced the ability to control the money supply and made currency a very important tool in ensuring stability and made it an effective means of exchange.
Cryptocurrencies are not a means of exchange. They're basically a pile of gold that people imagine themselves sitting on and that increases in value without them doing anything. Which is good if you're one of the few people who have much of it, but it is absolutely detrimental to society at large, which is uninteresting to the libertarianism on steroids that underpins the crypto-craze.
> it is absolutely detrimental to society at large
This may or may not be the case, but regardless, individual decisions are usually made based on an individual's self-interest rather than the warm fuzzy feelings about helping society.
> Cryptocurrencies are not a means of exchange.
It is not necessarily true that that will remain the case. Either way, that doesn't matter if society coalesces around one of them as the "WorldWide Ledger" because individuals are incentivized to avoid inflation.
There is no incentive for any average citizen to engage in crypto-currency activity if it is a shoddy means of exchange and favours individuals who own a large pile of it. Most individuals are debtors, not creditors, and for them, inflation is a boon because it benefits them and erodes debt burden over time at expense of creditors.
It is in the interest of someone who uses money as a means of exchange and not as a store of value (which is the overwhelming majority of individuals) to reject crypto-currencies in favour of fiat currency.
Your average citizen gains nothing from adopting a currency that is highly volatile, concentrated in the hands of a few, and awful at completing everyday transactions.
> Your average citizen gains nothing from adopting a currency that is highly volatile, concentrated in the hands of a few, and awful at completing everyday transactions.
Maybe, but those problems are not necessarily permanent, and it can be valuable without targeting the "average citizen." As a counterpoint to your "gains nothing" view, many "average citizens" used bitcoin for transactions in the early days, only to find out that holding some of their currency in bitcoin made them rich, so they did gain something. Even if that is only a small possibility, the possibility is still there.
> As a counterpoint to your "gains nothing" view, many "average citizens" used bitcoin for transactions in the early days, only to find out that holding some of their currency in bitcoin made them rich, so they did gain something. Even if that is only a small possibility, the possibility is still there.
Confirming that it is preferable to "HODL" than spend this currency...and any gains were at the cost of new bagholders who in turn expected to sell their coins for an even higher price.
I'm sure that's intended as some sort of damning critique, but that's true of most assets in a portfolio. Real estate, commodities, stocks, currency, options, index funds, precious metals--all held in hopes of selling for a higher price.
Then don't call it a crypto currency and pretend its going to replace fiat currencies any time soon.
A currency is something people want as means of exchanging goods and services, an investment instrument isn't that.
And crypto has performed terribly as an investment instrument as well...can you imagine the chaos if the stock market swung 5-15% from one week to the next for months on end?
Lastly you're conveniently ignoring that stocks can pay dividends, and don't serve only as a speculative asset... Because publicly traded companies generally produce something useful and of tangible value to society, unlike cryptocurrency projects.
Market forces will ensure a bitcoin exodus until central bankers stop acting like mental patients with unlimited printer ink cartridges.
The Bretton Woods System ended under Nixon to prop up the military ambitions of this nation, bad move using debt to start pointless wars. Current monetary system structure is deeply flawed, until reform BTFD in BTC.
Note: If you don't feel like paying for the next war, it's a nice bonus in BTC's favor.
Individual cryptocurrencies are inflationary or deflationary depending on their algorithm. Monero, for example, has a 2% tail inflation built in to account for lost money and an expanding economy. Inflation can also be tweaked or changed based on an agreement of the miners, or whatever else is built into the protocol - it could be a consensus mechanism beyond mining, or a central administration (which wouldn't be different than fiat really). My main gripe is that the popular logarithmic inflation curves are WAY out of whack and reward the early adopters way too much, making it more of a ponzi scheme than something useful, and ultimately making themselves poorer because if someone feels ripped off, they can invest in a new coin where they are the early adopter, essentially inflating the total money supply.
That said, inflation isn't always a good thing for some people. Neither is a centrally controlled money supply which also has its own winners and losers
Sure you can conceive of a crypto currency that is centrally controlled, targets inflation and focusses on price stability, but then you have simply invented 'crypto-fiat'.
The key point is that Ponzi schemes you are talking about are not an accidental byproduct, they're the entire point. There's no reason to reinvent electronic banking, we already have electronic fiat money that works very reliably.
Cryptocurrencies are digital gold and basing an economy on them would be like bringing the gold standard back. The problem being that every economist will tell you, for good reasons, that this is a terrible idea. Cryptocurrencies are not a technological innovation, they're a technological regression. They make things that we already can do more complicated and energy expensive which is the opposite of what technology is supposed to do.
It has already gained adoption outside of that sub cult. All of the major financial institutions are researching applications of blockchains - many on top of private ethereum deployments. Ethereum is PROGRAMMABLE MONEY. It's a technology that's certainly here to stay, but that's not to say that ethereum itself will be the dominant platform.
What if someone just programs more programmable money?
Also why would anyone want to use a system where over 71% of all money is owned by a small group of anonymous oligarchs?
Presale ICO / Premine ( max cost $0.50 USD per ETH )
= 72,009,990 ETH
Total Supply today (Jul 19th 2018)
= 100,773,797 ETH
Source:
https://etherscan.io/stat/supply
Not to mention, Ethereum and DAPPs can't validate any interesting external data without relying on centralized trusted data sources - at which point it becomes cheaper, more efficient, practical and legally accountable to just set up a normal database app.
Most financial applications or experiments in the space are not conducted on the public blockchain so the oligarch aspect of it is not applicable - and no someone can't just program more money on the public blockchain - the blockchain exists to validate the amount of money flowing through it. If someone "programs more" that's essentially a fork and they can't transact with it unless other folks in the network fork as well.
I hear this argument a lot, I'm not old enough to remember these times but I wonder how much is true. According to Wikipedia ARPANET was established in 1969, then:
> In 1971, Ray Tomlinson, of BBN sent the first network e-mail (RFC 524, RFC 561).[59] By 1973, e-mail constituted 75 percent of ARPANET traffic.
> By 1973, the File Transfer Protocol (FTP) specification had been defined (RFC 354) and implemented, enabling file transfers over the ARPANET.
Meanwhile Bitcoin is almost 10 years old and all we have is speculation, scams, a near-useless currency and many promises. What we don't have is a useful application that showcases what only cryptocurrencies can do.
I hear many influencers in the space (mainly in podcasts from a16z and Laura Shin) claim that crypto is matured to the equivalent of the mid-90's internet, but I'm skeptical it's anywhere close to that far along. Thanks for sharing these ARPANET milestones, it puts the situation in a better perspective.
NCSA Mosaic was released in 1993[1], so I think "mid-90's" is convenient for people who can't easily grasp what the internet was prior to the web. The TCP spec was published in 1974 and became a standard in 1983[2].
Consider these milestones in the crypto space[3]...
> 2008 bitcoin whitepaper published
> 2009 first bitcoin transaction sent
> 2010 Mt Gox bitcoin exchange established
> 2011 BTC market cap exceeds $1 billion (indicates activity and liquidity)
> 2015 ethereum launched
> 2017 crypto market cap exceeds $100 billion
We're still seeing early protocols contending to become standards. The analogy isn't perfect, but Bitcoin and Ethereum seem more likely to be analogous to ARPANET than to TCP/IP. It will be a couple more years before blockchain interoperability platforms (like Cosmos[4]) are fully operational, and another year or two after that before we get a killer app that's as widely accepted as the first web browser.
Returning to your final point, the amount of utility already derived from Bitcoin and Ethereum is fairly impressive considering their young age.
>What we don't have is a useful application that showcases what only cryptocurrencies can do.
We do, and it's been running basically since shortly after bitcoin's creation: buying drugs. If mainstream cryptocurrency interest goes back to zero (which seems totally possible), people will still be using bitcoin/monero to buy drugs online.
TBF I own some Ethereum as a hedge in case I'm wrong, but everything I read about it in terms of practical use, I find not impressive (beyond the tech which is impressive).
The computer and internet I think it was easier to see the practical value there (though not necessarily the final form both would take) but that may be 20/20 hindsight.
A similar sentiment drove the technocratic elite to a centralized economy in communist china. It may not be beautiful or ideal but it’s fairly effective at coordinating large amounts of people, especially if they are at odds somewhat.
Light wallets validates blocks by checking PoW in their headers. In Bitcoin, a header is 80 byte long, that's about 40MB as of today (~500k blocks since inception). Saving 39.9 MB on light wallets isn't worth starting a new blockchain.
No, my point is that light wallets don’t validate the content of blocks...so an attacker could create a long chain of invalid blocks and fool light clients. That’s a significant issue.
On this point, light wallets are just as protected as full node.
Light wallets are able to validate blocks just from their header, thanks to the structure of a block. Indeed, all transactions in a block are commited in a merkle root hash that is included in the block header, therefore it's impossible to invent fake transactions without forging a whole new branch of blocks, from the fake block up to the last block. Just like for a full node !
Light wallets are not perfect though, potential attacks are summed up here, slide 13: https://breaking-bitcoin.com/slides/SPVSecurity.pdf
Coda doesn't improve on this issues at all.
Also, argument of authority: I work for a wallet provider
I’m not sure why there is a misunderstanding here, especially given that you work for a wallet provider. The attack is as described: an attacker forks, mines invalid blocks, which are caught by full nodes, since they validate the contents of blocks, but not by the light client - assume for simplicity that the client connects to the malicious node and doesn’t do anything more than calculate PoW. The SPV client trusts an invalid blockchain, fault occurs. Coda is designed precisely to avoid this problem, and any solution that requires trusting full nodes, because it provides constant time verification of all of the contents of every block.
