EDIT: NVM, I RTFMed https://ethereum.org/en/layer-2/. I wonder what the trade-offs of layer 2 protocols are. Less secure?
Layer 2's are currently significantly more centralized than Ethereum L1, but this is not a fundamental technological limitation, and can be improved significantly with more resources and time. L2s right now have fragmentation concerns, where users have poor experiences interacting with a ecosystem of 10+ L2s, but this is largely a UX concern that can be solved in my view, and shared sequencers can help this on a technical level. I wouldn't say Layer 2's have many intrinsic fundamental technological trade-offs.
The actual proving mechanism is either a) too amazing for me to understand or b) not quite figured out yet. But first blobs..
Rollups use either optimistic proofs or zero knowledge proofs for settlement
> The blob stuff is about "ok if that's a thing how do I prove to some L1 contract that I own these L2 funds?".
It's about data availability
> there's (supposedly) a way for you to slowly get your funds back by exiting on L1
They're called escape hatches, you can view the state of each L2 on L2Beat: https://l2beat.com/scaling/summary
Can you post an example of a code that implements either of these? (the fraud proof, not the happy path)
I ask because in the past when I researched L2s such as Optimism, that code was "to be developed".
Sequencer side:
https://github.com/OffchainLabs/nitro/blob/d28682b9300d50214...
Contract side:
https://github.com/OffchainLabs/nitro-contracts/blob/90037b9...
It seems to me that Optimism is lagging way behind - I'm not sure if they have fraud proofs yet even to this day. I consider Arbitrum to have "picked up the torch" so-to-say.
https://github.com/search?q=test+repo%3AOffchainLabs%2Fnitro...
Not sure why it doesn't show up in GitHub's search, but it's right there in front of both of our faces.
https://l2beat.com/scaling/summary
Note the pie chart in the row for each L2. That pie chart notes the security risks for each one based on their tech stack. Ideally, the top L2s should strive to reach "Stage 2" which could be considered as secure as L1 itself, but no general purpose L2 is at that stage yet and most are still at Stage 0.
I predict that by 2030 the cryptocurrency nerds will have discovered SQL and transaction audits.
Furthermore, she is only one example of staggeringly recognized, decorated, acknowledged pillars of computer science who either is now or recently has been doing real, substantial, academically sound, and novel research in this field: Philip Waller of functional programming and category theory fame works (or did recently) at IO/HK, a shop with more Fields/Turing/ACM-type honors than they have places to put all the plaques.
So with all respect to a fellow community member, easy there with the “junk” stuff.
If you want to say: “2017-era ICO exit scam junk”, or “pump-and dump altcoin junk circa 2022”, be my guest as long as you cite examples, there was plenty of fraud during those bubbles just like there is always fraud in speculative asset bubbles, and fraud is bad (whatever Greenspan and Summers are on the record as saying and they are both on the record as saying financial fraud shouldn’t be prosecuted).
But even there, I’ll remind you that the conventional financial system sets no enviable record for either asset bubbles or the attendant fraud: quite the contrary, no one of any real seniority suffered so much as house arrest or community service let alone prison in 1987, 1999, 2001, 2008-2009, or whatever we’re calling this. With three notable exceptions: Sam Blankman Fried is serving 25 years for things that happen on Wall St. every day of the week, and if CZ misses serious jail time it will be by the skin of his teeth. Do Kwon is facing trial for felony market manipulation in absentia.
To my untrained eye, it looks an awful lot like cryptocurrency is the one place in the modern digital financial system where fraud is investigated, prosecuted, where people who go jail, and where the victims of that fraud receive at least some of their money back (I have a friend who held significant FTT and is already sure he’s getting something back, though these proceedings are complicated and we’re frankly a bit out of practice because we stopped prosecuting financial fraud in the late-Reagan/early-Clinton era, so he’s not sure how much yet).
Maybe I’m missing something here, my financial credentials are modest if that, perhaps you or another commenter could explain why the superficial analysis that indicates that crypto is the best-regulated of all the very rough financial markets in the world is deficient?
Consens algorithms are important in both safety and distributed (High Availibility) scenarios. There's no necessary link from research into that, and blockchain in general, and the Proof of X style crypto blockchains specifically.
Can you point to some research of Waller? I've tried to find it to see if it's more directly related, but the only somewhat famous person under that name I can find is a historian, not an expert in computing related topics.
There's some interesting technology blockchains lean on (remember, git storage is a blockchain) but the value proposition of crypo currency blockchains (largely 0 trust) have so far not materialized outside speculative currency.
Which is partially due to misaligned incentives (the developers of e.g. game assets in the NFT case) where the party that would have to enable something do not have an incentive to give up controle.
The contributions that merited his inclusion in a group that includes (in computing alone) people like Charles Babbage KH FRS and Alan Turing OBE FRS are too numerous for any HN comment: he’s got something like 20k citations of hundreds of papers.
His contributions while working at IOHK were IIRC substantially around advanced formal proof systems for typed lambda calculus, the most recent of his IOHK papers I read was describing a System F implementation in the Agda proof system.
I don’t think it’s called an appeal to authority when the topic is the merit of a field of study (I’ve never heard it referred to as a call to authority at all): when an overwhelming consensus of basically every reputable academic and scientific honor and award in the field (some among the highest honors in any field) are attached to research done over decades and reviewed, debated, cited, and recognized by a robust consensus, that’s an argument that the study was important, novel, rigorous, valuable and worthwhile. I cited the consensus of the entire reputable academic and scientific world because that’s how we codify a consensus into a formal recognition that a researcher has in the past, is currently, or is likely to again do important research. I think GP was mistaken to call this research junk or even imply it if someone is going to try to parse it that finely, and I thought that citing the ACM was a better citation than my own opinion.
I agree that git is a blockchain, though not a particularly Byzantine Fault Tolerant one, along with Mercurial and Nix and many other tools many of us use daily.
The broader convergence around previously disparate parts of the digital financial economy is just unambiguously happening: things like FedNow at the high end or Apple Pay / Venmo / Zelle / Wize / WeChat / etc. on a more retail level are arriving faster and faster, placing ever-greater demands on the technology involved, and similar pressures are producing related solutions: the NBBO system in US equities trading to name one example, the consolidated tape that results and the records around it used to be backed by all trades taking place on recorded phone lines, before that by taking place in a room full of witnesses, and before that in coffee shops and other gathering places. All of these systems were workable if imperfect solutions to questions of trust, escrow, reversibility or its converse, and broadly the ways in which Ricardian contracts are generally, in isolation, inadequate to promote a sufficient atmosphere of trust to admit active and reasonably efficient markets.
Cryptographically durable and tamper-resistant ledgers remain in a sort of transitional state where they back non-trivial commerce and much more but still comparatively small amounts of speculation/price discovery: the jury is out on whether or not cryptography and BFT research is going to hit the truly big leagues in terms of notional value: right now they’re somewhere in the rough ballpark of equities transactions daily in the maybe mid tens to low hundreds of billions in notional USD, making both a flea on the ass of an elephant compared to say global forex at something like 5-10 trillion a day, and derivatives are just really hard to estimate, but the notional value of all derivatives contracts is like, easily in the hundreds of trillions and there are days when a lot of that moves quickly.
But I wasn’t making the case that this stuff is like 100% locked-in the future, I was making a much weaker claim: that it’s dismissive and ignorant to call it “junk” and that what limited consequences fraudsters face for financial fraud are tightly clustered in this area.
It’s well-understood that the mechanism design of a combination of a floating transaction fee structure (gas) and a market in that unit of account with a lot of speculative activity in it is problematic to put it mildly: transactions become too expensive to facilitate significant commerce rather often. A lot of things are being tried to improve the emergent incentives, some with more noble motives than others, but that’s finance: if you’re under any illusion that innovation in finance is a constant battle between people trying to generate better outcomes and people trying to game the thing then you can easily disabuse yourself of that notion by learning about the history of finance and I’ll recommend two excellent places to start: the emergence of massive OTC derivatives markets that began in the 1980s but really got big a decade later, and the emergence of fully-digital equities and futures markets around the turn of the millennium.
Wait, isn't git more of a merkle-DAG? I thought one of the defining features of blockchains was effectively branchless global state. My understanding is that DAGs are a superset of blockchains. Is this a wrong?
My comment about "blockchain junk" is mainly in response to the fact that it's nearly impossible to find any investment/involvement in the space without running into complete fraudsters and starry-eyed "entrepreneurs" who view blockchain as some sort of god technology that will lift us out of poverty and/or upend the corporate control mechanisms. AKA a bandaid fix by people who don't understand its actual limitations or the dynamics of the systems they supposedly oppose.
As a system for maintaining auditable global state/knowledge in the face of sybil attacks, yes, it's impressive. However 99% of the projects people reach for it do not require it, hence the term "junk." It's more a condemnation of the space surrounding the technology than the technology itself. I figured that would be somewhat obvious.
There isn’t really a robust consensus that I’m aware of as to what constitutes a blockchain per se: Wikipedia lists git as one, and I suppose that’s as good a source as any absent such consensus.
git is an (often if not typically in practice degenerate) Merkle Tree, the contents of one atomic (and sometimes de facto immutable) node contain a hash (O(1)-verifiably k-equivalent… you know the drill) of ancestors.
In more pragmatic/colloquial usage I might define a blockchain loosely as a “tamper-resistant, directed, and typically acyclic / bounded-cyclic data structure with an implied machine economics optimization around infrequent but critically important fully-verifiable history subject to heuristically-determined / freely parameterized bounds on branching factor, duration in branched states, and a bounded susceptibility to adversarial interference in a verifiable consensus on the periodic elimination of branching on some semi-predicable cadence”, which is pretty hand-wavy but I think captures the spirit of the general usage. By that definition git is only a blockchain by common convention, there’s nothing preventing or even discouraging arbitrary, unbounded branching other than it doesn’t have a ton of widely valued use cases: most any time you’re fine with a branch that never has any scope to interact with any other via rebase or merge you could just make a copy or maybe a copy and a copy of some metadata/history, though git in practical terms is a good tool for such a copy.
And I think you’re right that as with any over-hyped technology, it gets attached to projects that don’t need it when it’s “hot”, preoccupies both investors and entrepreneurs without better ideas for how to deploy their time and money when it’s “in”, and is therefore constantly oscillating between being a magnet for snake-oil types and being out in the cold.
Throw in a bunch of electricity consumption that’s maybe net driving up carbon emissions and maybe net attaching a financial incentive to electricity so cheap that it basically has to be renewable but it’s kinda too soon to tell, and I think I’m now having trouble seeing how crypto three years ago and “AI” last year are any different along these dimensions.
The difference in my view is that AI is probably higher variance by a lot on social welfare, and not because of some dumbass “paperclip-indifferent AGI” tripe.
Blockchain as applied to finance has the scope to create transparency into financial markets and compel governments to open the books on what is and isn’t legal regarding money, for who, and why. It will never like, totally disintermediate government from money, because money is the #1 national security priority of any functioning government, so inventing money that the government can’t control is more likely to buy you a R9x than a Turing Award (in a macabre way it’s darkly amusing to contemplate the fact that it could buy you both). It also has a positive (in my view) externality of creating broad-spectrum incentives for the public to understand a little better how important digital identity, security, privacy, and autonomy are in 2024 and build at least a little muscle memory around running a slightly or maybe even substantially tighter ship on personal digital footprint. I’ve apologized to two friends this week because I lied to them about something that is now news that recently broke on the Onavo/Meta thing TechCrunch ran and I wanted them to hear it from me. I lied about this because before it hit the press, I felt it would have been detrimental to the national security of the United States to talk about it, but what I really wish is that we wouldn’t end up in situations where anyone faces such dilemmas in private industry.
AI has more obviously useful applications at the consumer level (though it’s largely a solution to itself as a way to get information one could previously get from a search engine before it ruined the indexes of search engines by making arbitrarily persuasive falsehoods too cheap to meter, we’ve had spam for a long time, but spam so good it’s convincing to experts in anything other than a bad mood? That’s new.). The danger with AI is that it winds up being something other than “available weight” and “operator-aligned”, i.e. whoever is the last man standing has arbitrary unaccountable power to convince anyone of anything and prevent that from being accessed by anyone else.
So probably higher stakes.
I didn't take offense at all. I find your knowledge of the space refreshing. I have watched blockchains carefully from the sidelines for some time because of my interest in the intersection of economics, state, and technology and how blockchains might change those things. Over time I became more jaded because the scaling problems blockchains run into seem to be almost insurmountable, so my interest has pivoted into less-global, more-scalable approaches (like merkle-DAG CRDTs with some external form of validation).
> There isn’t really a robust consensus that I’m aware of as to what constitutes a blockchain per se: Wikipedia lists git as one, and I suppose that’s as good a source as any absent such consensus.
Fair enough.
> In more pragmatic/colloquial usage I might define a blockchain loosely as a “tamper-resistant, directed, and typically acyclic / bounded-cyclic data structure with an implied machine economics optimization around infrequent but critically important fully-verifiable history subject to heuristically-determined / freely parameterized bounds on branching factor, duration in branched states, and a bounded susceptibility to adversarial interference in a verifiable consensus on the periodic elimination of branching on some semi-predicable cadence”, which is pretty hand-wavy but I think captures the spirit of the general usage.
Have you considered a career in poetry?? Joking aside, this pretty much sums up my view as well. A DAG with a strong gravitational pull towards a master branch with somewhat infrequently changing data. Which also includes git, so you're right.
> Throw in a bunch of electricity consumption that’s maybe net driving up carbon emissions and maybe net attaching a financial incentive to electricity so cheap that it basically has to be renewable but it’s kinda too soon to tell, and I think I’m now having trouble seeing how crypto three years ago and “AI” last year are any different along these dimensions.
Yes, agreed. Let's spin up an immense amount of computing power to train a model that hallucinates when asked basic questions. Again, there is a space where the marriage of a large dataset of knowledge and an automated linguistic system searching that knowledge has great use-cases, but throwing "AI" at every problem is just another eye-rolley fad.
> The difference in my view is that AI is probably higher variance by a lot on social welfare, and not because of some dumbass “paperclip-indifferent AGI” tripe.
What do you mean by this?
> Blockchain as applied to finance has the scope to create transparency into financial markets and compel governments to open the books on what is and isn’t legal regarding money, for who, and why.
This is one of the things I was originally most excited about. Make politicians receive all wages, contributions, etc through some auditable public currency. If you're going to work in the public sector, then you have to consent to transparency. I have my own issues with money (mainly its deficiency for economic transactions) but cryptocurrencies are certainly a step up from it. But like you said, you can't just release some new currency and expect the government to bless it. Some empires had their armies, some their navies, but we have our banks. Our empire is a financial one, and a currency that replaces the USD is a direct attack against the heart of the empire. Many mountains would have to move before that is possible, unless Wall St finds some extra utility in it that allows them to extract more.
> the public to understand a little better how important digital identity, security, privacy, and autonomy are in 2024
100%...cryptographic identity is going to be huge in the next few years (I'm betting on it quite heavily https://stamp-protocol.github.io/).
> AI has more obviously useful applications at the consumer level (though it’s largely a solution to itself as a way to get information one could previously get from a search engine before it ruined the indexes of search engines by making arbitrarily persuasive falsehoods too cheap to meter, we’ve had spam for a long time, but spam so good it’s convincing to experts in anything other than a bad mood? That’s new.).
Yes, exactly, LLMs as they are are a glorified search engine. Search engines have consumed themselves trying to tailor results to their users instead of just fucking showing objective information and are becoming essentially obsolete pay-to-play ad machines.
> The danger with AI is that it winds up being something other than “available weight” and “operator-aligned”, i.e. whoever is the last man standing has arbitrary unaccountable power to convince anyone of anything and prevent that from being accessed by anyone else.
Well, there's more here. As AI markets itself as this sort of objective intelligence machine, it garners more and more trust. As this solidifies it has the potential to shape perception over time to the benefit of the operators/controllers. The obvious conclusion is snuck in advertising, but I'm thinking much more sinister like the editorialization of information to protect the owner classes and the state from any kind of scrutiny. Effectively, a Big Brother that instead of using fear for compliance, softly whispers in your ear. Pair this with the immense surveillance apparatus we've spent decades perfecting (but it's in the private sector! so it's ok!!1) and we're setting ourselves up for a hellscape dystopia.
Definitely higher stakes, I'd say.
It’s also why I gave the example of Dr. Wadler FRS FRSE as someone who even more explicitly did extremely sophisticated research well-received by the academic community unambiguously in this context.
I just do not understand what it is about this topic that turns a normally very thoughtful community of people with a generally very high regard for the research agendas of noted computer scientists into spinal-reflex, knee-jerk critics on a dime.
You hear all kinds of arguments but they generally boil down to some version of “blockchain is for cryptocurrency, cryptocurrency is for financial fraud and financial fraud has no place in our society”.
But every link in that chain is either incorrect or selectively tolerant of wrongdoing.
Tamper-resistant ledgers have applications outside of finance.
“Everyone knows I mean crypto.”
Ok, cryptocurrencies have complex outcomes attached to them, some positive, some negative.
“I’m talking about the fraud which is rampant and the main/only use case.”
Eh, not really, speculation isn’t by itself fraud, although it is true that markets (say, OTC derivatives markets) with a high ratio of speculation to other non-speculative activities generally have more fraud in them, so more or less rampant than conventional digital finance?
“More both relatively and absolutely.”
No. Neither. Trivially false. You think that because you have no fucking clue how much fraud happens either relatively or absolutely in conventional financial markets. Furthermore, the fraud that does happen is dramatically more likely to be investigated and prosecuted, and the victims of that fraud are dramatically more likely to be recognized as victims (unlike for example victims of predatory lending in 2005-2007 who lost their homes and got called criminals into the bargain”.
I don’t want to believe that it’s just the banal fact that we all know someone who got stinking rich speculating in crypto markets and we didn’t (I’ve never speculated in crypto markets and am not in the same galaxy as well-off: I went into debt when the tech job market collapsed last year, with uh, a little help).
But I’m really struggling here to find a more charitable conclusion.
If you’re going to nitpick in an effort to discredit a robust argument by beam-searching for the weakest-looking link in the argument and go directly after it, then at a minimum, be right.
Don’t refute intentional misquotations.
If I’m mistaken about this, I want to be educated about it, I hold all of the famous computer scientists I’ve mentioned in extremely high regard and if I’ve been inadvertently spreading falsehoods it’s a priority to me to both stop doing that and depending on the degree of such an error if it is one, a formal apology could even be in order.
I do not think that this remains a useful or even acceptable forum for that conversation however.
If you know or strongly believe that I’ve made statements about public figures that I admire and that are inaccurate, I’d not only invite but explicitly request that you contact me directly to straighten it out without the distractions and perverse incentives of an audience to a debate that has become more than “spirited”.
I’m reachable at b7r6@b7r6.net and I hope to continue this discussion in a more productive forum than this particular sub-thread.
"Please don't comment about the voting on comments. It never does any good, and it makes boring reading."
It doesn’t really make sense anyway; if the joke is that the comment is “backed” by the fact that HN users are upvoting it—comment scores aren’t visible to people other than the posters, right? So all we know is that it wasn’t smote into the hidden state.
Etherium is a decentralized model. Its model is an actually inverted hierarchy.
Unless I’m remembering incorrectly, you need a ton of eth to partake in eth PoS voting.