Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Introducing A16Z Crypto

a16zcrypto.com

Post reply on HN
Print view View original

181–190 of 312 posts

Re: Introducing A16Z Crypto

#181

Post by EthanHeilman »

cbhl wrote: post #154
narrator wrote: post #127

Earlier quoted context omitted.

The best thing I've seen blockchains used for, besides currency, is as an integration platform for a non-centralized supply chain. There's really no money in it, since there aren't any gatekeepers, but that's sort of the point. The idea is that the farmer picks his carrots, puts them in a box and sends them off to the carrot juice guy. When he does this he puts a upc code on the side of the carrot box and then puts t…

What? For this system to be truely trustless, the blockchain needs to be owned and maintained by all the people who use it. That means farmer, carrot juice guy, distributor, and retailer need to all have beefy computers with large hard drives to hold the entire blockchain and keep mining it. Sure, it's great that you don't have to trust Amazon or Wal-Mart to trace the links back, but you could do just as well way mor…

>For this system to be truely trustless, the blockchain needs to be owned and maintained by all the people who use it. That means farmer, carrot juice guy, distributor, and retailer need to all have beefy computers with large hard drives to hold the entire blockchain and keep mining it.

Trustless isn't a well defined term so it is hard to disagree directly with what you are saying. However I would argue that most relying parties do not need to have the full state of the system to ensure they aren't cheated. In that they don't need information about other assets to track the assets they care about. For instance the carrot juice guy only needs to validate the chain of custody of the carrots. You could put a merkle proof for each custody change on a QR code on the side of the carrot box. The block headers are 70MB, the chain of custody is probably less than 10KB.

>Sure, it's great that you don't have to trust Amazon or Wal-Mart to trace the links back, but you could do just as well way more cheaply with a well-maintained independent 3rd-party with an AJAX API on top of a Postgres database.

I agree that a running PoW blockchain may be overkill for supply chain tracking. However a BFT distributed database maintained by several trustworthy institutions is probably better a Postgres database. Often blockchain is used as shorthand for "BFT distributed database maintained by several trustworthy institutions".

Re: Introducing A16Z Crypto

#183

Post by mgummelt »

root_axis wrote: post #48

More breathless fawning rhetoric and nebulous jargon, but I guess the target audience is blockchain enthusiasts so that makes sense. At the end of the day it's their money, so what I do I care? I find it frustrating when I hear comparisons of the blockchain to internet and cellphone communication platforms. The potential use cases for the internet and cellphones were immediately obvious (instant remote communication)…

> The potential use cases for the internet and cellphones were immediately obvious (instant remote communication) and available on day one.

"Instant remote communication" is abstract. There was no consensus on the specific applications of the internet in the 70s.

The same holds for blockchain. We know the applications are "trusted computing". We don't know what the specific apps will be.

Re: Introducing A16Z Crypto

#184

Post by endlessvoid94 »

jancsika wrote: post #152
endlessvoid94 wrote: post #29

The skepticism of blockchain technology is borne out of a lack of specific problems it solves. Which happens to also be the bedrock of the conventional startup wisdom -- value derived from solving a problem or exploiting an opportunity. So much of the blockchain hype is focused purely on the technology and the valuation of the tokens / coins. Not the problem it solves. It's distributed trust. That's the problem it so…

I'm not sure you've framed the problem and solution correctly, but here is one obvious societal problem solved by Bitcoin: How does Scihub receive funding to support the service it provides? If you tried to fund it using Paypal, Visa/Mastercard, etc., it's quite likely that those payment systems would have already blocked the payments from reaching the service providers. Because it is substantially more difficult to…

This is an excellent example. Thank you!

Re: Introducing A16Z Crypto

#185

Post by sjeohp »

wpietri wrote: post #132
endlessvoid94 wrote: post #56

Earlier quoted context omitted.

You're right, I pointed to the solution instead of the problem. I'm still an engineer at heart, it seems! What should I have said instead? Centralization of data? Unreliability of single-points-of-failure?

No, those aren't real problems either. Problems are experienced by humans. Humans willing to pay money for solutions. Actual problems are things like "I want to buy this candy bar," or "I want to fly to New York in August and I need to pay for the ticket." Those are problems Bitcoin, as "a purely peer-to-peer version of electronic cash" [1], was intended to solve. It didn't. As far as I can tell, the main problems bl…

"I would like to pay/donate to an organization even though no major credit card will let me."

Re: Introducing A16Z Crypto

#186

Post by kirillseva »

raverbashing wrote: post #98
endlessvoid94 wrote: post #29

The skepticism of blockchain technology is borne out of a lack of specific problems it solves. Which happens to also be the bedrock of the conventional startup wisdom -- value derived from solving a problem or exploiting an opportunity. So much of the blockchain hype is focused purely on the technology and the valuation of the tokens / coins. Not the problem it solves. It's distributed trust. That's the problem it so…

> It's distributed trust. That's the problem it solves. As someone else has pointed out, trust is still not distributed. You still have to trust whoever created the rules and wrote the code (from maliciousness and errors). Especially when that distributed trust is paid/earned with CPU cycles. The incentive is not clear besides periodic speculatory bursts.

> You still have to trust whoever created the rules and wrote the code (from maliciousness and errors).

100% agree with you. Seems like a bunch of people are neglecting this factor or misunderstanding what they are buying into.

Blockchain is not 'distributed trust'. In practice, blockchain centralizes trust. You trust your bitcoin wallet to correctly send money to the addresses that you supply, and correctly generate receiving addresses.

There is only one thing to audit/trust now - the bitcoin protocol and its implementation - instead of studying how modern banking and legal systems work. You trust one thing instead of trusting dozens of middlemen. While this sounds like a non-issue to those who live in "first world" countries like USA and only transact internally, it is a big deal for international transactions and some of the less developed economies.

Re: Introducing A16Z Crypto

#187

Post by vntok »

endlessvoid94 wrote: post #119
dsco wrote: post #110

Earlier quoted context omitted.

I think you're doing blockchain and human ingenuity a disservice by framing the question that way. Many corporations haven't revolved around "solving a problem", but instead making existing solutions more efficient. There were people letting out apartments before AirBnb, and there were taxi cars before Uber - yet we feel that these companies are making our lives easier. Also your question is very much centered around…

Sure. Bitcoin in Venezuela is solving the problem of their existing currency imploding. Anything that performs a function inefficiently can be a problem waiting to be solved. What I’m asking for is a concrete example of a societal problem (something inefficient, costly, painful, or impossible) that blockchain addresses.

Any industry or operations where intermediaries are needed becauses of a lack of trust between first parties can be desintermediated by a distributed trust network:

- Art transactions: registry of origin and property transfers for any object (with authenticated certificate), currently done by lawyers & auction houses (costly, inefficient, painful, near-impossible in some cases)

- International shipping: INCOTERMS, currently handled through various specialized intermediaries (costly, inefficient, painful)

- Registry of operations over the life of a company (creation/address change/acquisitions/foreclosure), currently done by national commerce courts & accredited local newspapers (costly, inefficient, painful)

- Wills: self-authentication of one's will preventing repudiation attempts by third parties, currently done by lawyers and witnesses (costly, inefficient, painful)

- Escrow accounts for service providers (ex freelancers) through multisig wallets (inefficient)

Re: Introducing A16Z Crypto

#188

Post by dd36 »

tomc1985 wrote: post #173
dd36 wrote: post #170

Earlier quoted context omitted.

How does it tamper-proof better than existing solutions?

By reducing the coordination of truth between so many parties to an offline, distributed data-structure, in a way that assures integrity mathematically. As alternative to the constellation of approaches (email, IT integrations, filing cabinets full of records, etc) currently in use. With a global peer-distributed database you can change the risks with your truth provider to risks that the data doesn't update. Dependi…

But you still have to trust the people doing the inspecting and certifying - or whatever. I'd be less worried about them altering information than them altering or substituting actual product.

Re: Introducing A16Z Crypto

#189

Post by plorkyeran »

dd36 wrote: post #153
CPLX wrote: post #148

Earlier quoted context omitted.

Indeed. And more to the point blockchain doesn't do a damm thing to improve trust. If someone in step three tosses the carrots out the window and uses a different batch the blockchain isn't going to get that information. The blockchain in this example doesn't add or reduce trust at all it's just a time stamped database of stuff people said. We have those already.

Bingo. I still don’t follow how blockchain solves current problems. If the problem in your country is a lack of good record keeping, that’s not a technological problem. It’s a political problem. Cheap solutions have existed for centuries. The real question is why haven’t they been adopted? Why would those in power allow blockchain to be adopted but not existing technologies?

In most cases it seems that the existing technologies have been adopted and the people advocating for blockchain-based solutions just aren't aware that it's a solved problem.

Re: Introducing A16Z Crypto

#190

Post by mlthoughts2018 »

foota wrote: post #167
cornholio wrote: post #106

Earlier quoted context omitted.

Except the early adopters don't need to actually participate in the network, just acquire and hoard tokens. There's almost no connection between the profits and the actual advocacy, resources and risks required for the network to succed. So the incentive structure is quite different from, say, the stock market or even a kickstarter. It's more like a tradeable Ponzi with strong incentives to overpromise and overhype i…

You could find a way to associate someone's activity with positive signals, like give them credits for generating positive value in the network. Detecting positive value is difficult though.

But organizing “payment” in the form of a cryptocurrency is no different than just depositing money in their bank account.

In some cases it would be: payment that evades government detection, payment that allows network participation to be anonymous, etc.

The trust issue would not be part of it in these cases, from a business perspective.

I guess I mean that predicating some network-effect-needing business on blockchain, for these reasons, still just seems like hype. Unless the business is fundamentally about anonymously being compensated in some way, then other existing financial institutions solve the problem (with just as much trust in 99.99999% of scenarios) in a far simpler way.

Post reply on HN