Hacker News Reader

Comment by verdverm | Hacker News Reader

He has a crypto fund that he needs to make returns on, so he has been shilling crypto / web3 / "decentralized" / blockchain hard with his personal brand.

What he doesn't say, that I find obvious, is that open source has been doing decentralized and permissionless innovation for a long time. Many closed source companies are also building APIs and ecosystems around their products which is much better than in the before times.

He's borderline lying saying that web3 is decentralized. Only its expensive data layer is, and maybe a bit of smart contracts, though those are moving towards permissioned calling, because malice... who thought making your code easily callable with arbitrary inputs and subroutines was a good idea... smh

View on HN

A quote I saw the other day:

“Decentralized blockchain, hosted on AWS”

So painful because it's true.

That said, I don't fault AWS for jumping on the bandwagon and making money.

I actually hope Amazon Managed Blockchain picks up so that the truth (and the hoax) emerges faster and we can put all this BS behind us.

The very reason a blockchain needs to exist alongwith PoW, consensus is to tackle the inherent risk of Decentralization and bad actors. Amazon Managed Blockchain is just needless complexity where federated databases would have sufficed and been far more efficient (perf., cost, complexity)

I believe central bank digital currencies are mostly built on managed block chains because the data structure works well. The Hamilton Project between the Boston bank and MIT Digital Currencies group are looking at two implementations for a digital dollar and one of those is a ‘classical’ centralized blockchain

Explain to me why banks would need blockchains over federated databases. All these federated databases have strict audit trails and controls.

Banks have been using these for nearly 50 years and the systems are very well understood. Event Sourcing as a concept and WORM drives as an implementation are mathematically proven to work already and don't need the complexity and overhead of a blockchain in a regulated, audited environment.

Look forward to what I'm missing.

I've been asking this question for the last 5 years every time someone pitches "blockchain for X". I have yet to see a satisfactory answer other than handwavy appeals like "but it's immutable".

This inability of advocates to clearly and concisely explain the pros and cons of their own technology is what constantly sets off my bullshit detector when people talk about blockchain.

Programmable money is the most obvious one. We don’t know each other but say we wanted to bet on a 1v1 video game match. We bet $100 on the outcome of a CoD game. I can write a smart contract that acts as an escrow account that accepts two payees, requires the amount to be $100, and queries the CoD API to check the result of a match ID paying out to the correct payee or pays back in 24 hours if it never happens. This is all within the payment system itself and can be represented in maybe 50 lines of code. Because it’s a blockchain you can see the code for yourself and you know it’s immutable so I can’t rug pull.

My favorite part is when I come along, exploit a bug in the smart contract, and neither of you gets the money, because Code Is Law on the immutable blockchain.

> This inability of advocates to clearly and concisely explain the pros and cons of their own technology is what constantly sets off my bullshit detector when people talk about blockchain.

I don't think you will be disappointed in your approach. I have mentioned some real problems that I think needs to be addressed in this very unpopular comment of mine: https://news.ycombinator.com/item?id=30349821

That quote reminds me of the Bo Burnham song “That Funny Feeling.” It would fit nicely alongside lines like “stunning 8k resolution meditation app.”

You can be "optimistic" with centralized L2 solutions on GCP

I really laughed out loud at this

What's missing from the open source model is the monetization. The incentives for contributing to open source and using open source products are misaligned - this is why it's been so hard to build businesses around it (with some notable exceptions).

What he's arguing is that web3 solves this problem by aligning incentives for creators and maintainers around the same token system.

I disagree. Monetization will ruin open source. The incentives w/o monetization are actually well aligned: the contributors are often the most avid users - whether private or commercial. Adding money into the mix will invite participants whose only interest is earning tokens.

This will probably mean that more OSS code gets written, but - as with the open web - financial incentives will invite bad actors to game the system and the signal to noise ratio in OSS contributions will decrease.

A good case study in this is what happened when Steam tried to add monetization to Skyrim mods[1].

[1] https://steamcommunity.com/games/SteamWorkshop/announcements...

why there needs to be a blockchain involved for this tho? Hard problem is who should pay for what work which a blockchain is not solving

> Hard problem is who should pay for what work which a blockchain is not solving

Yes, this is one of the many hard problems.

I would be 100% behind MA if he used his influence, wealth and power to fund solving these hard problems.

Unfortunately, his and most of current funds are building Managed Blockchains that's just needless complexity where federated databases would have sufficed and been far more efficient (perf., cost, complexity)

That said, Decentralization is the future and will happen with or without him. Unfortunately he's going down in history as one of those who slowed down progress to make monetary gains.

Because you can offload much of the logistic behind it onto the chain. Otherwise you need a legal entity willing to work with banks and auditors. It's what gitcoin does.

> He's borderline lying saying that web3 is decentralized

The "web3" he is funding is only decentralized in his mind a decade into the future. So either he doesn't understand decentralization or is knowingly lying for monetary gain.

> He has a crypto fund that he needs to make returns on, so he has been shilling crypto / web3 / "decentralized" / blockchain hard with his personal brand.

That's unfair. What if the reason he has a crypto fund in the first place is because he genuinely believes it's the future? That seems to be the case with everything he's said for years and how he makes investment decisions.

Here how you prove the fraud:

The hard problem that needs to be solved in a Decentralized blockchain is the risk and effect of bad actors. None of the current funds, including that of MA, addresses this risk.

Instead, all current plays are about building Managed Blockchains that's just needless complexity where federated databases would have sufficed and been far more efficient (perf., cost, complexity)

There's a fine line between vision and orchestrated fraud. It's always been easier to commit fraud and once you're managing billions of other people's money, it's extremely hard to not take the easy path.

Bottomline: This is a pump and dump speculation play. HN is too smart to fall for it.

Which bad actors? One way MA manages at least some of them is by not investing in them, and instead investing in high quality teams. That’s a large part of the value-add of professional VCs, be it in blockchain/web3 or any other domain, that they have the expertise to discern legit teams from scam ones.

I see it as more how do we allocate resources to better causes in society, something blockchain based ecosystems purported to solve, but have this far not, exactly because it is a social problem not a technology one.

Blockchain makes all of these things harder and more complex. I have yet to see something a blockchain can do that we could not with existing technology. The supposedly better incentive structures aren't and are rather the opposite from what I can tell.

> Which bad actors

actors on the blockchain

> That’s a large part of the value-add of professional VCs, be it in blockchain/web3 or any other domain, that they have the expertise to discern legit teams from scam ones.

That's a tiny part of it and as useful to society as hiring away nuclear physicists to work on ad prediction and tiktok recommendation algorithms.

Vetting early stage ventures is actually a large part of the VC job, not a tiny part. It's the entire reason for the accredited investor laws, to make it more difficult for scammers to scam the public.

People keep forgetting that blockchains are decentralized databases. The data is out there on an open database for everybody to use. That’s the feature. The app you build on that can be closed source, private, centralized, decentralized, it’s up to you.

We aren't forgetting it. We just know the database sucks at being a database.

* Transaction volumes are abysmal. Which means I have to build some kind of layer 2 transactional layer over it. (i.e. I have to build a second database to work around the shortcomings of the blockchain database)

* The mechanism of BFT that a blockchain uses is expensive. Proof of work, Proof of Stake, the others. This expense has to escalate in order to retain Fault Tolerance. It is among the least efficient way to achieve fault tolerance that you can imagine. It also exerts a push toward centralization.

* The storage capacity of the database is also terrible See my point above about needing to build a Layer 2 storage mechanism to be at all viable.

All of this is playing out in practice right now in web3 space. Merkle trees make a fantastic version of a public audit log. But blockchain consensus mechanisms all seem to suck. You'd frankly probably be better off going with Raft or something instead. But that would be too centralized for the adherents so they don't. Meanwhile for the exact same reasons that Raft leads to centralized nodes blockchains lead to mining pools at the cost of massive inefficiency.

[EDIT: mis-worded]

> Transaction volumes are abysmal

There are many database engines, some can actually go quite fast.

> The mechanism of BFT that a blockchain uses is expensive. Proof of work, Proof of Stake, the others

Proof of work is not BFT, proof of stake is not always BFT (ouroboros). Proof of stake is not expensive. I’m not sure why you’re saying it forces you towards centralization, these are decentralization technologies, literally.

> The storage capacity of the database is also terrible

How so? It seems like this limitation hasn’t stopped people from creating usecases.

> But blockchain consensus mechanisms all seem to suck. You'd frankly probably be better off going with Raft or something instead

Have you heard of solana, hotstuff, avalanche, algorand, to say a few?

> Meanwhile for the exact same reasons that Raft leads to centralized nodes blockchains lead to mining pools at the cost of massive inefficiency

Mining pools are only a thing in proof of work

If he believes in the technology I imagine he would have some investments in the space. It would be weird if he didn't. Brushing it off as shilling for his fund is not helpful.

If you can't answer the question at face value, you should just not answer it.

Try steelmaning crypto / web3. You should strive to understand thing from the perspective of their biggest advocates. That's a prerequisite to having a solid understanding of the space, unless 100% of people interested or optimistic in the technology and being dishonest.

Reply on news.ycombinator.com