Why an Ethereum based DAO won't outcompete us

The enthusiasm for Ethereum is quite naive. Here’s my impersonation of an Ethereum fan boy:

Ethereum is like a VPS, except it costs 10,000 times as much, only uses JavaScript, and can’t communicate with the world that exists outside the blockchain.

If you think Bitcoin has scalability problems, it is a model of efficiency compared to Ethereum. People talk about smart contracts being developed on it. Automated contract execution is and will be a thing, but most people haven’t thought through how extraordinarily difficult they are to implement on a blockchain. This is because a blockchain is its own isolated world that knows nothing about the rest of the world.

Consider the ability of Ethereum to parse if statement. Let’s say we want to use the Ethereum blockchain to execute a smart contract that will allow two people to bet on the outcome of a major political election. This is actually a pretty straight forward smart contract (when described in human language), but Ethereum can’t do it. Among the multitude of problems is even determining who won. The entire world will agree about who won, but the blockchain will remain unaware. Think about how this swarm of redundant computers can evaluate the statement “if candidate A wins”. It can’t resolve that if statement to true or false. How would it, even if every human on the planet knows the answer? It could be modified to resolve that statement, but this is impractical. Ethereum miners would all have to agree on a way to determine who won. They could agree to have CNN publish the results using a predictable API so that each miner can call the CNN API to determine who won. Given that Bitcoin miners can’t even agree to increase the transaction capacity of Bitcoin, it is silly to think Ethereum miners are going to standardise on an API to determine the outcome of specific smart contracts. Even if they did, the data source is still centralised and constitutes a single point of failure. The entire approach is simply impractical.

I pity anyone who attempts to do what we have done with a Peershare on Ethereum. Any such attempt will fail regardless of the quantity of funding involved.

Ethereum has done well because of the hype it has offered and also because it offers a solution to the currently very limited transaction capacity of Bitcoin. Ethereum scales better for simple transactions that transfer value merely because it doesn’t have its transaction capacity capped at 1 MB per 10 minutes. For Ethereum, hype + transaction capacity = success.

Proper functioning as a DAO requires proof of stake and a dedicated blockchain. Dash, which I consider to be the most developed DAO besides NuBits and B&C because they adopted a version of our custodial grant system as best their architecture allows, has confirmed this by adding a proof of stake system on top of their proof of work system using masternodes to develop their DAO functions. There is room to evolve, and Ethereum has a lot of money to do it. But the current structure and foundation are absolutely abysmal for the purpose of supporting DAOs.

5 Likes

You’ve must have misunderstood me if you have tailored that text as a response to (me) the Ethereum fanboy.
I know that Ethereum is empty hype, I know that it will end up like niche product. But that was not my point, I have simply stated that Nu should have reacted in advance to upcoming Ethereum projects. Do not fool yourself into thinking that Nu will survive in present state long enough to see the death of Ethereum hype. The hype is just getting started. Don’t you see how well do they do marketing? It will just be a “DAO” after “DAO” until they suck the money out of every naive soul out there. Don’t you see they have taken over acronym “DAO”?

Also, if Peershares are so good why did not anyone else utilized for their project but yourself?

P.S.

Nice try with topic name.

The marketing has been spectacular, I agree. This is an accurate characterisation of the situation.

Dash has been working to evolve to function more like a Peershare, as evidenced by their recent adoption of our custodial grant system. It doesn’t work as well as ours because it wasn’t a part of the original design. It is, however, a validation that the usefulness of Peershares are being recognised. Peershares are difficult to understand. The only way most people will be able to recognise the usefulness of them is to see them work. That takes years, just as it does with any startup.

We have discussed recently about how the crypto community is mostly hype and scams. Because Peershares typically require ongoing funding for their operations just as any business would, they are poor vehicles for making quick money using scams and hype.

2 Likes

Scam+hype coins can become very successful (long term) if million users use them and evolve them.
On the other hand, the most perfect of the coins cannot be a success, without users!
I am not referring to any specific coin, just generally speaking :wink:

1 Like

As is typical for startups that are 2 years old, we have not had much commercial success to date.

Let’s think about the issue of “survival” and what that actually means for our network. By not survive, does that mean our blockchain will stop stacking blocks on top of one another? That will only happen if no one, and I mean no one literally, cares about the network. Where we are the 4th most active community in peer-to-peer finance and growing, that seems unlikely.

If you are concerned about the survival of the network in the case that our product disappoints consumers, consider all the blockchains that have been running for years with no product at all.

The community is our most valuable asset. As long as people keep working for and caring about the network, I am not concerned about its “survival”.

Prospering is another matter altogether, and it is more difficult than mere survival. We have accomplished incredible things here. We haven’t been financially rewarded for those accomplishments yet. If we keep growing our hard working, non-hyping community, our prospects are very good. We are more like a regular non-blockchain startup than like a crypto scam in this regard. It takes time and development.

The way the system is designed we won’t ever require hype to have the NuShare price go the moon, because we don’t need any sort of broad adoption of NuShares. In fact, if we never gain another shareholder but do stimulate vigorous demand for NuBits, NuShares will go to the moon via NuShare buybacks. This fact is not well understood outside very involved community members, because it is a situation that is very different from any other crypto.

1 Like

By saying “Dash has been trying to evolve something like Peershare”, you mean having a votable coinbase tx? If so, that is awesome feature but not a selling point of a platform. DAO is much more than that.

They are not difficult to understand, it is a blockchain based token ownership tracker with two special functions: votable coinbase tx and dividend payout via external value tokens. That is about it, motions are not part of original design right? Also motions are just for show as they have no real value. It is glorified poll that considers stake of a shareholder. They have no action.

I do not really know what to comment on this. You should look wider than that.


[quote=“JordanLee, post:1, topic:4022”]
Let’s say we want to use the Ethereum blockchain to execute a smart contract that will allow two people to bet on the outcome of a major political election. This is actually a pretty straight forward smart contract (when described in human language), but Ethereum can’t do it. Among the multitude of problems is even determining who won. The entire world will agree about who won, but the blockchain will remain unaware.[/quote]

Let me get back to this. Are you sure a Ethereum can not run a script which will parse a news sites and return a result to a function? As Ethereum is turing complete this is theoretically feasible. Also, maybe in the future you will have an API for “who is current president of XY?” so a smart contract can just pull json and tell which is what. In this case such script would be a weak AI in service of DAO, but without weakness of being centralized.

About a year ago I’ve started writing a paper on DAO’s in general, I have never published but let me just show few definitions for sake of this debate:

Decentralized organization
Abbreviated “DO”

The idea of a decentralized organization takes the same concept of an organization, and decentralizes it. Instead of a hierarchical structure managed by humans interacting in person and controlling it via the set of agreed rules and laws, a decentralized organization involves a set of humans interacting with each other according to a protocol specified in code, and enforced by the public ledger, ie. the blockchain.

Decentralized autonomous organization
Abbreviated (DAO)

The main difference between a DA and a DAO is that a DAO has internal capital; that is, a DAO contains some kind of internal property that is valuable in some way.

Decentralized Autonomous Corporation
abbreviated “DAC”

A Decentralized Autonomous Corporation is a decentralized network composed of WeakAI autonomous agents which perform repetitive tasks and human agents which participate in decision making through democratic and meritocratic process. DACs are basically a subclass of DAO’s that pays dividends on purchasable and tradeable shares which potentially entitle their holders to continual receipts based on the DAC’s ability to generate profit.

Now in this theoretical case where we use Ethereum to host a DAO, where you have a:

  • infinite multisig wallet (owned by all shareholders),
  • you can employ trusted and decentralized Weak AI’s for external data input,

In this case motions have direct effect on shared wallet (no need for trusted executor) or on shared and commonly owned data.
Is this not closer to DAO than what Peershares based implementation is?

Motions were a part of the original design first circulated in April 2014. They are very important. It is isn’t trivial to design a motion system that is effective. Many people are currently suggesting The DAO place a temporary moratorium on voting (for motion-like proposals) because the system is broken (remember, that’s on the Ethereum blockchain). Your comment indicates even you, smart, capable and involved as you are, are having trouble understanding Peershares.

I’m not sure whether Ethereum can support use of the HTTP protocol within its scripting language. Let’s assume that it does support it, because I agree with you that it is possible to do so, but not very practical to actually use HTTP in Ethereum scripts, as I will explain.

Returning to the example of using a smart contract to bet on the outcome of a political election, as I had explained, it is possible for CNN to create an API specific to the purpose. However, weak AI parsing standard news sites just won’t work. What is this weak AI and how does that fit into the system?

With thousands or tens of thousands of nodes what happens when the script is run? For one thing, they might cause the servers hosting the API to become unresponsive for some time because it is almost like a coordinated DDOS attack in some regards, because every Ethereum node is suddenly calling the API. This means results are not returned in a timely manner. Using threading you can have nodes wait patiently without adversely impacting other tasks, but you can see how this is quickly becoming a scalability nightmare far worse than Bitcoin. What timeout value should you have when a script is waiting for a response? 30 seconds? What if this DDOS like attack moved response times past 30 seconds? The script can’t evaluate. Blockchains can only utilise processes that are quite reliable, otherwise consensus cannot be maintained. How many of these slow running scripts could an Ethereum node handle per second? I don’t know and it depends on hardware specs, but I would be surprised if the majority of nodes can handle more than 100 per second. I wouldn’t be surprised if the majority could only handle 5.

Use of APIs like this are also centralised and trusted systems. Now the IT staff at CNN have a financial motive to place an inexpensive bet on the apparent loser, then corrupt the results returned. Even changing the result of the API call for a few seconds could be very lucrative. When you make very reliable and predictable systems (like blockchains) dependent on less reliable systems, like a theoretical CNN API, they become unreliable. Not a good idea.

5 Likes

Very true and relevant. However, proof of work systems that don’t have a well defined and active leader such as Satoshi or Evan Duffield of Dash can’t evolve very much at all, which we have seen with Bitcoin. I suspect Ethereum is now so distributed that change will be quite difficult.

I am arguing that motion system in Nu is just a glorified opinion poll. It has no effect on the DAO it self.

But anyway maybe you should tell me what Peershares are if I am wrong with this?

I have just proposed an example how power of turing complete scripting can be utilize to widen the reach of DAO and make it more autonomous. It is not important that >over HTTP< scraping or whatever can actually be done or not.
What is important is that you can link almost any external signal to shared wallet without middle man. If I understood correctly shared wallet (like in TheDAO) is just a another smart contract on Ethereum and it can use return from another function as input signal and act on it. This is very important step in evolution of DAO concept as it cuts out the middle man between shareholder decision and execution.

I am not saying that what can be done with Ethereum is perfect, only that is closer to what DAO should be than what Peershares offer. And in world where Ethereum and Peershares exist I argue that more people will try to build on Ethereum.

Weak AI: https://en.wikipedia.org/wiki/Weak_AI (It was also linked above)

Basically a bot which serves the DAO. In mentioned paper I argue that those are very important for DAO as they can be used to execute simple tasks effectively. I’ve been watching you guys and I remember that you have too argued that more automation is needed. And that is what those Weak AI’s are.

Oh, a lot more automation is needed here.
But we need at least one more step till we can really do it. We need more processes, thresholds, triggers, etc.
A lot is based on gut feeling and hopefully appropriate discretion.

Would you have known that you can draw a demarcation line at 5% buyside offset? We didn’t know either, but it worked remarkably well so far. But we won’t be fast enough in adjusting the offset in case the market changes.
Market awareness - isn’t that a kind of AI? - is required.

1 Like

single point of failure? dictator?
I don’t like that.
The reason i like NU.

1 Like

What about a benevolent dictator :stuck_out_tongue_winking_eye:

I find PoS very nice. I like the idea that PoS considers those who have a stake in the system, whereas PoW rather has wageworkers that might work here today and tomorrow somewhere else.
PoS has more to do with loyalty than PoW.

4 Likes

Well real PoS is in fact proof of shares, so it s financial incentive beyond loyalty.

2 Likes

You guys said it all!

I’m tired of those endless discussions about PoW vs PoS that go on and on @ bitcointalk. A never ending PoW fan boys against PoS fan boys small talk…

The truth is: PoS is as near as we can get to a decentralized “Crypto Venture Capital” model and system promoted by “Crypto Angel Investors”.

i.e. with PoS you’re really putting your mouth where your coins are!

5 Likes

There is at least one

It is possible to handle answering questions in a crowdsourced manner
entirely within the blockchain, thus eliminating the risk of a
single corrupt third party, or limited network resources allocated to the
oracle.

You could create a motion: “Did X do Y”, then the first one thousand (or some other number) of votes between blocks A and B are used to calculate the answer. If some threshold (say 75%) of the votes are positive and at least a certain minimum number of votes are cast in that timeframe then then the question is answered as true. If less than the minimum number of votes were cast then it is considered undetermined, otherwise the question is answered as false.

After the criteria are met (1k votes cast or block B is passed) then a transaction is created codifying the final result and paying the people who voted. This transaction would also be used by the smart contract to determine the answer.

I think this is similar to, but a bit different from Cybnate’s NuOracle post.

I must confess that jordan is correct and saying it all in this thread –
Ethereum, in its current implementation will evidently and absolutely, by construction fail, There is absolutely no doubt about that.

But this is not really important.
What matters is to do topnotch engineering, topnotch marketing and topnotch recruiting for Nu.

Any data for this, btw?:

1 Like

Etherem is one of the have-it-all web portals – Yahoo!, MSN, Lycos, Excite of 1997. Don’t know if it can survive the crypto-bubble that is yet to come.

1 Like

OK, let us take a look at: access #3 CALLDATASIZE in https://live.ether.camp/transaction/4ed58ea866adc0/vmtrace#0
I would like the proof that each node “has run” operation #3.
Satoshi has certainly showed us a good means: take your nodes, and make them create a chain of time-stamped hashes of the stuff.

1 Like

Please join the conversation:

Anyone
has the proof of this: “every ethereum node executes the code”?
More concretely, anyone has any proof that every node has parsed and
executed " if (msg.value < 1 ether) { msg.sender.send(msg.value);
return;" in the line above?

As per some reference, Bitcoin is giving proof that every node has
verified the tx (one type of code execution in itself). It is
constructing it every 10min and in that sense we can see bitcoin as a
simplistic instruction machine, but does not cheat at least.