Is the DAO going to be DOA? | Hacker News Reader

Is the DAO going to be DOA?

steemit.com

88 pointsOsiris3061 comments

He makes three key points:

1. Funding a startup makes the value of Etherium go down. So holders have an incentive not to fund anything. This would be less of a problem if Etherium was bigger. The currency you're using needs to be much bigger than the biggest individual transactions, or the money system stresses out.

2. "The DAO is just a committee of non-professional voters who have relatively little ability to do proper due diligence."

3. All this complexity solves a non-problem. VCs seldom take the money and run. The things they fund sometimes do, but not the people who put the fund together. On the other hand, taking the money and running is close to standard operating procedure in the crypto-currency world.

1. Well, it's a bit more complex- DAO investors give up direct ownership in ether when they participate in the DAO in exchange for DAO Tokens. You're right that there could be some maligned incentives, but they fall on both sides of the ether valuation to a certain degree.

2. While the non-professional voters have little ability to do proper diligence, they DO have a mechanism for hiring outside parties to do diligence.

3. What complexity? Here's the source code, it's pretty darn simple: (https://github.com/slockit/DAO/blob/master/DAO.sol) Also, I think the DAO solves two problems: First of all, VCs and VC lawyers and Hedge funds have enormous secondary costs that the DAO avoids entirely. Secondly, VCs and fund managers also have enormously maligned incentives which may affect their performance.

(Note: I am still a skeptic of the DAO and did not participate in the experiment)

So how come you ended up not participating? What are the downsides of this experiment?

1. reading the source code, I'm not convinced that it is entirely safe from various soft attacks large shareholders can take to influence the agenda of the fund. (This is hard to predict though... I don't see any full-on bugs, it's more an issue of shortcuts taken in the code for efficiency reasons that may or may not cause problems)

2. Many people are locking all their eth holdings at the same time in this fund- This will prevent them from acting on eth price fluctuations that I might possibly wish to exploit- I want to leave my options open.

3. The idea of creating a company that makes executive decisions by direct democracy is untested and may simply be nonviable.

4. There's some hairy legal issues that could cause liability for shareholders, and I'd rather not have that anxiety in my life.

> The idea of creating a company that makes executive decisions by direct democracy is untested and may simply be nonviable.

This isn't true - it been tested and runs successfully at scale:

https://en.wikipedia.org/wiki/Workplace_democracy#Current_ap...

Two examples with Wikipedia entries that were easy to find are mentioned there:

* https://en.wikipedia.org/wiki/Mondragon_Corporation - turns over ~$16B USD and employed ~74,000 people.

* https://en.wikipedia.org/wiki/Ricardo_Semler - Secmo

1. If the DAO funded a huge startup and paid it out all at once, sure. But I suspect we'll see smaller projects funded and they'll get small recurring payouts.

2. Most of them, sure. Though the large holders are likely wealthy enough to be incentivized to perform due diligence / pay experts to do so.

3. I think the "problem" being solved is that DAOs provide a way for individuals to route around regulatory roadblocks such as accredited investor requirements.

1. is a fundamental problem with all deflationary currencies

Ethereum is fundamentally inflationary now, with no clear indication on how inflationary if and when PoS is implemented.

How? Does it have no mining cap?

Correct. The plan is to move to proof of stake once that's done.

Well some indication, it won't be more inflationary than it is now. That's been the promise since the beginning, was actually written in the legal docs for the crowdsale, and the community is unlikely to run software that violates that expectation.

Not really; it's a fundamental problem with extremely tiny currencies that don't have a ton of trading liquidity to absorb large buy & sell orders.

a) Ethereum is not deflationary b) Deflationariness is completely orthogonal to this problem. This is an issue of liquidity, which would exist in an inflationary or deflationary currency equally.

"The DAO is just a committee of non-professional voters who have relatively little ability to do proper due diligence."

Isn't that true of the stock market too?

As someone who strongly believes that individuals should invest in indexed funds, probably a similar approach should come into play here, where you don't invest a large amount in any one company, but instead, invest a little in lots of companies.

The stock market, of course, has over time had lots of checks and balances applied to it over the years that you're not going to see here, so it's a much more dangerous game.

> Isn't that true of the stock market too?

No, the stock market is a system in which 95% of people have "relatively little ability to do proper due diligence" and 5% of people who know what they're doing (or have private information) and eat everyone else's lunch.

The reason stock pricing is so accurate and predictive when analyzed scientifically is because that 5% has a disproportionate impact on the marginal price of shares.

> 5% of people who know what they're doing (or have private information) and eat everyone else's lunch.

Ah, the myth of the expert investor. Time and again, it's been proven that the experts don't pick stocks better than an index fund. If they did, managed funds would do better than index funds.

http://www.cnbc.com/2015/06/26/index-funds-trounce-actively-...

Your assumption is that the experts would necessarily be fund managers rather than individual investors or people getting away with insider trading.

Nothing he said is contradicted by the mediocre/sub-par performance of recognized experts. In fact it's explained by it:

>that 5% has a disproportionate impact on the marginal price of shares.

>> Ah, the myth of the expert investor. Time and again, it's been proven that the experts don't pick stocks better than an index fund.

"Myth" and "proven" are interesting words to use. Your citation doesn't support the conclusion that experts are proven to perform worse than index funds. What your citation supports is a claim that it is very difficult to beat the market.

I see this zeitgeist expressed on Hacker News very frequently, as though we can't simply look at the numbers and find out if there are experts capable of beating the market consistently.

In point of fact, whether or not a fund performs better than index funds has nothing to do with whether or not it's managed, but rather with how it is managed. You've made a falsifiable claim here, so let's pursue that:

>> If they did, managed funds would do better than index funds.

Pick a timeline, ideally one that seems rigorous enough for consistency. How about 15 years?

The PRPFX and JENSX funds have both beaten the S&P 500's average annual returns over a 15 year timeline. [1]

You might say that 15 years is not long enough to be sure, so let's go with 20 years.

The legendary Renaissance Technologies Medallion fund averaged a whopping 71.8% annual return between 1994 and 2014, which trivializes the S&P 500's ~7.4% during that time. [2]

How about 30 years, you ask? In that case, Baupost Group has enjoyed a 19% average annual return, soundly defeating the S&P 500's ~7.3% average annual return over the same time span. [3]

How many of these can we find?

Bridgewater's PureAlpha had an 18% average annual return between 1991 and 2011 [4].

Farallon Capital had a 13.4% average annual return for 27 years, between 1985 and 2012 [5].

This list goes on. One could look at this and say, "but index funds beat most actively managed funds" but that is not at all the same thing as claiming it is a provable fact that experts cannot beat index funds.

We do not gasp in awe and terror when software developers are capable of actually writing a for loop despite wide reports that most candidates are unable to complete FizzBuzz in an interview. Successful programming is not a "myth." In the same way, we should not act as though it is impossible to beat the market just because it is difficult and requires effort.

---

[1]: http://www.marketwatch.com/story/this-is-how-a-mutual-fund-c...

[2]: http://www.bloomberg.com/news/articles/2015-06-16/how-an-exc...

[3]: http://www.wsj.com/articles/SB100014240527487041677045752584...

[4]: http://nymag.com/news/business/wallstreet/ray-dalio-2011-4/i...

[5]: http://www.forbes.com/sites/kerryadolan/2012/10/23/californi...

As other posters have said, I did not state that this 5% represents people deemed as "experts" or that there is any objective way of determining which market participants are in that 5%- Just checking performance is not enough because of many false positives due to the large random element and the overwhelming number of less-skilled investors (and because skilled investors have an interest in remaining non-public)

In other words, I personally believe in weak forms of EMH, since I believe having an "edge" is very very hard but not impossible. I base this belief on my 15 years of experience investing in the stock market. Without a doubt, the default assumption for any analysis of the markets should be "no one can outperform index funds" but I think there is evidence that reality deviates slightly from this default assumption.

(1) is also a short-term problem, just depressing the market while the ether is cashed out. Longer-term investors benefit by funding businesses from which they can profit. On top of that, the value of ether goes up if there are more applications, hence more people using ether in some kind of commerce.

In response to 3, I think the main problem it solves is that most people don't have access to invest in VC funds, and the DAO represents a way of democratizing that access.

I'm still pretty sure that the thing will be hacked before long, or will have critical bugs that will necessitate complex negotiations to fix.

It's not just the size of the code that is the problem, although that alone practically guarantees that there will be bugs. Ethereum has a couple of unique properties that can cause all sorts of chaos in ways that 'normal' language developers never need worry about.

First of all, code requires 'gas' to run. If there isn't enough gas, then the execution stops. But critically, some but not all of the code's actions are rolled back. Transfers from one party to another still take place. You can imagine a situation where an attacker runs a contract with just enough gas to make the code transfer some money but halt before it records the payout, allowing the withdrawal to be repeated.

Secondly, there's a stack limit for code, again triggering an exception and a partial rollback. Developers are used to having to reason about stack overflows in normal programming situations, e.g. a recursive algorithm. But Ethereum's stack has more problems. It allows attackers to control the available stack space before your code gets run. e.g. an evil-doer could use up almost all of the stack, then call your contract. If the stack runs out at a select inopportune moment, there is potential for abuse.

> But critically, some but not all of the code's actions are rolled back.

This is not true - the actions within the contract are rolled back. You do, however, lose the ether committed to paying the gas cost (cost of execution).

https://github.com/ethereum/wiki/wiki/Subtleties says:

If a transaction triggers an exception, then:

* The value transfer from sender to recipient still takes place

* The fee transfer from sender to miner still takes place

* The entire STARTGAS is added to gas_used

* All other execution is reverted

Perhaps I am reading that wrong, the 'value transfer' is unclear now that you mention the gas cost. I read it as talking about actual transfers. Can anyone confirm?

I don't have a document on hand to support my claim, but it's alluded to in the header of the document you linked:

THIS DOCUMENT HAS BEEN LAST UPDATED IN MARCH 2015 AND SOME SUBTLETIES HAVE CHANGED SINCE, ESPECIALLY CONCERNING REVERTING ETHER TRANSFERS IN CASE OF OUT-OF-GAS.

Based on my understanding, your bullet points are valid excluding the first one.

More likely is sometime the system will be abused to enrich an individual or group who has authority over a part of it. Maybe the developer, maybe a software distributer, maybe someone who writes up it's contracts.

Somewhere down the line someone with access will crush it to enrich themselves.

Yeah, totally. I've been cooking up ways to do this specifically to the DAO, and I'm a pretty nice guy. There's no way this thing won't have brilliant minds and tons of resources working to break it.

Do you have any more information about the stack overflow attack? It's quite concerning to me and searching google for "ethereum stack overflow" is as useless as you might expect.

Sorry, I can't find the good example which I read recently, which detailed the problem well. The best source I have right now is:

https://github.com/ethereum/wiki/wiki/Subtleties

Which details many of the corner cases of the virtual machine, including the stack limits - but it's just a single line mention.

Even if TheDAO "fails", we'll have learnt something very important. We'll have found a DAO configuration that doesn't work. That knowledge will mean that future DAOs will have a higher chance of "succeeding".

You're making the critically flawed assumption that the space of all DAOs is a finite set!

Elucidate.

He writes: "It was because of these articles that Vitalik Buterin, one of the founders of The DAO, started exploring the concepts in a three part series."

Vitalik is in no way a founder of The DAO. He's a "curator" which only means he can whitelist wallet addresses that can receive funds. He had no hand in it's creation or in the white paper that inspired it.

Getting such a basic fact wrong makes me doubt his understanding in general.

I wouldn't be so quick to judge. Most of this Ethereum stuff even confuses me half the time and I've been following BTC closely for a long time now. I'm not surprised the journo mixed up him being a co-founder of Ethereum, which underpins the tech behind DAO, and being a co-founder of The DAO.

Off-topic: This type of nerd superiority complex comes up in every article written for a non-technical audience on a tech subject. Can't we ever just point out the error and move on, without turning it into an attack? It'd do a lot to improve the civility around here.

> Perhaps even more interesting with The DAO, once you vote for something you are no longer allowed to split your ETH out and form a new DAO. This means that you have much to lose by voting and much to gain by not voting.

I'm not an expert on the DAO but I don't think this is correct. I think that voting on a proposal prevents you from splitting the DAO for a certain period of time, but not forever.

Looking at various discussions happening around the web on this blog post, it's alarming how many people are disagreeing about this point. Even the author admits that he's not clear on it:

> The DAO is complicated and I admit that I am not sure I fully understand how, when, and where ETH can be split relative to payouts and rewards.

How exactly has this project accumulated over $100 million in funding, when there's no real understanding among the community about how the DAO actually works on a day-to-day basis? This telegraphs the sickly condition of the DAO investing community, in my opinion. People seem to have bought DAO tokens to speculate, not to vote. That's a problem for the DAO because the only way for it to take more ETH into the system is for projects to be profitable.

You're right! It's just for the duration of the voting period (two weeks) of the proposal that you voted on, to avoid people being able to vote for a proposal without the outcome having any potential to affect them (because they split away).

This massively undermines the OP's article criticism.

I'm surprised that this is posted to a (mostly) technical forum, and no one else assumed that it was talking about data access objects.

TL;DR - It's really about Bitcoin-ish stuff

This is the first time I've read content on Steem, and out of curiosity I went to check out the site. Its a pretty impressive alternative to reddit. Here's the whitepaper: https://steem.io/SteemWhitePaper.pdf

Time will tell. While he makes some interesting points, I think he misses a core distinction on this point "BitShares had all of the tools, the talent, and the money... [and failed]."

I had never heard of "BitShares" before, and I'm an active researcher in this area. The DAO made it to the front page of the New York Times, and not because of its amazing coding. Sometimes life is just a popularity contest, or nepotism (landing Buterin surely helped, since he is proving to be a masterful marketer). If the DAO succeeds where BitShares failed, it is probably due to completely _social_ reasons.

> I had never heard of "BitShares" before, and I'm an active researcher in this area.

How long have you been researching this area? I've never invested or participated in the project, but Bitshares or one of the other Larimer/Hoskins DAC projects/renames/reincarnations filled most of the top search results for "Distributed Autonomous Corporation" and "Distributed Autonomous Organization" for several years, up to this past year (indeed, I'm surprised that Ethereum's success has pushed out almost all trace of the project from the top search results).

They promoted it pretty heavily, so you either are very new to the field or you didn't spend much time looking around for actual implementations of the technology.

In fact, to the best of recollection, Larimer actually coined the phrase "distributed autonomous corporation". And Larimer and Bitshares are even featured prominently in the Wikipedia article on "Distributed Autonomous Corporations" (and it's not a new addition to the WP article).

Here's an earlier piece from 2013 discussing DACs and Larimer by none other than Vitalik Buterin himself, back when he was a writer:

https://bitcoinmagazine.com/articles/bootstrapping-a-decentr...

Also, most of Larimer's criticisms in this piece address problems with The DAO's incentives. If the DAO gets those wrong, it won't matter how well it's marketed.

That said, I do agree that these kinds of projects frequently fail or succeed for social, not technological, reasons (indeed, that's one of the main thrusts of the article).

Well, in my defense, I have not been actively looking for actual implementations (I'm an academic researcher), so I'll have to concede naivety. But, I think my point still stands, and I think we agree about the capricious nature of these things.

"Sometimes"... The older I get, the more I realize that everything comes down to a popularity contest. The only thing that doesn't abide by that law is physics and warfare. And warfare is a lot of physics.

> Who is right? Well odds are in favor of it being wasted as 9 in 10 startups fail.

This commonly cited 90% startup failure rate is actually based on a 7yr timescale (if I'm not mistaken). I remember it being an example of a common misconception in the book "The Illusions of Entrepreneurship" [1]. Just because a small company eventually fails, it doesn't mean the investors didn't get any amount of ROI out of it.

For example: It's possible a company makes 10's of millions in revenue but fails from cash flow issues after a long-term of being operational.

So this isn't a good way to measure return potential when investing in companies. Unlike VC-style investments, not all businesses (even with $1 million capital) are billion dollar or nothing businesses.

Not everything The DAO invests in has to be 100x return to be viable (unless they want to be VCs). So this changes the math quite a bit.

[1] http://www.amazon.com/Illusions-Entrepreneurship-Costly-Entr...

> With an initial quorum of 20% it will be very challenging to get enough agreement, especially with the downsides associated with actually voting.

20% of tokens are held by the top 12 account holders. Getting 12 people (perhaps fewer, if someone had multiple wallets) who each invested millions of dollars into the DAO to vote doesn't seem that hard.

Presumably those who have invested millions are going to be even more aware of the downsides to voting (losing the ability to split) than most.

The DAO represents about 13% of all Ethereum in existence. Considering over 75% of Ethereum was part of a premine crowd sale for a price between 2,000 - 1,337 BTC per ETH it's a reasonable assumption that the DAO was mostly founded with severely discounted ETH, and doesn't really represent +100M USD investment.

Anyone could have sold their ETH at around USD $8 per 1 ETH just prior to joining the Dao. I don't think "it's not real money" applies.

Anyone could have sold 1 ETH for 8 USD, maybe, but that is very different from selling all ETH in a large account.

Do you think they could have raised ~150 million USD by selling 12% of the existing ETH? I suspect $150 million couldn't be raised by selling all the ETH.

It's disingenuous to take the current ETH-USD conversion rate based on a relatively low trading volume and imply that's what the DAO is worth, especially when so much ETH was bought so cheaply.

It's likely that at least some of those large holders are exchanges like Poloniex that allow you to purchase DAO tokens directly on the exchange.

Forgive me, but the website of DAO reminds me a bit of this one -> http://zombo.com/ :)

I know ethereum and "The DAO" sounds like a ton of mumbo-jumbo, but I assure you at its heart everything is actually very straight forward- For instance, the DAO is mainly just this small piece of computer code that is very readable: https://github.com/slockit/DAO/blob/master/DAO.sol

(Whether the DAO is good idea though is another matter entirely)

I thought the application for DAOs was far broader than setting up VC type corporations.

Another problem in general with blockchain technologies is that the power elite will eventually get around to trying to squash them. Anything that facilitates local businesses keeping value inside communities, businesses that don't give up value to the centralized monetary system, etc. will probably be targets.

here's a typical deleware court lawsuit http://courts.state.de.us/opinions/download.aspx?ID=207080 skim thru that and now imagine that same disagreement over a DAO company vs. deleware court. Power players need to learn all sorts of different rules.

I read through the whole thing, and honestly, that whole thing seemed really straightforward.

What would you see the issue as?

I guess I'm a DAO novice, but when I heard about an org with 150 million and it's NOT a deleware company, I wonder how the power dynamic changes when things go wrong. The company in that pdf had a huge disagreement over who was running that company. Board members were forced out. New CEO placed in. In DOA what takes the place of the deleware court when there is this type of fight?

For TheDAO itself, there's no possibility of such a dispute or resolution. There's a little bit of code that runs everything, and what it does can't be overridden. With this particular code, you get a vote in proportion to how many shares you own.

The companies employed by TheDAO will resolve things like any other company.

this explains it well: https://news.ycombinator.com/item?id=11787475

All this assumes that people who invested in DAO are random blokes off the street which they are demonstrably not. Have a little faith.

DAO was dead as soon as ODBC entered the room.

Reply on news.ycombinator.com