The DAO is effectively on hold to decide whether flaws exist and how to fix them | Hacker News Reader

The DAO is effectively on hold to decide whether flaws exist and how to fix them

wired.com

88 pointsbpierre54 comments

My 2 cents:

I started and (briefly) ran the sole Ethereum-focused meetup in the Atlanta area. I shut down shop due to lack of interest in the meetup. ATL has a pretty decent crypto scene. Bitpay and Storj are located here and the local Bitcoin meetup has ~500 members.

IMO Ethereum's problem is complexity and a lack of anyone in the inner circle who has the skills or desire to bring the complexity down to a place where the cost/benefit begins to makes sense (from a developer's standpoint).

Decentralization and trustless-ness are better means than they are ends. The platform's learning curve is steep and long. Outside of fintech and a few other niche applications, it hardly seems worth hiking up that curve just to say "yeah, but we're decentralized."

I understand ETh is an experiment and I commend the core dev team for putting in so much hard work, but a platform is only as good as what people build with it. Unless they get some people in the inner circle who can focus on evangelizing the tech to regular devs (not just crypto-enthusiasts) it will fail.

EDIT: let me clarify when I say the core team lack skills, I'm clearly not talking about technical skills. I'm talk about API design and marketing to developers. I know we're not supposed to comment on down voting but at least comment something saying what you disagree with. I don't feel like I am being un-constructive or unnecessarily harsh or combative with my comment.

That runs counter to what I have been hearing/reading. AFAIK the dev community is the fastest growing in the cryptocurrency space.

With respect, I don't think that's saying much. ETH is pretty much the only crypto platform [anyone has ever heard of] specifically built to support on-chain app development so they kind of win "the fastest growing" category by default, don't they?

Did any of them produce a killer app though

I've been trying to follow things pretty closely too, and to me it looks like that is (partially) the next step. Disclosure: I'm definitely a supporter, and think it's a cool project.

Right now it seems like the big things are building out the network and tools, to make the next stage easier. Their release schedule states that the next major release (what they call Metropolis) will happen when they have 1.0 stable release of the native wallet/browser. A couple of IDEs are also in the works, built to work with developing smart contracts for Ethereum. Even Microsoft has added Solidity (essentially a DSL for Ethereum, for those who don't know) support to the latest preview of Visual Studio.

Also, from everything I've been seeing so far is that the dev community forming around Ethereum is very passionate and is grow faster than most expected. If it continues at this pace, and the tools continue to improve with it, Ethereum could continue to become a very robust platform.

All this said, things could still not work out and the project could fail. I would hope it doesn't, because it really is a cool experiment thus far.

I've been working in the Ethereum space for a year and a half, so I know my perspective is warped, but I really don't think the learning curve is that bad any more. Solidity isn't a very complicated language and has decent documentation[0], and Truffle[1] and TestRPC[2] can get you writing automated tests against a simulated blockchain more or less instantly.

It's still relatively early days, and there is a lot more work to do in making this stuff as friendly as possible, but progress has been pretty rapid.

[0] http://solidity.readthedocs.io/en/latest/ [1] https://github.com/ConsenSys/truffle [2] https://github.com/ethereumjs/testrpc

I specifically mentioned the platform, not solidity. It's pretty hard to be proficient with solidity if you are ignorant of the many abstractions, rules and constraints of the ethereum platform as the language is (intentionally) very tightly coupled to the smart contract abstraction and it's low level nature make it imperative that one understands a great deal about the protocol (bit lengths of various classes of data for example) before you can really get up and running.

Anything becomes easy if you look at it long enough. The major issue ethereum is going to have is making a case that there's a real-world value add relative to centralized, traditional architecture before many people will be willing to stare at it long enough for the easiness to set in.

I mostly disagree. You definitely need to know the basic abstractions and concepts, like blocks, transactions, and addresses. But they're not particularly harder than concepts like inheritance or pointers, and way easier than something like monads.

I can't think of any reason you'd need to know "bit lengths of various classes of data" for normal development.

> The major issue ethereum is going to have is making a case that there's a real-world value add relative to centralized, traditional architecture before many people will be willing to stare at it long enough for the easiness to set in.

I definitely agree with that.

I generally agree with those sentiments, and I've been working lately on some developer education on gotchas and best practices for devs using solidity, in particular.

There is a huge need for a focused consumer and developer-oriented rework on the marketing and ui side for ethereum. It's very, very confusing, even for highly technical people.

That said, things are much easier to work with than a year ago. Communication people would still help a lot.

On the plus side, the community is really great. People are nice, and motivated. There is a huge amount of pending work to get Ethereum ready for real prime time, and people do seem to be kicking in.

I co-founded and ran the Silicon Valley Ethereum meetup for some time. There was lots of interest right after the whitepaper up through the crowdsale, then a waning of interest in the details, and now resurgence of interest. Like the Bitcoin meetups there does seem to be some correlation between price of the asset and number of attendees but that is unavoidable IMHO.

As far as marketing, Tual was the community lead for a bit and he left to do the DAO which has broken crowdfunding records so I don't think you can say he's been unsuccessful. As to the ultimate success of the developer ecosystem, I'm not sure but I used to track metrics like github stars quite rigorously and Ethereum was doing better than any other crypto project.

I think when Mist is officially released you will see a lot of Dapps that will come out and will make using Ethereum fun and easy... Here is a sneak preview: https://www.youtube.com/watch?v=qghBWznAvoM&index=5&list=PLx...

Matt Levine at Bloomberg has some insightful commentary on the DAO from a legal/finance perspective:

* https://www.bloomberg.com/view/articles/2016-05-17/blockchai...

* https://www.bloomberg.com/view/articles/2016-05-23/bailout-f...

* https://www.bloomberg.com/view/articles/2016-05-31/complianc...

I found this helpful. From the first linked-to article, regarding the DAO:

> All of this is trivially replicable using old-fashioned governance structures. Berkshire Hathaway would look a bit like the DAO if, instead of leaving things to Warren Buffett, all of its shareholders got to vote on every investment decision. When I put it like that you can probably see why this structure has not been super popular so far. But who knows! Maybe the DAOers will be really good at picking winners, and putting them into smart contracts.

What's missing is that this DAO structure has a self-evolving nature and can modify its own decision making structure, something that no "old-fashioned" structure can do.

To extend the (simplistic) Berkshire Hathaway analogy, why couldn't Warren Buffet change the structure to shareholder voting? If he did, why couldn't the shareholders then vote to change the structure to something else. That would be self evolving in the same sense that the DAO is "self evolving".

The actual difference is the concensus rules and that everything is public in a shared chain of transactions / events. It means anybody can interact with it on equal grounds.

Last time I checked, I was able to vote in my shareholder's meeting when I only owned a few shares in a company. (I mean yeah... I know GOOGL exists, but most people seem to buy GOOG and pay the slight premium for the voting rights)

Most people just don't care about voting rights in the companies they own (which is why GOOGL exists)

http://investorplace.com/2014/07/goog-google-stock-split/

But that's still not on equal grounds and fully public - what the blockchain systems implement in public using the computers of the collective is instead implemented by humans with imperfect transparency and accountability.

And that was the entire point. The capabilities aren't different, it is the guarantees around its behavior that are.

> But that's still not on equal grounds and fully public - what the blockchain systems implement in public using the computers of the collective is instead implemented by humans with imperfect transparency and accountability.

Who are your major coinvestors in the DAO?

Did any of your coinvestors perform an inside-trade to bet against the company through either a put-option or a short sell?

Because all inside trades of public corporations are publicly documented. Can you really trust your fellow DAO "investors" when you aren't even sure if they're betting on your side?

Nah man, I think classical companies are more transparent. Its very important to keep in mind the insider trades.

Here's all of the insider trades from Microsoft in the past two years: http://finance.yahoo.com/q/it?s=MSFT

Now where's the equivalent information on the DAO?

-------------

The DAO is a baby experiment. The first of its kind: a digitally run company. I'll give it credit for that. Unfortunately, it still has a LOT to learn about political trust structures.

No one gives a crap about whether or not board policy was followed. What people care about are inside trades and conflicts of interest, of which the DAO doesn't seem to offer any solace.

Those trades are still just the public ones - a trade doesn't need to have an official paper trail involving every intermediate buyer.

Yeah they do, if you don't feel like going to jail.

https://www.theguardian.com/business/2014/sep/08/matthew-mar...

So once again, I ask what is the mechanism that allows you to trust your fellow DAO investors? What are the protections the DAO has against the classic Ponzi scheme or Pyramid Scheme?

Again: the DAO is a baby in a young field. I'm sure eventually a digital scheme will be invented that addresses these concerns. But from my understanding, the DAO is not a good digital scheme yet. Maybe a future project will be better at addressing real concerns and solving real issues. Or maybe the DAO will be able to self-mutate itself to solve the issues that truly do arise.

The very basis of parliamentary procedure is to create a self-changing human-organization / structure.

https://en.wikipedia.org/wiki/Robert's_Rules_of_Order

The details of parliamentary procedure change from organization-to-organization... but the goals are always the same.

1. To create decisions

2. To self-evolve the decision-making structure over time

Errr, except this is the whole point of these structures.

DAO makes it easier to administer/more automatic, but it is not new in what it is trying to do, and has a long history that hopefully folks involved are aware of.

Do we know for sure that a single individual / investment group doesn't hold a majority of shares and is just a normal VC with the added benefit of random people adding their money to the fund for no real benefit to them? If it's the case this whole thing could be a way to take risks you couldn't take with your name + the additional PR from the model.

It seems like in such an open organization, and considering the problems (1%) that this is trying to address, there would be some tools for querying the wealth distribution. (There may/should be, idk; a few Googles came up empty.)

Yeah but it's anonymous / via bitcoin, no one could stop you from doing a 1000 'small' donations under a 1000 different identities, yet you control them all?

Damn. Of course you're right. And it could all be automated, transparent to you, the cheater.

Maybe they could charge a small per-share voting fee which would be destroyed (or used in some clever, non-game-able way).

I could even imagine automatically refunding the fee for those casting losing votes to avoid wars of attrition... I guess that's essentially an auction.

Wouldn't work, people could group together and remain anonymous.

I suspect that if N honest backers would expect a net gain from a decision even if they spend X/N voting for it, a single owner of N fake accounts would expect a net gain from the same decision even if he spent X voting for it.

Anonymity on the internet pretty much inherently means that you can't tell how many actual individuals a given set of apparent individuals really are. You can tell how much money they have (or CPU power, bandwidth, or various other proxies for wealth), but that's about it. I don't see how this wouldn't be an inherent problem for any cryptocurrency (or crypto-whathaveyou-system) attempting to force decentralization (see; the primary actors in Bitcoin split war.)

If it makes a return for the random people, isn't that a benefit?

That depends. Ponzi schemes generate returns for random people for a while as well. Not that this is a Ponzi scheme but generating returns for random people in and of itself is not always beneficial, you need to look at the longer term to establish whether or not the net result is positive.

Anyone who invests in the DAO can split their money out at any time

The fundamental question with the DAO is "what happens when something goes wrong". Contract law is simple when everything goes right - item gets delivered and paid for, contract is fulfilled and complete. It's the error and fraud cases that are hard. As Bitcoin has taught us.

I hope this reality check cools off all the zealots. To me, this is, like, "duh!" At best, extremely risky to spend your money on; at worst, obviously stupid.

And at best, extremely high returns!

Sure, same applies to the lottery, except there's substantially more evidence of the lottery generating extremely high returns for ordinary players than an unregulated investment pool governed by the votes of anonymous entities.

Of course the DAO is very likely to fail. It's an experiment. Like the system it's based on.

It will fail, we're going to learn from these failures, and then try again. And then again. And again. And after enough iterations and learning from earlier mistakes there could be a chance for it to succeed.

It's not about the money, but about the technology. Yes - people are throwing lots of money at the DAO, but it's their own decision to place an extremely high risk investment for the chance of a potential gain in the future. Don't blame the technology for that.

I'm really glad that there are people who try to continue pushing things forward, although everyone tells them that their ideas are never going to work out. If those people wouldn't exist we'd still be throwing rocks together to light a fire.

Thanks for posting this defense. There are so many detractors that are calling this stupid because it's not fully baked. People voted with their wallets to make this happen and effect change. Let them.

There are risks and unknowns. Will it fail? Probably. But most ideas and ventures fail. So a high likely hood of failure (which is inherent in all novel ideas and ventures) is an intellectually lazy reason to reject it.

It also bothers me that you have these lawyers calling this sloppy and poorly executed. Like the founder had any idea that this would take off as it did. This is a cool project and I think it can self correct. If not, it's a wonderful experiment to eat hand the worlds a better, more interesting, place because of it because it's things like this that could be zero to one events. And that is rare.

I only recently because aware of Ethereum and the concept of DAO so my understanding is pretty hazy. It strikes me that this is version .1 of something could be really cool, but which like most products doesn't really become usable for the masses until version 3.0. Sort of like Friendster -> MySpace -> Facebook

I hope the evangelists keep iterating but I'll stay away from it until I feel like it's something that will stick around for at least 10 years.

DAO tokens can be redeemed for the Ether it represents at any time. With a delay. People aren't doing that because they can sell the DAO token on exchanges much faster.

I will say it has a shot, it's not as risky as say buying a shady altcoin during a pump. But it sure as hell has the possibility of crapping out entirely.

So it turns out that the DAO is buggy. Hardly a surprise, given that it has hundreds of lines of code. What is surprising though is that the code was allegedly audited for bugs and security problems before release: https://blog.slock.it/deja-vu-dao-smart-contracts-audit-resu... ... and they found none of these big problems. What does that say about code audits?

At a first approximation, it sounds like the problem is perverse incentives, not buggy code. The code does what it's supposed to, but it has consequences that were either not noticed or not publicised.

> given that it has hundreds of lines of code

That's our benchmark for massive project now? Hundreds of lines of code? Any decent open-source project should be able to handle hundreds of lines without breaking.

I didn't say anything about massive. On average, computer code has about one mistake per ten lines of code (very roughly, of course) So once you have over a few hundred lines of code, you are practically assured of having bugs. And then you throw over $100 million at the buggy code, and the results are ... as expected.

I thought Star Citizen was the largest crowdfunded thing ever. Source?

https://en.wikipedia.org/wiki/List_of_highest_funded_crowdfu...

Space nerd checking in: Star Citizen is currently at ~115m[1], DAO reads to be at 16something, from my skim of the article. Additionally, people who get more picky as to what counts as "crowdfunding" sometimes only cite the early kickstarter campaign for SC, but at least "the big number" seems smaller.

[1]https://robertsspaceindustries.com/funding-goals

Me too. DAO is more recent, though. It might be the largest number of backers.

Maybe they really meant croudfunding instead of croudfunded? The DAO is supposed to be a decentralized system for managing a fund that invests in other projects.

We replaced the baity title (in which direction Wired sadly took a noticeable step a while ago) and replaced it with what seems like a representative sentence from the article. If anybody suggests a better (more accurate and neutral) title, we can change it again.

"Kind of" is the understatement of the year when you consider the amount of money involved and how stupid this idea actually is.

Please don't post unsubstantive dismissals. If you want to make a point like this, the way to do so is to make it thoughtfully and include actual information to help readers evaluate the truth for themselves.

Reply on news.ycombinator.com