为之漫笔

Don’t use a blockchain unless you need to – Hacker Noon

为之漫笔 · 2017-06-13推荐 · 36阅读 原文链接

Don’t use a blockchain unless you need to

An easy way to spot a startup that won’t provide return on investment is to look for the words “blockchain” or “decentralized” on their landing pages.

I divide them mainly into two categories:

  • the bitcoin replacements

  • the “built-with-blockchain” crowd

Let’s start with the first.

Betsy Bitterly Bit A Better Bitcoin

I bought my first bitcoin in 2011 when I wanted to make sure my friends and I had good MDMA for a festival on the Isle of Wight. I was new to London, I had no hookups, and drug dealers are in any case not fun to spend time with.

Six years later, there are still only two ways bitcoin is being used at any real scale:

  1. to help people hide their money from governments (mainly in China, where the government at all levels is more corrupt and capricious than anything we’re used to in the West), and

  2. to buy drugs over the internet.

If anyone could do these things without a complex, unfamiliar, risky system like bitcoin, they totally would. There are just few good alternatives (ask anyone who’s ever bought drugs in person or tried to evade tax).

Now, bitcoin isn’t perfect. It takes over twenty minutes for a transaction to come through properly, and if you’re not careful you can make it easy to trace whom you sent money to.

As a result, there are a raft of alternative coins (altcoins) which are like bitcoin, but:

  • faster (e.g. Dash)

  • more anonymous (e.g. Zcash)

  • with lower transaction fees (e.g. Litecoin)

  • with dadaist pretensions (okay, this one’s just for DOGE)

What they fail to mention is that at the same time, their coin is like bitcoin but without:

  • network effects

  • brand recognition

  • years of being unsuccessfully attacked by the world’s most sophisticated hackers

  • relative price stability (emphasis on relative) and liquidity

Buying your monthly dose of opiates is not the same as buying a soy latte: the fact that you have to wait twenty minutes for the money to reach the drug marketplace just does not matter.

And it’s a similar deal with money laundering. Sure, bitcoin could’ve been made more anonymous by default. But all you have to do is google “bitcoin cleaning” and you’re fine.

So, that dispenses with the people trying to sell us a more perfect and beautiful internet money. Let’s turn to the other category, a much bigger and more complex one.

Your Business Probably Doesn’t Need To Be Decentralised

Let’s take as an example Steem, and their flagship product Steemit.

It’s in a space that’s near and dear to my heart: rewarding people for creating, discovering and sharing good stuff. People who trawl through the gravel pile of online content to find gemstones should be rewarded with something better than Reddit karma points. Few people give a shit about Reddit or HN karma. At least with retweets they get exposure and new followers, but that’s still chump change compared to the value that comes from curation, discovery and sharing.

The best way society has for rewarding people for doing useful work is with money, so let’s do that instead, the idea goes. So far so good.

But Steemit sucks as a product, and has barely improved since their launch over a year ago. They will never take people away from Reddit at the rate they’re going, even though they raised millions with their ICO. I tried to use it, I really did. Now, if Steem were one of the coins available for margin trading on Poloniex I’d be shorting them. Their product is built on top of a brittle, slow (compared to, you know, a normal database), potentially unsafe, potentially irrelevant-in-the-future blockchain platform, for no good reason.

Solving the “rewarding-creators-and-discoverers“ problem doesn’t require decentralisation. No one is going to shut down a version of Reddit that rewards people with money for being the first to discover good stuff. Steemit doesn’t need a blockchain, it needs better branding, design, copywriting, and most importantly the ability to iterate their product quickly. They need to be able to throw features at the funnel and see what converts. You know, that thing where you build a business on the internet.

Centralisation has huge advantages

As another example, let’s look at IPFS and Filecoin. I think they’re super fun and cool. But the only things I’d build on top of them would either be for activists or criminals. Those are the only people who need a way to store data like that, whatever the third season of Silicon Valley is telling us.

Richard is actually a lesson in what not to do. More like Silly Con Valley amirite

For the vast majority of businesses, it’s more important that your files are stored safely and cheaply, and can reach your users quickly. There is currently no better way to do this than to use Google, Amazon et al’s futuristic datacenters and CDN tech, and by the time Filecoin is ready, those corporations will be even further ahead in whatever robot- and AI-driven stuff they’re doing in those big rectangular prisms.

I’m all for playing with fun crypto tech, but I sure wouldn’t invest in a business- or consumer-facing startup that planned to use it for their primary datastore.

What about prediction markets?

I’ve been obsessed with prediction markets ever since I read Robin Hanson’s work on Futarchy.

I’ve been following contenders Augur and Gnosis since their inceptions, and I don’t have a lot of faith in their teams to reliably deliver. Augur in particular has been going since 2014 and they don’t even have an MVP. Each platform raised tens of millions of dollars in their ICOs.

It’s a shame that Intrade got shut down. But compromising with the US government is obviously possible: just look at PredictIt.org. They did a roaring trade on the 2016 election, while the Augur team were presumably reading Ethereum creator Vitalik Buterin’s latest whitepaper on how the network could theoretically do more than seven transactions per second, and scratching their heads. (Side note: Vitalik is a genius, but still human.)

In the worst case, where the US government just wouldn’t play ball, they could’ve tried to find a place that would. In fact, that’s what PredictIt did initially. They started in New Zealand, where they could experiment unencumbered before making the jump to the US. But, working around governments? Lobbying? Moving countries? That sounds like a lot of annoying, boring work. How much cooler and more fun would it be to do it with Ethereum, instead, right? It’s like they haven’t read PG’s essay on schlep blindness.

If I sound frustrated, it’s because I want Augur to be a real thing that I can use right now to make a bet on when George R. R. Martin’s next book is coming out, but I can’t, and the reason is that the founders thought that the key to making it work was the blockchain, and it’s unfortunately a less sexy problem. I’ve been waiting to give them my money since 2014. They are clearly doing something wrong.

What about all the corporate interest?

The marketing departments of big, old, uncool corporations love opportunities to look sexy, especially when all it costs them is a press release and sending some people to conferences. Let me know when they’re actually using it for any significant part of their day to day operations.

C comes before D in the alphabet

Napster came before BitTorrent. The walled gardens of CompuServe and AOL came before the web. Hilton came before Airbnb. Taxi companies came before Uber. And HSBC and government-issued coins came before bitcoin. Centralisation has advantages for businesses trying to find product-market fit.

If you have an idea for a new business that doesn’t need to be protected from the prying eyes or arbitrary gavel of a government, then don’t involve a blockchain. Your lunch will get eaten by the team using technology more appropriate for solving the customer’s actual problem — the team who just sucks up the fact that they can be shut down at any time by The Man, that they are the single point of failure, and that they don’t get to put as many buzzwords on their landing page.

相关文章