Re: [ExI] Bitcoins Again

3 views
Skip to first unread message

Brent Allsop

unread,
May 14, 2013, 9:57:54 AM5/14/13
to ExI chat list, transf...@googlegroups.com

Hi Kelly,

Thanks for this analysis.  It provides a key piece of thinking to my understanding.

One possible weakness I've been concerned about, is if some 'network' of scam peers could start overwhelming the legit peers.  If it gains a majority of computer power, they can start hacking with the master ledger, creating counterfeit coins and or double spending them.

This article about how the bit coin network is now way more powerful than the top 500 supper computers combined greatly falsifies my worries about the possibility of legitimate bit coin peers being overwhelmed.  Though I still wonder about how much compute power is being installed by the NSA, at the new Utah mega computer sight and other such facilities.  Certainly any secret computing power they have isn't going to be registered with the top 500 publicly known supper computers?

So, now that that worry has been mostly falsified for me, the only remaining worry (which is now compounded by the above information) is the profound effect bit coins could have on the economy, if they really become popular.

Immoral investments, like Gold, have always been a drain on economies, as it amplifies bubbles and crashes, or the tendencies for the world economy to be like a dramatic roller coaster.  Federal reserves try to put more money in the system, to compensate for the people hording cash, and abandoning stock markets.  But, thankfully, this effect hasn't been too bad.  But Bitcouns could take this badness to a whole new level.

I heard one naive pundit argue that it is good for a currency to deflate, as it encourages people to save, rather than spend.  But that isn't what you want.  You want people to invest in factories, technologies, and other things that improve the world.  In other words, you want people to invest in the stock market.  If people are 'saving' by putting their money in something that doesn't contribute, like Gold, or Bit coins, this can be very detrimental to what is required for technological progress.  If technology / stock market is only returning at best 10% / year, while each bit coin 'bubble' is 10 time the last bubble (as history is starting to show) this could suck every last penny out of the world economy, in a very terrible bubble ish way (i.e. as the economy get's worse, it only encourages people to buy ever more Bitcoins as the new replacement for gold both of which destroy the ability of the worlds federal reserves to 'stimulate' the economy when needed.)

So, my question is, am I making any mistakes in my reasoning?  How many of you agree with all this?  Are there any other significant 'risks' to Bitcoins?   Could this lead to some kind of global war against people using Bitcoins, and people trying to give the power back to governments, to stop the wild economy destroying swings, that just drives bit coins to ever higher 'bubbles' every time the economy turns down?

It's sure be great to know, concisely and quantitatively, what all the experts think on this kind of stuff, to get a much better view of various possible futures.  All the individual pundits are just so clueless on so many things.

Brent Allsop























On Mon, May 13, 2013 at 9:59 PM, Kelly Anderson <kellyc...@gmail.com> wrote:
I've been trying to do an estimate of what I might expect to make if I bought a bitcoin miner from Butterfly Labs, and I thought I'd share my calculations in case any of you found it interesting (and to double check my math... LOL)

Ok, So according to http://www.bitcoinwatch.com/
The network's current power is 82.77 TeraHashes/Second (I'm making the assumption that there are zero butterfly labs machines out there at this point, which may not be entirely correct)

The network is automatically adjusted to target the creation of 7200 bitcoins per day (though it varies both up and down, this is the average that the math attempts to maintain).

So, you currently need to do 41.385 TeraHashes to create a bitcoin, on average. This makes sense since 7200 is about twice the number of seconds in a day... So far, I think my math is about right.

So the vapor hardware 50 GH/s Bitcoin Miner at $2,499 (that's right, not yet available) would create a bitcoin, on average every 827 seconds, or every 13.7 minutes. (Of course, they are really created 50 at a time...)

So now the interesting question is what happens when there are thousands of these machines actually out there... What happens is the difficulty goes up. Some of the serial numbers on ebay suggest that they have presold as many as 53,000 of the 50 GH/s Miners, though that is a guess based merely on serial numbers beginning with 1 and going up by 1, which is probably a reasonable guess. So that gives 50 GH/s * 53000, which ups the network capacity by 2650 THashes/second, meaning that the entire network will be 1587 TH/Sec, making the difficulty roughly 31 times more difficult (and making all the old bitmining hardware pretty worthless). And that is just the big boxes, doesn't count all their smaller units, and those of their competitors... Is my math messed up here?

So if the guy on Ebay is right that one of the 60GH/s machines will make $336/day (I didn't check his math). Then once these are out there, you could expect to make about $10 a day off an investment of $2500, which is a return of $3650 on an investment of $2500 if nothing better comes out during the year... for a profit of $1150 minus the cost of electricity. That seems like a lot of risk.

I found another article where the guy said he thought the difficulty was rising about 5% per week. If a bunch of these new machines ship quickly, that will be a fond memory of the past, for a little while anyway.

Another interesting article on how the bitcoin network compares to supercomputers.

So, I can only assume that it is going to get worse, as this chart seems to show. 

Bottom line... If you can get an ASICS machine in the VERY short term, it's probably not too horrible... but if you get one in six months, it will take a long time to earn a return, if you ever do. But damn, that's some number crunching for the ages folks!

-Kelly


_______________________________________________
extropy-chat mailing list
extrop...@lists.extropy.org
http://lists.extropy.org/mailman/listinfo.cgi/extropy-chat


Carl Youngblood

unread,
May 14, 2013, 10:08:49 AM5/14/13
to transf...@googlegroups.com, ExI chat list
Khan Academy also has some really good videos explaining how the bitcoin protocol works in detail, in case you're interested.


--
To learn more about the Mormon Transhumanist Association, visit http://transfigurism.org
---
You received this message because you are subscribed to the Google Groups "Mormon Transhumanist Association" group.
To unsubscribe from this group and stop receiving emails from it, send an email to transfiguris...@googlegroups.com.
To post to this group, send email to transf...@googlegroups.com.
Visit this group at http://groups.google.com/group/transfigurism?hl=en-US.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

David Hamilton

unread,
May 14, 2013, 4:45:34 PM5/14/13
to transf...@googlegroups.com
You are spot on with your analysis of a 51% attack and the likelihood being reduced by a large decentralized network of peers. All I just had a couple things I wanted to add. Even if coins are lost in old wallets or taken out of the market by hoarders causing a deflation, the coins will eventually be infinitely divisible (although only 8 decimal places are written in as of the current code) Also for the lack of goodwill contribution, there is a coin that isnt in use much but its out there and its called DevCoin, a large portion of all the mined coins are given to open source software developers. People could build on this idea, I could imagine a SETI Coin or something along those lines. The whole idea of mining digital currency is in its infancy but I think it can be applied to a whole host of projects. If you think of something you could build a coin yourself, no promise of it catching on but the option is there to build a better mousetrap.

Brent Allsop

unread,
May 14, 2013, 6:41:38 PM5/14/13
to Gordon, ExI chat list, transf...@googlegroups.com

Hi Gordon,

Wouldn't it be great to know, concisely and quantitatively, just what all the transhumanist Bitcoin experts believed about this?  Just how bullish are you/they?  And I bet tracking this real time, over lots of experts, all communicating concisely and quantitatively about why they believe, what they believe, would provide the most intelligent possible predictions about both the near and long term future of Bitcoins, and any effect they may have on the world economy, and so on.

I, like you, Gordon, am very bullish.  I think it would be much better to invest in the coins, at this point, and am heavily investing in them, rather than mining hardware.  James, are you investing in mining hardware or Bitcoins?  Is anyone else?  I consider you a good expert, and would like to know which way you really lean, why, and if you are actually investing either way?

Wouldn't it be great to know, concisely and quantitatively, what all possible risks the experts believe, and their odds, and all that?

If bit coins turn out to be catastrophic in any way, for either holders or the economy, it'd be great to know, sooner, rather than later, and a crowd of experts, amplifying their wisdom, concisely and quantitatively, about potential risks, would be way beyond any other intelligent system.

It does take a lot of work (one person can't do it) for everyone to build consensus, and to communicate concisely and quantitatively, so you know who does and does not agree, and why, but, as can now be shown on the consciousness survey project, the intelligence of everyone working on such blows anything else (like the primitive ivory tower, and their worthless 20 thousand "peer reviewed" documents) or any other individual "talking head pundit" away on any moral issue such as this.


Brent Allsop



Brent Allsop











On Tue, May 14, 2013 at 10:06 AM, Gordon <gts_...@yahoo.com> wrote:
Mirco Romanato <painl...@libero.it> wrote:

Il 14/05/2013 15:39, Gordon ha scritto:

>> Here is a live stream of an inexpensive Butterfly ASIC in operation:
>>
>> http://www.wired.com/wiredenterprise/2013/05/butterfly_live/
>>
>> Pooled, it is mining about 2 BTC per 10 days. That is a fantastic ROI at
>> current prices, but I suspect you cannot buy one anytime soon at the
>> list price. As I understand it, it is backlogged into the thousands.
>> There is no telling what the economics will look like by the time your
>> order is filled, if it is ever filled.

>Currently the difficulty of mining went up five times from January.
>If it keep this, by the end of September it will be up other ten times
>from January. The raising appear to be linear, so it is proportional to
>the ASIC production (or it appear so to me).

Given the uncertainty about delivery dates on these Butterfly ASIC miners, my thought for the moment is that anyone fundamentally bullish on the outlook for BTC over the next year or so should invest in BTC rather than commit those same funds to a miner order. Would you agree, Mirco?

Gordon



David Hamilton

unread,
May 15, 2013, 1:00:26 AM5/15/13
to transf...@googlegroups.com
I've invested several hundred in bit/crypto coins but I haven't bought any mining hardware. I think buying hardware is an okay way to go about investing. Because then if all the bitcoins are worthless, you still have a video card! Unlike if you invested pure cash. GPU mining uses SHA-256. SHA-256 is used to mine Bitcoin, Namecoin, PPCoin, Devoin, Terracoin, many others.

I dont know much about building a custom mining rig but bitcoin uses GPU mining so if you are thinking about spending some dough on a video card, many high end video cards arent really cost effective if you are getting the top of the line card. Mostly because it just wont pay for itself. But if you want a card already or if you are getting a good deal, or buying a cheaper value card, go for it.

https://en.bitcoin.it/wiki/Mining_hardware_comparison has a list of the different hashrates of video cards, or even ascis (the new miners that are supposed to mine SHA-256 like crazy when they come out). The above link, it even has the hashrate per dollar spent on the equipment so it could come in handy. Not sure its 100% accurate and even if you buy a couple video cards that mine coins well, when something better (asics) come out and the difficulty shoots up you may not be bringing in the coins you were before, making it not pay for itself totally. Many people do have their cards pay for themselves but if you are getting a top of the line card it can be hard to make all of that cash up. So if you are going to use the video card to watch movies or play games too then have no qualms buying a good card or two. Otherwise I'd go with the cheapest you could spend for the highest hashrate, There are lots of online discussions about which cheap cards work best. Also watch out for the watts you are using up, that can cut into profits.

No expert here, and I want to hear the experts step up too. Just explaining things the way I was lead to believe they are.

If you are looking to mine litecoins, novavoin, feathercoin, chinacoin, mincoin, bbqcoin it uses scrypt so its CPU based rather than GPU. So thats a whole different discussion.


Carl Youngblood

unread,
May 15, 2013, 2:36:09 AM5/15/13
to transf...@googlegroups.com
My understanding is that GPU-based mining will be rendered ineffective by FPGA-based mining. It probably hasn't happened yet because these devices only barely started coming out and haven't yet been produced in large quantities.

David Hamilton

unread,
May 15, 2013, 4:09:51 AM5/15/13
to transf...@googlegroups.com
I buy/mine Litecoin mostly which is CPU based. One of the few alt-digital currencies that is not only usable to trade to bitcoin but also USD. Plus the most web commerce support of any alt-crypto coin.

The thing with the ASICs and especially BFL, they are only delivering to the press and the first couple weeks of pre-orders. Lots of people gave them bitcoin when it was cheap and lots of people think they are running with the money despite the fact that some people have asics already When they do get released I think the people who are currently gpu mining will switch over to a scrypt coin or something of that nature. If people start using FPGAs with insane hashrates, there is only a short time before the difficulty increases to compensate for the increased hash power. Its built to only give out so many coins at regular intervals by adjusting the difficulty of mining a block. And it's built for hundreds of years of computer progress (maybe quantum computing could change the game though.)

I think that it will turn people more towards specialized set ups when it comes to mining. The average joe wont be mining bitcoins from his family pic, but scrypt and altcoins and people okay with taking little or no profit will help keep it up.

My main worry is when the regular block rewards are no longer given out automatically, after that block rewards will be completely based on transaction fees. With no new money coming in then deflation is a problem but like I said the coin is very divisible so even if the price goes up slightly for everyone it evens out. And old hoards of money could probably be found like opening any old safe or wallet.

Traders will compete for higher fees to get faster transactions. Miners could charge less fees for a slower transaction. If I am correct, the blocks that are mined to give you bitcoins are actually a data ledger of the different transfers that happened during that period. Transfers are confirmed each block when the different ledger blocks are checked against each other, its like 8 confirmations with bitcoin sometimes before a transfer is completely accepted. Litecoin blocks are mined once every 2.5 minutes, while bitcoin blocks are mined every 10. This makes transfers a lot faster too if you are using litecoins. At the same time the difficulty would increase making each supercomputer mine barely more than the computers they started with, and the dude with the regular computer who wanted to mine cuz it sounded fun is kind of out of the picture. It will make bitcoin mining for people who know what they are doing and are willing to invest in equipment. If the reward is not as high for miners based profit from other people's fees being paid out then less people will want to mine and that decreases security (51% attack) it could take forever to get a transaction confirmed until difficulty decreases again and people want to start mining it once more, increasing security.

So I stick with Litecoin (scrypt/cpu ming) most of the time, Its mostly protected from those kinds of mining rigs. You can build one of those types of rigs for scrypt but the price makes it pointless. So I buy and trade and horribly mine litecoins. I trade over to bitcoin when its undervalued while litecoin is looking bubble-ish, but I switch back as soon as litecoin goes back down. I just like scrypt better and I can actually make coins with it, with bitcoin I cant even make slivers as far as I can tell. But the quad core processor helps with scrypt. Another good thing about litecoin is transfers, between bitcoin and litecoin they are shown instantly, but confirmations happen each block (like I explained each block is a list of who sent money to who.

The litecoin price is lower. There are about 4x more litecoins than bitcoins. If you dont like litecoin but like scrypt there are many options (like feather, its a scrypt coin that is to litecoin what litecoin is to bitcoin although they are having difficulty adjustment problems) but stay way from scam pump and dump coins. You can buy the new cool coin at .02 with a hundred bucks and then it goes down to .0002 and its all gone. You never know, research each coin., only buy what you believe in. If you want to mine that new currency that gives you 5000 coins every second thats great, save them for a few months and see how the price looks. Dont buy them. Buy what you can trust.

http://altcoins.com/ has an okay list math based coins in scrypt and sha sections
https://bitcointalk.org/index.php?topic=134179.0 has a bigger list with stuff alive and dead



 Sorry I wrote this on a phone.

Gordon

unread,
May 15, 2013, 7:26:29 AM5/15/13
to ExI chat list, transf...@googlegroups.com
Hi Brent,

> Wouldn't it be great to know, concisely and quantitatively, just what all the transhumanist Bitcoin experts believed about this?  Just how bullish are you/they?

I think it would be great to have an indicator of the sentiment of all bitcoin investors and traders, including but not limited to transhumanists. Such measures can have value as contrary indicators. When most everyone thinks the price of an asset is going one way or the other, it is often wise to bet against the consensus.

I once wrote a program to analyze the P/C ratio (the ratio of the volume of put options [bearish bets] to call options [bullish bets]) on the S&P 100. I found a statistically significant relationship between high P/C ratios and market rallies. I used my model successfully with my clients for several years.

Gordon




Reply all
Reply to author
Forward
0 new messages