This article is more than 1 year old
Well YES, Silicon Valley VCs do think you're a CRETIN
Bitcoin-mining toasters leap forth from stealth
Worstall on Wednesday A couple of weeks back I pointed to some rumours about what the stealth (and hugely VC backed) company 21 Inc was up to, with the idea being, seemingly, to stick a Bitcoin mining chip in your toaster. So, as well as that breakfast comestible you could lose money on your electricity bill by attempting to mine for Bitcoin, obviously.
21 has now come out of stealth and, well, that's what it's trying to do: put a Bitcoin mining chip in your toaster. Here's its post announcing the grand opportunity for everyone to lose money.
Now, there are some possibly sensible things being suggested as uses. For example, one thing the blockchain does achieve is a register of who owns what and how – or whether – it has been transferred. Some are much more excited about the possibility of it being a register of ownership than they are of the currency itself.
So, what follows doesn't look entirely ridiculous:
Decentralised device authentication. Many applications of bitcoin are not strictly monetary. Embedded mining means that any device can authenticate itself to the network by sending one Satoshi to a specified address.
When used in combination with multisig, a 21 BitShare offers new paradigms for device authentication.
Machine Twitter. The constant stream of Satoshis provided by a 21 embedded mining chip means that any device with an internet connection can now write to a globally readable, always accessible database – with obvious benefits for device coordination, ad-hoc networks, and the like.
Whether it's worth having spent $116m to get that into a chipset that can then be used by, say, smartphone or video card manufacturers is unknown. But it's not an obviously ridiculous method of pissing capital up against the wall.
However, the rest of the pitch is, as I compared it with last time, something out of the South Sea Bubble. For it really is saying that there should be hundreds of millions of these mining chip cores out there, all beavering away to mine Bitcoins. The problems with this being slightly obvious once we start considering the basics of what is really being said.
For a start, the more computational power that is set to the task of mining the less any specific amount of computational power makes in Bitcoins. This is baked into the basic assumptions of the whole protocol.
Stick hundreds of millions of low grade chips out there, all mining away, and the current economics of mining go ever further south. It is already true that only the most advanced ASICs, using the lowest cost power, actually make a profit from mining.
Such ASICs, to be competitive, tend to last only a few months before the next iteration and generation overcome their profitability. Having a standardised chipset in devices that will last for some few years (yes, some change smartphones once a year but two or three is more normal) means that the installed base will be falling ever further behind that profitability frontier.
Bitcoin – fundamental system resource
But they really are saying that this is the main point of the whole adventure. Conceptually, we believe that embedded mining will ultimately establish Bitcoin as a fundamental system resource on par with CPU, bandwidth, hard drive space, and RAM.
That is, one can imagine the ultimate thin client in which a system designer consciously chooses a relatively slow CPU but a relatively strong 21 mining chip, using the Bitcoin generated therein to purchase computation in the cloud.
Crucial to this is the idea that Bitcoin generated by embedded mining is more convenient – and hence more valuable – than Bitcoin bought at market price and manually moved over to the site of utility.
Sure, there's that utility value from ease of access. But that's got to be hellaciously large to overcome the basic economic problem here: mining Bitcoin, except on the most advanced rig with the cheapest power, loses money.
These will be, embedded in devices as they are, using electricity at average retail and consumer prices. And it'll be a chip that within months, if not weeks, will be well behind the profitability curve even in the best of circumstances. So everyone attempting to mine in this manner is going to be losing money. Whatever the convenience, it would be cheaper to go and buy already mined coin and then store it locally.
Further, any single chip of this kind is going to be mining such incredibly tiny fractions of a coin over time that there's absolutely nothing at all that one could possibly buy with it.