[espians] Amp: A proposal for a global reference currency
32 views
Skip to first unread message
tav
May 13, 2010, 10:22:44 AM5/13/10
to Espians
Hey all,
As some of you know, a few of us have been busy finalising the spec
for "Ampify" — the new name for "everything" ;p
As part of this process we tackled the issue of a global reference
currency in recent days and I figured I'd share the outcome to get
everyone's thoughts...
But, first, a quick intro to the current system where currencies like
the Dollar and Sterling are used for mixed purposes:
1). As a medium of exchange — so that you can easily compare
the relative prices of apples and oranges without having to
resort to doing complex calculations like 1 apple = 0.43872 of
an orange.
2). As a store of value — so that you can put your money into a
account and use it as an asset at some point in the future.
3). As a tool to manage resources — especially with regards to
people's time (labour).
And the failings of the present monetary system are now quite evident:
a). When economic growth is tied to getting into debt, the spectacular
deficits of our leading economies should come as no surprise.
http://video.google.com/videoplay?docid=5352106773770802849
b). When electronic trading fueled by complex algorithms are
adopted wholesale without much regard to the systemic
implications, the 500-point drop in the Dow within 5 minutes last
week shouldn't be a surprise.
c). When Wall Street's primary interest stops being about investing
in businesses, the unsustainable overleveraging of the debt in the
economy through derivatives is quite inevitable.
http://blogmaverick.com/2010/05/09/what-business-is-wall-street-in/
d). When currency speculators trade over $4 trillion every day on
the foreign exchange markets, it should be no surprise that entire
countries can be bankrupted overnight due to pure hysteria.
e). When politicians have no clue what they're talking about and
believe that they can resort to the fairy dust that is "regulation",
we can be sure that it's only going to get even more fucked up.
To quote Keynes:
“Speculators may do no harm as bubbles on a steady stream of
enterprise. But the position is serious when enterprise becomes
the bubble on a whirlpool of speculation. When the capital
development of a country becomes a by-product of the activities
of a casino, the job is likely to be ill-done.”
Our global economy is now subject to the whims of the casino and there
isn't much that is being done to tackle the root causes.
The "best" mainstream ideas involve SDRs (Special Drawing Rights) and
the creation of an International Monetary Policy Committee — because
central banks have been so awesome in managing the global economy!!
(*sarcasm*)
The "radical" monetary reform world isn't too bright either. It is
littered with flawed ideas like LETS, complementary currencies pegged
to fiat currencies (e.g. The Brixton Pound), time banks, etc.
But amidst all the crap, there is one idea that gets too little
attention – global reference currencies. There are many variants of
the idea and Keynes even proposed one (the bancor) in 1943 during the
lead-up to Bretton Woods.
But the Terra proposal by Bernard Lietaer, one of the architects of
the European currency system, is perhaps the most interesting. I'd
encourage you to read the white paper. It is very readable:
http://www.terratrc.org/PDF/Terra_WhitePaper_2.27.04.pdf
A Terra is defined as being equivalent in value to a fixed basket of a
dozen or so commodities, e.g.
100 Terras = 1 barrel of oil + 5 bushels of wheat + 10 pounds of
copper + 3 pounds of tin + 1/10th ounce of gold + 1 carbon emissions
right
Since Terras only come into "creation" when actual commodities have
been handed over to the Terra central authority, anyone using Terras
can be confident that there is *real* value backing the currency. That
is, they can "cash in" their Terras for real commodities at any time.
And the use of multiple commodities is meant to act as a safeguard by
averaging out price volatility in any individual commodity and thus
make the Terra relatively inflation resistant.
Unfortunately that last point is no longer true due to across the
board volatility in commodity markets and the large volume of
speculative trades in the forward markets.
But the Terra also incorporates another idea that makes it interesting
— it has a built-in demurrage charge (negative interest) of 3-4% a
year.
To understand demurrage, consider a scenario of two farmers, Alice and
Bob. Let's say Bob has had a shit harvest and can't afford to plant
his next crop.
So Alice loans him what he needs and Bob promises to pay her back from
the proceeds of his next harvest. But as collateral Alice holds on to
Bob's cows, just in case Bob has a shit harvest again.
The next year Bob has a great harvest, sells it on the market, repays
Alice and stores the surplus in his granary. As "interest" for the
loan, Alice keeps all the milk from the cows and the calves born in
the interim.
Now, from the scenario so far, it's not a giant leap to the present
day situation of interest-bearing currencies. Unfortunately, there are
undesirable properties of such a currency: hoarding and boom/bust
amplification.
Hoarding is the easiest to understand. If the currency can be a store
of value and bears interest, then it generally makes sense to hold
onto it for as long as possible.
Now, this by itself can be benign as there will be a few willing to
risk some of their capital for greater returns. But when coupled with
the current system of banking, it can be deadly.
As Lietaer states:
"The money-creation process by the banking
system tends to amplify the fluctuations of the
business cycle.
"The banking system simultaneously tends to
either making credit available or restricting it.
Specifically, when business is good in a particular
market, banks tend to be more generous in terms
of credit availability, thereby pushing the “good
times” into a potentially inflationary boom period.
"Conversely, as soon as the business horizon
looks less promising, banks logically tend to
reduce credit availability, thereby contributing to
the deterioration of a minor business dip into a
full-blown recession."
Sound familiar?
Monetary policy is meant to counter this through interest rates. But
given the massive amount of debt in today's global economy, the effect
is pretty feeble nowadays.
For example consider that the Bank of England has frozen the interest
rate at 0.5% for the last 14 months, yet credit cards continue to hike
interest rates above 30% APR!
In contrast, a currency like the Terra which incurs demurrage will
result in the currency losing value over time and thus encourage flow
(and thus participation) within the economy.
So say you were able to buy 5 apples for 10 Terras today and that the
annual demurrage charge is 4.36%, then in 5 years you'll only be able
to buy 4 apples with those same 10 Terras.
Bringing us back to the example with Alice and Bob, the demurrage
could be considered analogous to the surplus that Bob will have lost
from his granary over time to insects, rodents and humidity.
The beauty about demurrage is that it leads to an economy that is far
better at self correcting than the painfully slow trickle down effect
from central banks to the high street.
I hope the above is reasonably cogent. I've tried to keep it simple,
but fear that it's still too complicated =(
Anyways, Amps.
The Amp is a global reference currency in the lineage of bancor and
Terra, except that instead of a basket of traditional commodities, it
consists of:
* Storage
* Processing
* Transfer
So the initial basket may be defined as:
1 Amp = (3 Processing Units for an hour) + (15 GB Storage for 10
days) + (5 GB of Transfer)
The beauty here is twofold. First, storage/processing/transfer is
something available to most people with a computer and, increasingly,
phones. So one would be able to "create" and "redeem" Amps for their
underlying value with a substantial portion of the global population.
Secondly, thanks to Moore's Law and technological advancements, the
value of the Amp will be constantly dropping — a truly built-in
demurrage charge!
Unfortunately the rate of Moore's Law is presently too high and price
stability is relatively important, so Mamading and I devised a
solution using dampening factors. That is, each of the individual
elements in the basket will have gradually adjusted dampening factors
which would aim to stabilise the demurrage at about 2% a year.
The dampening factors also gives us a way to counter the eventual
demise of Moore's Law and any new technological advancements that
might come along — not to mention being able to change the composition
of the basket itself...
Now, one of the challenges of any complementary currency is of course
getting enough distribution. For this, we'd use a similar strategy to
Facebook Credits and Second Life's Linden Dollars. That is, to build
it into the Ampify platform.
http://blogs.secondlife.com/community/features/blog/2007/08/14/the-second-life-economy
When one wants to pay for the hosting of an AmpNode, it'd be
denominated in Amps and we'll provide an integrated AmpExchange for
people to buy and sell Amps. Pecu payouts would be denominated in Amps
— as will the charge that people set for others to pay if they want to
be messaged by unknown individuals.
Most people would probably just treat it as yet another "in-world
virtual currency" for a while and perhaps come to value its stability
properties later on. But that doesn't matter as long as the Amp gains
distribution =)
With regards the creation of Amps, there will only be a single mint.
This deeply pains the Minister of Decentralisation in me, but sadly,
it's just too insane (from a user experience perspective) to have
multiple mints.
We'd "issue" Amps by selling them on the AmpExchange just like others.
And maintain a reserve in other (hopefully stable) currencies/assets
to buy back Amps as needed.
The automated trading in recent years (over 80% on some exchanges!)
leads to potentially insane spikes like the recent 500-point Dow
free-fall. Circuit breakers which halt trading are meant to counter
this:
http://en.wikipedia.org/wiki/Trading_curb
But I'm not too convinced about them. The recent free-fall might have
even be triggered by certain algorithms shutting down in recognition
of hitting the circuit breaker and thus failing to buy back stocks for
a while.
Instead I'd like to propose a TCP-like congestion control mechanism
for use on AmpExchange. That is, instead of fixing limits ahead of
time like with circuit breakers, it'd be a constantly adaptive flow
with sane levels of rate increase/decrease — giving time for human
sanity checks to override algorithmic insanity.
As for legal and taxation issues that might be barriers to the
implementation of Amps, like with the Terra, Amps are simply a
standardisation of counter-trade, and as such there is existing
legislation in nearly all countries.
Hopefully we wouldn't need to be heavily regulated as we could limit
issuing Amps to only users with accounts on already regulated
institutions, e.g. HSBC, Paypal, etc. This will hopefully help avoid
the tragic fate of e-gold:
http://www.wired.com/threatlevel/2009/06/e-gold/
There'd be nothing stopping an anonymous user from getting hold of
already issued Amps and using it for evil purposes. But, here, we can
hopefully leverage Trust Maps as a decent counter-mechanism.
While financial institutions providing facilities like escrow would be
useful in the Amp Economy, traditional banks have no incentive or
place. With its value eroding every minute, it makes no sense for Amps
to be used as long-term *store* of value.
Currency speculators also have little to gain by shorting the Amp and
their usefulness in terms of providing liquidity in the market is not
so important since Amps already encourage flow!
There is however scope for market arbitrage, e.g. Transfer costs more
in the U.K. than it does in Germany. I see arbitrage here as an
overall good thing as it'd help counter market inefficiencies and
bring down purchasing power parity differentials across regions.
I would be quite happy for us to take over ¤, the universal currency
symbol, for Amps, e.g. ¤240.10. And perhaps Amps should be denominated
in powers of 2 to make calculations easy, e.g. 1 Amp, 2 Amps, 4 Amps,
&c. Users wouldn't ever have to see that aspect though.
And perhaps there should also be a term for the fractional part? Cents
and pennies are deeply tied to units of 100 and I'd like to support up
to 9 decimal places — meaning that one could have transactions of
0.000000001 Amps.
Now there were many other things considered which didn't make it into
the final spec. For example, instead of defining Transfer as total
bytes transferred, it could've been defined as being proportional to
throughput over a period of time, e.g. 1Mbps for 2 hours.
The advantages of this would be that it'd encourage locality (i.e.
preferring Nodes closer to you) and caching. To see why locality is
important, consider the fact that with a ping (round trip time) of 200
milliseconds, you'd only be able to download a 100KB file over TCP at
a rate of about 1Mbps
You're going to be stuck at that limit even if you had a 100Mbps
connection or even a 1Gbps one! Even in a perfect system, the hard
limit set by the speed of light would be a bitch. So encouraging
locality makes a lot of good sense for the overall Amp Economy.
However, this complicated matters. Especially when it came to another
aspect that was considered and abandoned — that of using auctions to
manage the resources. This has been explored in the literature decades
ago, e.g. see the Agoric Papers from 1988:
http://e-drexler.com/d/09/00/AgoricsPapers/agoricpapers.html
And even attempted partially by our old friends at Mojonation:
http://en.wikipedia.org/wiki/Mnet#Mojo_Economy
In such a context, using throughput over a period of time as a metric
would certainly be more attractive in terms of efficient allocation of
resources, but it complicates auction algorithms so much that they'd
require human intervention. And asking users to make decisions
constantly may not lead to such a brilliant user experience.
Furthermore, while it would be attractive for say video broadcasting
services to know that Nodes are willing to offer a certain throughput
over a period of time, it is quite impractical with the fundamental
design of our current IP networks which may experience congestion at
any time.
And thus limits such use to high-speed institutional networks and
discriminates against the average home user who is, at best, on a
rather shit network.
To make things fair, there'd have to be widespread adoption of
fundamental changes to how IP is routed, e.g. this paper provides an
interesting financial instrument-based mechanism to route internet
traffic:
http://soda.swedish-ict.se/2270/1/SICS-T--2002-22--SE.pdf
Unfortunately, while possible, until the transition to such an
alternate routing mechanism is completed, it'd create a class system
in the internet. With those that can afford it being given first-class
treatment and others being routed at what would then be a severely
handicapped best effort.
So even though such a change would be immensely beneficial in the end,
e.g. everyone being able to open up their wireless networks and share
access with others in a fair way, the transitionary period is too much
of a threat to the existing fairness of the internet.
Less fundamental changes like requiring nodes to pay Amps when
contacting other nodes that they didn't have a previous relationship
with would be beneficial in many ways, e.g. providing an economic
counter-measure to denial of service attacks.
However, this puts up a massive barrier to adoption of the Ampify
network. It's easy to throw up a web server and no money changes hands
when a browser contacts the server. If Ampify required Amps to be
exchanged, I don't see how we could get enough nodes going in the
first place.
There is also user psychology to deal with. The majority of people are
now used to having (advertising-supported) free services where others
bear the cost of their use of resources.
And, finally, applying Trust Maps to inter-node transactions is
vulnerable to segmenting the network heavily as multiple users would
likely be co-located on any one node and a default on payments by any
individual user would have a negative impact on the node's reputation.
As Mamading put it, the market-based approach ends up in a
nickel-and-dimed scenario which isn't brilliant and ends up favouring
"cloud computing" over "peer computing".
As a consequence, I opted for a much simpler solution. In short:
* Users can exchange Amps for goods and services — in the way we use cash today.
* Users have accounts on AmpHosts — a collection of AmpNodes.
* Users can add others to their Trust Map to let them use resources on
their hosts, e.g. if Karina and Luke are happy for me to use their
hosts, I can run services there. Karina and Luke will perhaps end up
paying for it (or maybe not if they're just running it on their own
computer) and I'll be sent a receipt for my use in terms of Amps
(resource used).
* It's then up to whatever social transaction mechanism as to if and
how the resource usage is paid for with Amps, e.g. the use could've
been for free and I could just reward them later with Pecus or maybe
I'd agreed beforehand to pay a specific amount.
That is, the settlement mechanism is de-coupled from the underlying
resource usage and the base policy would be for free usage as long as
you were trusted for it.
And, finally, Amps are complementary to Pecus:
http://thruflo.com/2009/06/09/tav-describes-pecus.html
Amps and Pecus are a bit like yin and yang. Amps provide a shared,
"objective", present-oriented, grounded base for exchange and trade —
a reflection of the Natural Commons if you will.
In contrast, Pecus provide an individual, subjective, future-oriented,
values-driven base for directing social forces — a reflection of the
Cultural Commons if you will.
Anyways, that should be enough to get going for now — let me know what
you think! Thanks.
--
love, tav
plex:espians/tav | t...@espians.com | +44 (0) 7809 569 369
http://tav.espians.com | http://twitter.com/tav | skype:tavespian
--
You received this message because you are subscribed to the Google Groups "espians" group.
To post to this group, send email to esp...@googlegroups.com.
To unsubscribe from this group, send email to espians+u...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/espians?hl=en.
As some of you know, a few of us have been busy finalising the spec
for "Ampify" — the new name for "everything" ;p
As part of this process we tackled the issue of a global reference
currency in recent days and I figured I'd share the outcome to get
everyone's thoughts...
But, first, a quick intro to the current system where currencies like
the Dollar and Sterling are used for mixed purposes:
1). As a medium of exchange — so that you can easily compare
the relative prices of apples and oranges without having to
resort to doing complex calculations like 1 apple = 0.43872 of
an orange.
2). As a store of value — so that you can put your money into a
account and use it as an asset at some point in the future.
3). As a tool to manage resources — especially with regards to
people's time (labour).
And the failings of the present monetary system are now quite evident:
a). When economic growth is tied to getting into debt, the spectacular
deficits of our leading economies should come as no surprise.
http://video.google.com/videoplay?docid=5352106773770802849
b). When electronic trading fueled by complex algorithms are
adopted wholesale without much regard to the systemic
implications, the 500-point drop in the Dow within 5 minutes last
week shouldn't be a surprise.
c). When Wall Street's primary interest stops being about investing
in businesses, the unsustainable overleveraging of the debt in the
economy through derivatives is quite inevitable.
http://blogmaverick.com/2010/05/09/what-business-is-wall-street-in/
d). When currency speculators trade over $4 trillion every day on
the foreign exchange markets, it should be no surprise that entire
countries can be bankrupted overnight due to pure hysteria.
e). When politicians have no clue what they're talking about and
believe that they can resort to the fairy dust that is "regulation",
we can be sure that it's only going to get even more fucked up.
To quote Keynes:
“Speculators may do no harm as bubbles on a steady stream of
enterprise. But the position is serious when enterprise becomes
the bubble on a whirlpool of speculation. When the capital
development of a country becomes a by-product of the activities
of a casino, the job is likely to be ill-done.”
Our global economy is now subject to the whims of the casino and there
isn't much that is being done to tackle the root causes.
The "best" mainstream ideas involve SDRs (Special Drawing Rights) and
the creation of an International Monetary Policy Committee — because
central banks have been so awesome in managing the global economy!!
(*sarcasm*)
The "radical" monetary reform world isn't too bright either. It is
littered with flawed ideas like LETS, complementary currencies pegged
to fiat currencies (e.g. The Brixton Pound), time banks, etc.
But amidst all the crap, there is one idea that gets too little
attention – global reference currencies. There are many variants of
the idea and Keynes even proposed one (the bancor) in 1943 during the
lead-up to Bretton Woods.
But the Terra proposal by Bernard Lietaer, one of the architects of
the European currency system, is perhaps the most interesting. I'd
encourage you to read the white paper. It is very readable:
http://www.terratrc.org/PDF/Terra_WhitePaper_2.27.04.pdf
A Terra is defined as being equivalent in value to a fixed basket of a
dozen or so commodities, e.g.
100 Terras = 1 barrel of oil + 5 bushels of wheat + 10 pounds of
copper + 3 pounds of tin + 1/10th ounce of gold + 1 carbon emissions
right
Since Terras only come into "creation" when actual commodities have
been handed over to the Terra central authority, anyone using Terras
can be confident that there is *real* value backing the currency. That
is, they can "cash in" their Terras for real commodities at any time.
And the use of multiple commodities is meant to act as a safeguard by
averaging out price volatility in any individual commodity and thus
make the Terra relatively inflation resistant.
Unfortunately that last point is no longer true due to across the
board volatility in commodity markets and the large volume of
speculative trades in the forward markets.
But the Terra also incorporates another idea that makes it interesting
— it has a built-in demurrage charge (negative interest) of 3-4% a
year.
To understand demurrage, consider a scenario of two farmers, Alice and
Bob. Let's say Bob has had a shit harvest and can't afford to plant
his next crop.
So Alice loans him what he needs and Bob promises to pay her back from
the proceeds of his next harvest. But as collateral Alice holds on to
Bob's cows, just in case Bob has a shit harvest again.
The next year Bob has a great harvest, sells it on the market, repays
Alice and stores the surplus in his granary. As "interest" for the
loan, Alice keeps all the milk from the cows and the calves born in
the interim.
Now, from the scenario so far, it's not a giant leap to the present
day situation of interest-bearing currencies. Unfortunately, there are
undesirable properties of such a currency: hoarding and boom/bust
amplification.
Hoarding is the easiest to understand. If the currency can be a store
of value and bears interest, then it generally makes sense to hold
onto it for as long as possible.
Now, this by itself can be benign as there will be a few willing to
risk some of their capital for greater returns. But when coupled with
the current system of banking, it can be deadly.
As Lietaer states:
"The money-creation process by the banking
system tends to amplify the fluctuations of the
business cycle.
"The banking system simultaneously tends to
either making credit available or restricting it.
Specifically, when business is good in a particular
market, banks tend to be more generous in terms
of credit availability, thereby pushing the “good
times” into a potentially inflationary boom period.
"Conversely, as soon as the business horizon
looks less promising, banks logically tend to
reduce credit availability, thereby contributing to
the deterioration of a minor business dip into a
full-blown recession."
Sound familiar?
Monetary policy is meant to counter this through interest rates. But
given the massive amount of debt in today's global economy, the effect
is pretty feeble nowadays.
For example consider that the Bank of England has frozen the interest
rate at 0.5% for the last 14 months, yet credit cards continue to hike
interest rates above 30% APR!
In contrast, a currency like the Terra which incurs demurrage will
result in the currency losing value over time and thus encourage flow
(and thus participation) within the economy.
So say you were able to buy 5 apples for 10 Terras today and that the
annual demurrage charge is 4.36%, then in 5 years you'll only be able
to buy 4 apples with those same 10 Terras.
Bringing us back to the example with Alice and Bob, the demurrage
could be considered analogous to the surplus that Bob will have lost
from his granary over time to insects, rodents and humidity.
The beauty about demurrage is that it leads to an economy that is far
better at self correcting than the painfully slow trickle down effect
from central banks to the high street.
I hope the above is reasonably cogent. I've tried to keep it simple,
but fear that it's still too complicated =(
Anyways, Amps.
The Amp is a global reference currency in the lineage of bancor and
Terra, except that instead of a basket of traditional commodities, it
consists of:
* Storage
* Processing
* Transfer
So the initial basket may be defined as:
1 Amp = (3 Processing Units for an hour) + (15 GB Storage for 10
days) + (5 GB of Transfer)
The beauty here is twofold. First, storage/processing/transfer is
something available to most people with a computer and, increasingly,
phones. So one would be able to "create" and "redeem" Amps for their
underlying value with a substantial portion of the global population.
Secondly, thanks to Moore's Law and technological advancements, the
value of the Amp will be constantly dropping — a truly built-in
demurrage charge!
Unfortunately the rate of Moore's Law is presently too high and price
stability is relatively important, so Mamading and I devised a
solution using dampening factors. That is, each of the individual
elements in the basket will have gradually adjusted dampening factors
which would aim to stabilise the demurrage at about 2% a year.
The dampening factors also gives us a way to counter the eventual
demise of Moore's Law and any new technological advancements that
might come along — not to mention being able to change the composition
of the basket itself...
Now, one of the challenges of any complementary currency is of course
getting enough distribution. For this, we'd use a similar strategy to
Facebook Credits and Second Life's Linden Dollars. That is, to build
it into the Ampify platform.
http://blogs.secondlife.com/community/features/blog/2007/08/14/the-second-life-economy
When one wants to pay for the hosting of an AmpNode, it'd be
denominated in Amps and we'll provide an integrated AmpExchange for
people to buy and sell Amps. Pecu payouts would be denominated in Amps
— as will the charge that people set for others to pay if they want to
be messaged by unknown individuals.
Most people would probably just treat it as yet another "in-world
virtual currency" for a while and perhaps come to value its stability
properties later on. But that doesn't matter as long as the Amp gains
distribution =)
With regards the creation of Amps, there will only be a single mint.
This deeply pains the Minister of Decentralisation in me, but sadly,
it's just too insane (from a user experience perspective) to have
multiple mints.
We'd "issue" Amps by selling them on the AmpExchange just like others.
And maintain a reserve in other (hopefully stable) currencies/assets
to buy back Amps as needed.
The automated trading in recent years (over 80% on some exchanges!)
leads to potentially insane spikes like the recent 500-point Dow
free-fall. Circuit breakers which halt trading are meant to counter
this:
http://en.wikipedia.org/wiki/Trading_curb
But I'm not too convinced about them. The recent free-fall might have
even be triggered by certain algorithms shutting down in recognition
of hitting the circuit breaker and thus failing to buy back stocks for
a while.
Instead I'd like to propose a TCP-like congestion control mechanism
for use on AmpExchange. That is, instead of fixing limits ahead of
time like with circuit breakers, it'd be a constantly adaptive flow
with sane levels of rate increase/decrease — giving time for human
sanity checks to override algorithmic insanity.
As for legal and taxation issues that might be barriers to the
implementation of Amps, like with the Terra, Amps are simply a
standardisation of counter-trade, and as such there is existing
legislation in nearly all countries.
Hopefully we wouldn't need to be heavily regulated as we could limit
issuing Amps to only users with accounts on already regulated
institutions, e.g. HSBC, Paypal, etc. This will hopefully help avoid
the tragic fate of e-gold:
http://www.wired.com/threatlevel/2009/06/e-gold/
There'd be nothing stopping an anonymous user from getting hold of
already issued Amps and using it for evil purposes. But, here, we can
hopefully leverage Trust Maps as a decent counter-mechanism.
While financial institutions providing facilities like escrow would be
useful in the Amp Economy, traditional banks have no incentive or
place. With its value eroding every minute, it makes no sense for Amps
to be used as long-term *store* of value.
Currency speculators also have little to gain by shorting the Amp and
their usefulness in terms of providing liquidity in the market is not
so important since Amps already encourage flow!
There is however scope for market arbitrage, e.g. Transfer costs more
in the U.K. than it does in Germany. I see arbitrage here as an
overall good thing as it'd help counter market inefficiencies and
bring down purchasing power parity differentials across regions.
I would be quite happy for us to take over ¤, the universal currency
symbol, for Amps, e.g. ¤240.10. And perhaps Amps should be denominated
in powers of 2 to make calculations easy, e.g. 1 Amp, 2 Amps, 4 Amps,
&c. Users wouldn't ever have to see that aspect though.
And perhaps there should also be a term for the fractional part? Cents
and pennies are deeply tied to units of 100 and I'd like to support up
to 9 decimal places — meaning that one could have transactions of
0.000000001 Amps.
Now there were many other things considered which didn't make it into
the final spec. For example, instead of defining Transfer as total
bytes transferred, it could've been defined as being proportional to
throughput over a period of time, e.g. 1Mbps for 2 hours.
The advantages of this would be that it'd encourage locality (i.e.
preferring Nodes closer to you) and caching. To see why locality is
important, consider the fact that with a ping (round trip time) of 200
milliseconds, you'd only be able to download a 100KB file over TCP at
a rate of about 1Mbps
You're going to be stuck at that limit even if you had a 100Mbps
connection or even a 1Gbps one! Even in a perfect system, the hard
limit set by the speed of light would be a bitch. So encouraging
locality makes a lot of good sense for the overall Amp Economy.
However, this complicated matters. Especially when it came to another
aspect that was considered and abandoned — that of using auctions to
manage the resources. This has been explored in the literature decades
ago, e.g. see the Agoric Papers from 1988:
http://e-drexler.com/d/09/00/AgoricsPapers/agoricpapers.html
And even attempted partially by our old friends at Mojonation:
http://en.wikipedia.org/wiki/Mnet#Mojo_Economy
In such a context, using throughput over a period of time as a metric
would certainly be more attractive in terms of efficient allocation of
resources, but it complicates auction algorithms so much that they'd
require human intervention. And asking users to make decisions
constantly may not lead to such a brilliant user experience.
Furthermore, while it would be attractive for say video broadcasting
services to know that Nodes are willing to offer a certain throughput
over a period of time, it is quite impractical with the fundamental
design of our current IP networks which may experience congestion at
any time.
And thus limits such use to high-speed institutional networks and
discriminates against the average home user who is, at best, on a
rather shit network.
To make things fair, there'd have to be widespread adoption of
fundamental changes to how IP is routed, e.g. this paper provides an
interesting financial instrument-based mechanism to route internet
traffic:
http://soda.swedish-ict.se/2270/1/SICS-T--2002-22--SE.pdf
Unfortunately, while possible, until the transition to such an
alternate routing mechanism is completed, it'd create a class system
in the internet. With those that can afford it being given first-class
treatment and others being routed at what would then be a severely
handicapped best effort.
So even though such a change would be immensely beneficial in the end,
e.g. everyone being able to open up their wireless networks and share
access with others in a fair way, the transitionary period is too much
of a threat to the existing fairness of the internet.
Less fundamental changes like requiring nodes to pay Amps when
contacting other nodes that they didn't have a previous relationship
with would be beneficial in many ways, e.g. providing an economic
counter-measure to denial of service attacks.
However, this puts up a massive barrier to adoption of the Ampify
network. It's easy to throw up a web server and no money changes hands
when a browser contacts the server. If Ampify required Amps to be
exchanged, I don't see how we could get enough nodes going in the
first place.
There is also user psychology to deal with. The majority of people are
now used to having (advertising-supported) free services where others
bear the cost of their use of resources.
And, finally, applying Trust Maps to inter-node transactions is
vulnerable to segmenting the network heavily as multiple users would
likely be co-located on any one node and a default on payments by any
individual user would have a negative impact on the node's reputation.
As Mamading put it, the market-based approach ends up in a
nickel-and-dimed scenario which isn't brilliant and ends up favouring
"cloud computing" over "peer computing".
As a consequence, I opted for a much simpler solution. In short:
* Users can exchange Amps for goods and services — in the way we use cash today.
* Users have accounts on AmpHosts — a collection of AmpNodes.
* Users can add others to their Trust Map to let them use resources on
their hosts, e.g. if Karina and Luke are happy for me to use their
hosts, I can run services there. Karina and Luke will perhaps end up
paying for it (or maybe not if they're just running it on their own
computer) and I'll be sent a receipt for my use in terms of Amps
(resource used).
* It's then up to whatever social transaction mechanism as to if and
how the resource usage is paid for with Amps, e.g. the use could've
been for free and I could just reward them later with Pecus or maybe
I'd agreed beforehand to pay a specific amount.
That is, the settlement mechanism is de-coupled from the underlying
resource usage and the base policy would be for free usage as long as
you were trusted for it.
And, finally, Amps are complementary to Pecus:
http://thruflo.com/2009/06/09/tav-describes-pecus.html
Amps and Pecus are a bit like yin and yang. Amps provide a shared,
"objective", present-oriented, grounded base for exchange and trade —
a reflection of the Natural Commons if you will.
In contrast, Pecus provide an individual, subjective, future-oriented,
values-driven base for directing social forces — a reflection of the
Cultural Commons if you will.
Anyways, that should be enough to get going for now — let me know what
you think! Thanks.
--
love, tav
plex:espians/tav | t...@espians.com | +44 (0) 7809 569 369
http://tav.espians.com | http://twitter.com/tav | skype:tavespian
--
You received this message because you are subscribed to the Google Groups "espians" group.
To post to this group, send email to esp...@googlegroups.com.
To unsubscribe from this group, send email to espians+u...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/espians?hl=en.
Reply all
Reply to author
Forward
0 new messages